Sorry to break it to you folks but if you hated your data being with Google, you are in for a surprise on how atrociously bad Microsoft is at privacy and security. Multiple close friends who have worked at Microsoft tell me that search history data - who is searching what, is basically sitting in systems with ACLs so bad that 20-30K employees have access through transitive membership of groups. To access a customer's data you just need to know token which is logged everywhere and is apparently very easy to generate. I have heard horror stories from them about privacy incidents which never went public.
From what I know about Google, they are serious about least privilege type of stuff internally and employees dont get arbitrary unbound access to systems or data.
> Both are evil but I agree that Google is less evil than Microsoft.
I'd agree they are both evil in that they both seek to leverage your personal data against you for their own gain, but it sounds like Google does it while being more protective of what they've taken from you while MS also leaves you vulnerable in new ways through carelessness.
I'm not too worried about Bing though. That's nothing compared to Microsoft's access to your computer at the OS level. We know anything we do online will be seen by others, but being able to snoop on our personal files and log our keystrokes is a whole lot worse than knowing what we type into a search engine.
However improved/more popular Bing becomes, eventually Bing will end up in the same position Google is in now where they'll have to make their web search worse in order to constantly shove ads in your face instead of returning results that are useful, and slowly SEO spam will adapt to pollute Bing's results further. Google may not even mind if Bing starts gaining users. At this point Google can collect so much data from android devices that I doubt they need google search to peer into our lives like they used to anyway.
Yes, exactly, I already swallowed the poisoned pill when I installed Windows. Although arguably some ChatGPT queries are far more personal than even having access to your entire filesystem.
It’s not the company so much as it’s the corporate conformists that swarmed the valley. What happens when the carpetbaggers are left holding the bag? Will they actually build things that matter?
> Azure is the worst of the 3 cloud providers. Horrible developer experience and documentation and reliability
Don't forget security! Multiple serious and extremely trivial cross-tenant security exploits only in the past couple of years. AWS and GCP have had none ever, the much smaller OCI has had one.
Both are driven by a profit motive, but Google takes some forms of privacy more seriously.
Lets not use emotionally driven, ambiguous terms like 'evil' to describe these kinds of things. Google is not, for instance, conspiring to commit genocides, nor is Microsoft planning a coup in an Eastern European nation with the goal of colonizing them to use as indentured technical support.
I'd prefer if it was driven by reputation, honor and status achieved by helping humanity.
The primary driver for humans is status. Status in the military is achieved through leadership and honor. Status in academia is achieved through capability. These sectors provably show that people can be driven by things other than money.
The sad thing here is that there are tradeoffs. Nothing is as efficient or as effective of a motivator as money. But money is the driving force that is most detached from ethics.
The media likes to make a big deal out of privacy but mostly people only give a shit up to a point... so it doesn't effect googles bottom line and therefore google also doesn't give a shit.
The dominant ideology revolves around the profit motive and markets being inherently ethical and the most effective way for progress.
There’s something to it. Ideas and implementations are pressure tested through market competition, so they get refined and adapted.
The issue with advertising specifically is that it’s inherently manipulative and often condescending. On the web it has a infectious nature, making everything a bit worse.
> People are taking the piss out of you everyday. They butt into your life, take a cheap shot at you and then disappear. They leer at you from tall buildings and make you feel small. They make flippant comments from buses that imply you're not sexy enough and that all the fun is happening somewhere else. They are on TV making your girlfriend feel inadequate. They have access to the most sophisticated technology the world has ever seen and they bully you with it. They are The Advertisers and they are laughing at you. You, however, are forbidden to touch them. Trademarks, intellectual property rights and copyright law mean advertisers can say what they like wherever they like with total impunity. Fuck that. Any advert in a public space that gives you no choice whether you see it or not is yours. It's yours to take, re-arrange and re-use. You can do whatever you like with it. Asking for permission is like asking to keep a rock someone just threw at your head. You owe the companies nothing. Less than nothing, you especially don't owe them any courtesy. They owe you. They have re-arranged the world to put themselves in front of you. They never asked for your permission, don't even start asking for theirs.
The purpose of the government is to legislate in order to align those kinds of goals with profit. Ex. taxes on emissions, fines for human rights abuses etc. This is what Adam Smith meant when he wrote about the invisible hand.
Profit is ethical in abstract if you're intelligent and can see how reputation is valuable. You will have a hard time getting people to buy your product if people hate you. We're not in the state we're in because of profit motive.
It's the short sightedness that causes the most unethical behavior from the top down. 10-Qs... a man considered changing that timeline to help fix that problem but he was ostracized. Maybe one day la naranja will return.
I would be interested to see what an ethics motive looks like that gets people out of the bed in the morning to do a job they barely care about.
I have spent a significant amount of time on each and would suggest it's the only actual cloud. The others are data centers masquerading as a cloud, notice when us-east has issues for AWS, Azure had a spof dependency on a datacenter that didn't even have AZs... GCP is architecturally on another level. The others seem to be unaware of what 'cloud native' means, so if you are attempting cloud native on non clouds, you inherit their lack of foresight.
Google Cloud seemed to be in a forever beta state. It was not uncommon to find out that the sdk version I was using was no longer working so I had to update it and of course the new version had different flags with different configuration files. Just a like start-up/beta product. The worst part was if you got caught in a locked-in/proprietary product(i.e google datastore) because there was no way to get out without massive costs(i.e you can't install google datastore on a compute instance)
It's a slightly different from of eeeevil, up to you to decide which is lesser, or if the distinction has any meaning at all.
Microsoft is the capitalist evil. Google is the technocratic AI evil. The first wants your money, as much of it as they can hoover, using dark patterns if useful for it. The second wants to optimize your life using _their_ metrics, without ever engaging with you (humanity is the problem - engaging with it never scales).
I wouldn't be extraordinarily surprised if this is the case, but I do not expect it to be as lax as you suggest and given how it's presented to us it reads like someone trying to get people to not look at something. If you have things you can present as evidence that would carry water for your argument, otherwise it's random inflammatory claim on in the internet.
And I agree, Google does take security more seriously than most places.
Of the major three cloud providers, Azure is by far the one whose security bugs I hear most about. Not that this necessarily has much relevance to search, but it doesn't inspire confidence in their culture.
> Azure is by far the one whose security bugs I hear most about.
Only* about. GCP and AWS haven't had cross-tenant security bugs, meanwhile Azure have had multiple, and trivial ones to boot. If a multitenant service can have such poor security, it's doubtful internal only stuff is any better.
Agreed, but parent poster made fabulous claims based on hearsay. It could be true. But it also might be a motivated competitor trying to distract their rival as well.
Related question, if security is so bad, how did they win Project J.E.D.I? (honestly curious)
Cloud security is different, especially for these government contracts. Everything has to be certified to a higher standard and the staff all have to have clearances. A clearance gives you a nice comp boost, so dorking around looking at stuff because you can is a good way to get it yanked.
Oh yeah IIRC I used to look at Bing search logs in Microsoft and I wasn't on Bing team. This was 2014. Things probably have changed.
IIRC there was some basic safeguards though, like a query must have had like 10 unique occurrences or something. I also have no idea if you could tie searches to people.
I was just looking at aggregate numbers. That was really cool. It was like Google Trends but with real numbers!
I hate Big Data(tm) as the next guy in line, but if I have to choose between Google and Microsoft nomming my data I'll happily pick Microsoft every time.
Why?
For one, I've had a Microsoft account for over 20 years now (anyone remember Hotmail?). It's long past due for me to be complaining, and Microsoft hasn't wronged me in that time anyway. By comparison, my oldest Google account only goes back just over 10 years, and horror stories abound even if I've been fortunate so far. I keep all my truly important correspondance and login tie-ins with Microsoft (read: my Hotmail).
For another, Microsoft nurtured over 30 years' worth of good will from me with Windows and Office; even though I hate many things about Windows from 8 and up, among other things, I will ultimately be a friend to Microsoft simply because they were a significant and positive part of my childhood and now my adult life.
Part of it is, but I wouldn't happily pick Microsoft even today if they were as horrible as Google. As far as I'm concerned, Microsoft is the lesser (and more familiar) of two evils so I'll happily pick them.
> Microsoft account for over 20 years now (anyone remember Hotmail?)
From memory, Microsoft used to be absolute bastards when it came to your data. They did everything they could to keep you gated in and without any control. Things like: refusing any export tools for your emails so you couldn't migrate to another provider; or, not allowing any non-web access to your email.
Then Google came along with GMail and revolutionised things: amazingly generous free data allowance, IMAP access, data download options, open standards, liberal approach to letting people set up various hacky app access, and so on. It was hugely successful for them, and forced MS to moderate their approach.
Google abandoned their Don't Be Evil strategy and sacrificed a lot of their principles (and, I would argue, a lot of their success with it), but at one point they had a positive impact and they've never been as evil as early-era Microsoft.
I remember when Hotmail still had an allowance of "only" 25MB worth of emails including attachments. No doubt a lot of the QoL improvements the free email provider scene has had stems from Google forcing a paradigm shift.
But on the other hand, Gmail was an absolute bitch to get early on unless you were lucky enough to be friends with someone who had an address already. Maybe it was justified, but I don't remember that period fondly.
And ultimately, whereas Microsoft at best improved and at worst stayed as EEEvil as before, Google went from hero to zero in just the last decade burning away all the good will they nurtured. As things stand, I'll happily throw my hat in with Microsoft here.
Dunno if it was incompetence or malice, but the straw that finally broke me over to Linux permanently was a privacy setting in Windows that appeared to work until you re-entered the menu and it wouldn't persist. In my time at YouTube though, privacy issues were an immediately fireable offense, and even when my role required data, use was gated and then still logged, and misuse was handled. That's an n=1 apples-to-oranges data point, but I personally trust Google a bit more than MS with my data.
I don’t hate Google for having my data because they have worse internal controls.
The monopolistic dominance is the issue. They have too much data and too much power, and leverage that to swallow other market categories. Today they have a deeply unhealthy dominance over the tech industry.
Google could do everything right and still be an existential problem. You say MS has data security issues, okay sure. I’d still rather have them or anyone at 25%+ of the market.
Apple isn't a monopoly yet it has huge power and domnance with safari. Do you avoid Android because it's the most common or Linux since it's got a monopoly on servers?
Google isn't forcing someone to use it. It's monopoly can disappear when something better it's there.
That anecdote reflects the data protection part, everyone dealing with Windows trying to disable tracking experienced their horrible data collection attitude
It's not much about Microsoft. At least it improved human life in serious ways (Typescript, VSCode,...). What did Google bring to my life ? Nothing except for spam stuffs from Ad, Email,...
You're responsible for your data, not Google, not Microsoft.
Uh... Maps? Open, Permissive phone OS compared to the alternative? Greatest entertainment/information system, especially one that's democraticsed producing media?
In terms dev-related stuff, TensorFlow - literally something that's made today's ML possible, Kubernetes, BigTable, MapReduce, and yknow, that good old thing called Chrome..
I understand hating on Google. They're not that great, but none of the FAANG are "great".. They all have some terrible bits, some good bits. However, to say that Google has not provided anything except "spam stuffs" is very wrong.
Search alone has changed the entire way that humans consume information. I would say that today's internet landscape, for better or worse, would not exist without Google.
Search alone has changed the entire way that humans consume information.
I believe this thread would have a much less dramatic headline if Google didn’t itself kill Google Search ten+ years ago.
Also, it bought most of the examples listed, not made these. It just saw an ads/dm opportunity.
I also remember using Opera on its last original engine and those “Get Chrome” banners which made a whole browser freeze and stutter on anything but Chrome. I’ve pointed that out a few times back then on forums, but not too many people noticed it, it seems. The “our ads our performance” trick fled completely under the radar.
Real Question: What is wrong with Netflix? (The "N" in FAANG, right?) The only thing I can think of: Work life balance does not sound good. It sounds like you need to be "pro-athlete" level and are constantly at risk of being laid-off or replaced by someone better. That said, they pay silly good money. But are they evil to their customers? I don't see that.
People sometimes fault them for removing content (not their fault, everyone saw Netflix works and made a streaming service of their own) and cancelling series too quickly and erratically (true). Also people have a problem with the average quality, but seem to misunderstand Netflix's business - they need to have lots of content, one very good series per year isn't going to keep subscribers on the service. They have great quality stuff, and meh filler that some enjoy.
So all in all, disagreements with direction, but nothing wrong/evil on the level of any of the FAAN.
You forgot to mention Google's contribution in open source projects. Even their hardwares are OEM unlock-able. If it wasn't for Google, smartphone market wouldn't be what it is today.
in the long history of ML, tensorflow is not that significant. realistically, nearly everything it did was available in other systems before. It was really just a grab for Google to try to get everybody to use their framework. But it wasn't a good framework, and it evolved terribly.
MapReduce was not and isn't something that made ML possible. It made data engineering at scale possible.
> in the long history of ML, tensorflow is not that significant.
I don't necessarily disagree, but that's only from a technical pov. However, I think the release of it as a open-source library made it so that it's far easier to learn the tech that already existed. Vast majority of university courses use it to teach lots of different concepts that were far more "mathy" before.
> MapReduce was not and isn't something that made ML possible. It made data engineering at scale possible.
Yeah. Maybe I worded that wrong. I wasn't saying it made ML possible. Just that it's a huge contribution to the Open-Source tech overall.
mapreduce wasn't a contribution to open source tech, except in the sense that it was published with enough detail for Doug Cutting to make an open source version.
Both MR and TF are net-negative for the outside world. I think more unis teach pytorch than TF now.
You literally said "made human life better" ... it doesn't matter if you use them or not.
Also, you don't use a Phone? or Maps? Really? Do you use Chrome? Do you use Firefox, because Google's contributed a massive amount to the web standards that are implemented in both Firefox, and Chrome.
this setup and response seem a little silly to me. "what did the romans ever do for me?" "roads? aquaducts?" "ha! I don't go anywhere and drink well water!"
You have already lost the argument. I'm sorry but nobody would care about anything you say because that almost just does not apply to anyone else.
Also the fact that those products have no effects on you has nothing to do with whether a company is important or if their products are significant or helpful to the vast majority of human population.
And to be a bit cynical, it feels the only purpose of this comment is to show that you have a unique lifestyle, not to contribute to the discussion constructively or based on consensus. (Of which there are many on HN)
Actually i can answer more clearly on the why Typescript (example) improved human life.
I used it, and also guided other juniors to implement real world products which're running in production to help others human life in a very short time, in a maintainable way.
This was impossible before with old toolings.
Is it more clear to you ? I'm also sorry for not so clearly said, but that's how conversation works ;)
While it might be fashionable to hate on Google, the question "What did Google bring to my life" seems really odd. I can't believe I have to say this but search, and more importantly good/fast search, is to Google's credit. Typescript and VSCode all have, and will continue to have alternatives, but for a long time, Google was (some may argue it still is) the bleeding edge of search.
Anyway, there might be a generational thing at play here, if so, I meant no condescension.
What an absurd statement. I don't have especially strong feelings for Google in any direction, but I can't deny that, even if Google had only done search or Maps, that alone would be a massive boon for society.
The article starts, "Nestled in the hills of Mountain View", which is worrisome, because as anyone with even a passing familiarity with Silicon Valley geography knows, Mountain View is mostly flat. And Google's Mountain View offices are located on reclaimed baylands, which are especially flat.
I wonder if the article authors were thinking of Xerox PARC's offices in Palo Alto's hills? Or the road named "Sand Hill Road" that used to have some venture capitalist offices? Other than those offices I can't think of any significant hill-based offices in Silicon Valley.
I do think there is worse bullshit to be concerned by than whether or not a British tech writer has physically stood in Mountain View, though. I thought the beauty of the internet was in reducing the significance of that sort of thing.
Truly for that you want to head to Milpitas. I wonder how the economist would describe that. “The fragrant town of Milpitas, adjacent to the unappreciated Asbestos Gem of Alviso…”
Although Google, Maps, Youtube are of daily use they are monetized exclusively by advertising which is annoying and hated by many.
It has been many years since Google has launched an innovative new product.
I don't think ChatGPT will gain daily traction after this hype. Anyway we could say that MSFT and AMZN have demostrated more power to innovate with different business models (not only adv) and products.
The GOOG stock has a PE (Price/earning) of 23, while Coca cola 26. So the stock market expect higher growth from CocaCola than Google. Quite surprising.
Google stopped being innovative the second their "20% Time" got gutted:
> Employees were encouraged to spend up to twenty-percent of their paid work time pursuing personal projects. The objective of the program was to inspire innovation in participating employees and ultimately increase company potential. For Google's part, Gmail and AdSense both arose out of side projects. In 2013, Google discontinued 20 percent time.
What has Google released since 2013[0]? Stadia?
Then you tack on that Google Search as of today is basically unusable/broken with people needing to append site:, double-quoting everything, or enabling verbatim search to get high quality results. Heck, on top of that, I also have a uBlock configuration that removes 280 domains from Google Results because they're all auto-generated spam of Stackoverflow answers.
There is something deeply wrong with the culture inside Google.
Google Workspaces originally released in 2006 (inc. Sheets, Docs, Drive was 2012, Calendar was 2009, etc), it was rebranded "G Suite" in 2016 then rebranded "Google Workspaces" in 2020. Rebranding an existing product isn't innovation.
Google Cloud is a worse clone of AWS, is that innovative? Redmond-style photocopying?
I'll give you TensorFlow, but we're talking about that being their death knell, so congratulations about signing their own death-warrant?
Agreed. I can actually just figure out how to do simple things in GCP console by clicking around the UI, including setting up networks, doing IAM, and getting stuff talking to each other, most of it without ever leaving the web browser or even peeking at documentation. Good luck doing that with AWS.
It was when it came out in the form of App Engine. It was when it released managed Kubernetes and then managed containers within Kubernetes (Cloud Run using Knative behind the scenes).
Also:
- Kubernetes has been incredibly transformative and adopted by all other cloud providers
- Golang - the language of choice for a lot of cloud native tools
- Flutter - it may not be the incumbent choice but it makes cross platform development a breeze
- Google Photos - no other photo platform comes close to it
Apple photos now offers full user controlled encryption. Google Photos has no such protections and never will. Apple may not be saints, but the fact is there’s one photo service where the photos remain yours, and another where a photo of your kid can get your entire digital life nuked with zero recourse.
Perhaps the original author should have specified profit generating products. Go, Kubernetes, Tensorflow are all open source and free. Google Photos makes a profit?
>I also have a uBlock configuration that removes 280 domains from Google Results because they're all auto-generated spam of Stackoverflow answers.
Can you describe exactly what you mean? If there is a way to hide specific domains from showing in search results, it would be an incredible boon for me.
They added so much red tape and bureaucracy to it that it is a distinction without a difference. Innovation went from a stream to a trickle after 2013.
It went from "spend 20% on whatever project you find interesting!" to "spend UP TO 20% on a project your manager approves of and can defend from their manager, with routine reporting on the status, direction, and potential outcomes."
Employees went from wanting to work at Google for the "20% Time" to not using it because it was a huge hassle that could risk their career advancement, if your 20% time project didn't pan out. It went from a perk to a gamble almost overnight.
The result in the same either way though: Google stopped innovating.
Some have postulated that this might be Google's Blackberry moment. On the flip side I wonder if its actually more like its Dropbox moment.
The press loves a David and Goliath story; the young disruptor versus the stodgy disrupted.[1] At that time, Dropbox was David, coming out of nowhere in a hurry and had a product that seemingly should be so obvious for Google to launch but hadn't - making Google seem antiquated and slow.
Fast forward a couple of years (aka a decade) and Dropbox is still going ok but Google Drive is by far more ubiquitous.
Google makes fuck all revenue from Drive, it never mattered. EVERYTHING at Google, most notably the cash fire that is Cloud, is funded by Search. This is the first time that is in danger and over 25 years they never came up with a fallback plan. This is an unmitigated disaster and purely the result of mismanagement.
Also the comparison is between Google versus Dropbox (the underdog David that's analogous to OpenAI), not Google versus Amazon or even Google versus Microsoft.
Google is 51% of office productivity software market versus Microsoft. Microsoft is at 47%, Dropbox isn't even listed. [1]
Dropbox has a decent share of the personal cloud storage market but has really failed to expand beyond that vertical or become significantly profitable. [2]
The point is Cloud is still 100% subsidized by Search, and if Search loses even 5% of market share in the next 12 mo it is a disaster as GCP cannot fill the gap as they are still losing money. The velocity at which they are losing money is immaterial to the fact they are not nearly ready to plug a hole if Bing take share.
Good point - I think the much more typical flow for Google Drive is cloud first for most documents and only uploading occasional offline document upload.
Which is perhaps why Google didn't really focus on creating a Dropbox like product until it was forced to because the more long term profitable for Google was a solution/vision based around a cloud first experience.
Dropbox iPad refuses to keep docs synced docs that I have specifically flagged for this. I routinely open the iPad somewhere without a connection and see that the docs are a month out of date.
Not as mission critical as the desktop app, but still pretty frustrating that they can't handle this core functionality after years and millions spent on other bells and whistles.
Apple is forcing Dropbox to cripple itself because they now have to use Apple's own cloud file sync API which is hot garbage. I don't know if this is the cause of your iPad issue, but Dropbox on MacOS is about to get a whole lot shittier.
They didn’t exactly stop innovating: ChatGPT itself is using tremendous amounts of innovation that was done and published by Google. What they stopped is beating their competition to launching compelling products based on the innovation they do.
IMO this whole narrative reflects a type of human bias and the reality that we didn't evolve to reason about trillion dollar complex business.
I love chatGPT but it clearly has a ton of issues. To declare Google dead that is point is completely absurd.
I can't think of a single transformative technology that actually has a first mover advantage. The first, second and third mover seem to make all the mistakes and setup the forth mover to learn from those mistakes for free and create a monopoly.
Counting out a company that owns Deepmind is just not rational.
Nearly all of the authors left Google years ago. If the company doesn’t keep the people innovating it’s not like they’re some magic middle manager who can make it happen without them.
I think YouTube, which has a paid and ad-free version along with its Live TV offering, somehow manages to still be a sleeper hit within the larger Google portfolio and doesn't get talked about enough.
What happened to the Waymo hype? If Waymo manages to break through and become a sufficient AI taxi, then car ownership becomes optional for a majority of Americans (unlike today).
1. Car infested ergo perfect venue area for testing autonomous cars, since the infrastructure is through and through car oriented at the detriment of every other form of mobility (walking, cycling, buses, trams, trains, etc)?
2. Super hot and sunny, ergo no fog, no rain, no sleet, no ice, no special weather conditions to handle.
3. Flat like a pancake, making terrain management, coupled with the nice, wide roads from point #1, so super easy mode.
Let's see them scale Waymo to at least 3 out of the following: San Francisco, NYC, Bucharest, Istanbul, Mumbai, St. Petersburg, etc.
Yes Phoenix is ideal, that is why they've been using it. I think it indicates they are more than 5% done though, is my point. Also they are operating fully autonomous public taxis in San Francisco, but not the run of the whole city, I think its just airport to downtown. But even navigating downtown SF I think is an accomplishment.
That maths sure make it a very hard problem to solve. But jokes aside I see your point. Isn’t that always the case with this kind of issues? Doing the first big chunk is relatively easy and then the final details are very hard?
Google is an ad-driven company and they haven't yet figured out to generate ad revenue from Waymo... (Having said that, it's cool that the Mountain View office building I used to go to is now occupied by Waymo).
"Urban" includes second, third, and fourth tier "cities" with low populations and no fleet of Ubers today. There won't be a market for automated ride share for most of the US.
Ads aside, the UI of Youtube and Maps is atrocious. There is so much room for improvement here, but it will never happen from Google. It's just not in their DNA. Unfortunately, they have such a huge moat here I doubt we'll see anything better for a long time.
They used to have pretty good UI for maps until some PM needed to make their mark... Given they count individual pixels and +/-1 deltas in colors I don't think they are ever going to get any UI correctly. The initial simple UI with just a single search input box that was highly praised by minimalists was just an accident.
The most infuriating thing with Maps is that if you're looking at directions and have got the viewport where you want it and want to do something there (e.g. look at the description of a restaurant that's near the station you're changing at, or get directions to somewhere along your route), you can't. There's no way to escape from "directions mode" except hitting the back button, which resets your viewport. So you have to memorize the name of the place you're interested in, exit directions, and then hope that searching finds the place you were looking at a second ago (which, maybe in English it does, but searching for a Japanese place by name works maybe 1/3 of the time).
Generated playlists (recomendations) are always (for days/weeks) the same. Not a bit randomization. It feels like they were approved by party commitete for years.
I don't understand why they decided social network in music service.
Ah hell. CUE-lists are still too novice technology for them.
I still think that last.fm 15 years ago was the best music service.
If search gets usurped, or even a marginal but meaningful reduction in usage due to competition from AIs, then its not surprising at all that Google’s near term prospects could be worse than Coke.
Even Google winning the AI wars leaves them worse off as the operating profit potential of querying an AI vs their search index has to be far lower.
As to usage, once you start to integrate ChatGPT into your workflows it can meaningfully benefit vs traditional search. Ive been able to find information on specific programming language concepts, with generated examples, far faster than searching.
I can ask it about GameDev concepts and ask for bulleted lists or higher/lower level of detail in the answers. Information is presented in a much easier to consume manner
That being said, most stocks that are considered defensive are quite overvalued on a fundamental basis. I would consider a PE of 26 for Coke quite undesirable, though there’s much worse
If Google ends up winning in the AI war, search and ads will be significantly cannibalized, but it will have tremendous opportunities outside of search, many of them in areas that don't even exist yet.
In some ways, it might even be good. It's suffering from a kind of advertising resource curse nowadays, and being forced and able to diversify its economy would be best in the long term.
I'm not sure, it seems many companies have comparable technology at this point, and Google is not known for being great at packaging tech into products.
They had an enormous moat and edge in search, but not seeing it for this next wave of AI. Barrier to entry seems far lower
>It has been many years since Google has launched an innovative new product.
I can say the exact same thing about countless other large companies: Facebook, Intel, Microsoft, Boeing, AMD, AT&T, I could go on and on. At some point, launching innovative new products isn't really that important for a company, when many people on the services that company provides.
Google used to be considered different from and better than all the companies you list (except Facebook). Being lumped together with those companies would have been seen as an insult to Google in the past.
Times change, and companies grow and mature. Intel used to be a big innovator too, like when they decided making RAM wasn't the best strategy and decided to try making CPUs instead. They've come a long way since inventing the 4004.
It is if you don't want to wind up like IBM, i.e. - a company that could have been so much more, knew it had to change at the leadership level, and yet still couldn't get out of their own way. Apple's renaissance came because they started releasing innovative products and cannibalizing their own business.
Most large companies would just destroy themselves if they tried this. IBM is still there, decades later. Why is it not good enough to just continue to exist and generate profits, even if you aren't the highest-market-cap company on the planet?
Decades ago, it used to good and normal for big companies to simply stay the same size, make a steady profit, and return dividends to their shareholders. That's how you get a stable society, not with constant boom-bust cycles.
They are kind of forced to keep up or die now, not a good place to be in. It feels like everyone always just assumed they must have super advanced AI behind closed doors. Perhaps that was a wrong assumption because they never actually seem to *do* anything with it, but we will probably know that for sure very soon.
Google are a mature corporation. The fact they managed to innovate for so long is really a huge achievement but nothing lasts forever. The same applies to the rest of FAANG one way or another. Netflix cannot innovate out of streaming tv and movies, and now face strong competition from others. Apple has always been more about marketing than real breakthrough tech innovation and the bits of real innovation they did do left with Jobs. Amazon has flopped most of it's recent attempts (again, respects for running BOTH the worlds best retail company AND AWS). Facebook has been flailing around doing crypto currency and VR and getting nowhere. Microsoft haven't innovated since Excel FFS. Coke at least has the occasional new flavour...
I'm not "blaming" them, it's the cycle of life. Gravity eventually wins.
This is also a key reason for all the layoffs: Growing, innovating companies cannot ONBOARD staff fast enough, because their value is decided by how big they will be in the future and how fast they can get there. People are a profit centre. But mature companies' (in mature industries) only source of increased value is in efficiency, getting the same revenues with less costs. So people are a cost centre and they succeed by showing they're efficient (aka laying off workers).
Stadia worked much better than the competition. It was the first time that cloud gaming had no tradeoffs. Definitely a lot of innovation behind the curtains.
The discussion was about innovative products. An already explored idea, but executed properly, only counts as innovative if Apple does it. And even then, it only ran for a couple years, so even if we want to give points for taking an idea and making it work, they don’t pass there either.
It is depressing that people dont recognize that Google has invented and perfected a parasitic business model that has destroyed the internet (and much of tech with it) and this episode feels like just another turn of the downward spiral.
The article aims to offer a preliminary analysis of whether Microsoft can become a better parasite: Grabbing the content people generate online, paying nothing, and using it deftly to serve advertisement on the basis of private behavioral traits that are gleaned by prying open and subverting the use of all pieces of IT people use.
It is not too difficult by now to imagine alternate tech universes (Philip K Dick style) that have nothing to do with this nightmare, where more or less the same technologies empower individuals and companies and organizations rather than squeezing them dry. The combination of oligopoly status, moral laxity and political dysfunction means we are simply sitting around like sheep discussing whether a new butcher is about to get sharper knives.
I grew up using the internet as a child and when Google came out it was way more a blessing than a parasite.
The web was hard to navigate and you relied on webpages "befriending" each other and helping you navigate to similar pages. That was real navigation and it had terrible recall.
Once google became the Internet's frontend everything was one or two hops away. It was awesome, and I think that it has evolved in a way that saves time with the snippets and knowledge graph panel at the expense of taking away clicks from sites that needed to serve their ads. Although today things are getting worse with walled gardens and SEO.
A problem that the internet has not solved yet is how to keep webpages up while there's increasingly more content. Most sites were maintained at a loss by people, or leeching resources at a university or company. The closest "people" have got to fix this IMO is through P2P, but it got too focused on piracy and got a bad reputation with malware too. And I think we never had the infrastructure to serve and share webpages we visited. Sharing today is still a mess and we still depend on centralized distribution + caching (and that's dying with https, only big companies who are better off giving caching servers away to ISPs can do this).
Ads came to "save" the internet from this problem. I really hate them and I wish we had a better model, but they seem a necessary evil as people just got used to believe that all digital things are free when they obviously cost something, and that still shows today in the apps stores.
> Ads came to "save" the internet from this problem. I really hate them and I wish we had a better model, but they seem a necessary evil as people just got used to believe that all digital things are free when they obviously cost something, and that still shows today in the apps stores.
Advertising worked just fine for more than a century, ever since the invention of consumerist society. (How much of that is really necessary in a society that is pushing headlong into unsustainability is another, bigger story.)
The disaster we have now in our hands is of a more immediate and crass nature.
Think about it: there are algorithmic breakthroughs that can help everybody on the planet raise their level of information retrieval and the only discussion is which of the former or the current consumer tech oligopolies will use it best to push ads and how to double down on a bad design.
I would have agreed with you in the past. Now, I have to disagree.
>the web was hard to navigate and you relied on webpages "befriending" each other and helping you navigate to similar pages. That was real navigation and it had terrible recall.
This is exactly why internet was decentralized. A network of social connections between small online spaces people congregated in was essential to find what you wanted. It was somewhat of a replication of the old world - where in order to find an answer, you first had to get into relevant circles (eg. a local club, university, etc). You didn't just get an answer to your question, you inevitably had to sift through a lot of other content and learn about new places.
Google cut through all of that. Ideally, you could just get the answer you wanted, and not even see the rest of the site. The first order effect was that everyone's life get better. The second order effect is that it killed the old decentralized web, because random discovery nearly died down.
LLM are only the next step on this, but I don't think they change much. I think it's mostly going to damage reddit. A lot of reddit questions are an attempt to find an answer to something that google is too dumb to find - but llms may be able to.
The defense to this - are chatrooms, like discord. Sometimes I see people complaining that so much information is now 'hidden' on discord. But that's the exact point. Making information hard to get means people are forced to interact with each other. This creates incentives to contribute.
That's the future. Chatrooms may be replaced by voice based vr, hard to say. We already passed the peak of public information dissemination, and are going back to old style "ask at the university", just more decentralized and online.
I don't think LLMs will replace Reddit. Reddit is mostly a social network for people with similar interests, or for funny cat pictures. LLMs don't provide that human connection.
Don't tell me you forgot Yahoo, Excite, Altavista, Lycos, and all those other search engines existed! Lycos also had hotbot and tripod.com, back when homepages were your internet presence instead of social media.
Maybe Google had results that were a bit more relevant and won that competition, but they didn't Change Everything.
Google changed everything but nowadays it feels as bad as Yahoo was at the time. Google search is now optimised for ad $$$, not for finding what you're looking for. If Google hadn't existed then maybe we would have got competitiveness in the search space.
It would have been another hegemony. Yahoo was google then. AskJeeves, AltaVista and many others were not ever going to be competition. Yahoo was the stronger one then and it failed.
Who do you think it would have been competitive for?
It is already back to that. I use scholar.google.com, semanticscholar, reddit search, chatgpt, public data sets, as well as all the search engines everyday. No one has everything indexed properly.
The GPT LLMs just raise the bar for information retrieval. And so it goes.
> It is depressing that people dont recognize that Google has invented and perfected a parasitic business model ... The article aims to offer a preliminary analysis of whether Microsoft can become a better parasite
In this model, are end users who use it to search for information parasites as well? After all, they are consuming content that other people generate online, and usually pay nothing for it. And they love to look for — and find — what they are interested in. If Google didn't satisfy that need well, people wouldn't have preferred it, and it wouldn't have grown.
Users pay with their data. That is the whole point of the business model and it is obviously lucrative enough so that the entire tech infrastructure (devices, OS, browsers etc) can be repurposed to be a user data collection channel. If you want to find a real accomplice that is essential for the model to work, it is not the users, it is the advertisers. It takes two to tango in the adtech market.
I don't dispute that search (in its various incarnations) is an essential service in a digitally interconnected world. There are countless ways to pay for it (as a digital public good, as a user subscription etc) that are fundamentally better than what we have. It is also obviously true that some decades ago Google innovated technically. A lot happened since and it wasn't positive. Normalizing it simply prolongs the agony.
> Search should be democratized by governments. It is central and should be for everybody. No control.
I am confused. Are you suggesting that governments should nationalize the search part of Google, Bing, DuckDuckGo, Yahoo, etc.? Or that they should build their own high-quality search engines? And what do we do until they have?
> No control.
Why do you think governments wouldn't control search engines if they were responsible for providing them?
> It is central and should be for everybody
How is the current model (Google, Bing, DuckDuckGo, Yahoo, etc.) not for everybody?
The role of a government in a capatilistic society is to basically do the stuff that isn't profitable for corporations to do without exploiting the population. For example build roads, because if a corporation did that there would be tolls everywhere and it would be hard for poor people to improve their living standards if they couldn't economically move around. That seems to have happened on the internet. The internet is in the commons and everyone expects it to be free, the problem is that the infrastructure to navigate it was built by corporations and now we have to pay tolls through ads and trackers
Nobody prevents government(s) from creating a search engine. Actually nobody stops anyone from creating a search engine. Also nobody stops anybody from using another search engine. The internet is the most level playing field and is more competitive than any other industry. Nowhere else is it as easy to create something new as a business and to switch services as a customer. If anything it's the customers fault and by extension a lack of education. So if you want to blame government then blame it for keeping its people dumb. And there you actually have a point.
Search provided by the Russian government to the Russian people, Chinese government to the Chinese people, British government to the British people, X government to the X people..
Google started out noble, but was poisoned by its own success and by extension, poisoned everything it touched. The web, as a platform, is nearly dead. Walled garden phone apps are almost completely dominate. The revenue model is on such shaky ground that almost all remaining websites either went clickbait with a pox of ads or paywall. I would not be able to list more than 5 websites that survived Ads and SEO, perhaps only Wikipedia.
These LLMs will be the end. They will capture enough of the user attention that the few remaining ad supported sites will be wiped out. “Content”, that low quality grist every site is filled with to capture users for a few seconds before they click on will become worthless, no different than the low quality crap the LLMs spew out.
Finally, we will stop trying to keep the bloated corpse of the web alive and will move on. It will be the death of a dream. At least we can all let go and focus our energies on a new dream. Whatever that dream is, I hope high quality, human created art, knowledge, opinion and creativity is at the center of it.
I wish I could find something to disagree. If there is any hope it ultimately comes from the fact that the technologies we talk about are at various stages of commoditization. Maybe at some point the gap between what-is and what-can-be will be so stark that it will set in motion positive forces that are currently dormant.
"sitting around like sheep discussing whether a new butcher is about to get sharper knive"
Speaking of that, I might as well provide some supprting evidence from a different part of the tech industry. Most people don't know that right now Blackrock and Coinbase are attempting a coup on the US dollar by seeking access to central bank liability for their stable coin? The Coinbase CEO even said it out loud that this year his stable coin will become the de facto central bank digital currency. I suspect that they have indirectly bribed someone at the Fed (how else does that work? esp since the Fed chairmen has come out strongly against stablecoins) and are basically trying to get the ability to get Fed loans at the Fed rate and loan out at the commercial bank rate without being a bank and by using a coin which they can mint themselves, so all the profits from interest create more reserves and the cycle keeps going. It is literally insane. The threat from Google and Microsoft to society pales in comparison. I may not have my understanding 100% correct but all banks are up in arms about it and the average person on HN has no clue this is happening. Just Google Blackrock, USDC and the Fed. The banks are calling it a backdoor CBDC.
Yes, Blackrock and Coinbase have partnered up to provide crypto trading access to Blackrock clients but how are they attempting a coup on the US dollar?
As someone else said, the problem is ChatGPT lies straight to your face, whereas at least Google's answers are based on structured data by someone. It's traceable, whereas I would not trust ChatGPT ever to tell me the correct temperature to cook a steak.
If it's citing all it's sources, it's not ChatGPT, it's a natural language interface around search, which Google has been doing for years including with voice...
"Get answers instead of being overwhelmed by options. Bing looks at search results across the web and summarizes responses to your specific questions and needs."
There's summarised cited searches and then there is the more ChatGPT option which is also available.
Presumably they have bolted on something that cites sources. This doesn't mean that the things ChatGPT says are true, just that you can check some sources that are plausibly relevant.
Traditional search practically forces you to read the sources directly. With ChatGPT, people will just assume its citation is correct due to laziness.
I've seen this with many human published articles on the Internet too. The litany of citations fool everyone into thinking that the article is credible, but when I actually read the citations, they don't support the article at all.
Specifically just the 400 best/recent ones. Google will never return more than 400 results for a query despite reporting ~80,000,000 results. If you click next enough it will run out after showing you ~38 pages of 10 results or almost 4 pages of 100 results.
This is the intended behavior. It kind of ruins serendipitious web surfing via search. I look through 4 pages of 100 results in less than a minute and the web as google search presents it just feels so shallow now.
Even this kind of project only helps so much. ChatGPT has absolutely no understanding of anything, it just regurgitates words and sentences in a way that sounds like it knows something. It's next sentence might be completely, dangerously wrong. You can't even claim that it lies to you. It needs to have an inkling of the truth to be able to lie in the first place. It's basically a well-trained parrot that has learned when to say the right combinations of phrases it's memorized at the right time in the right order to earn a treat. It has no idea what any given word means, let alone what it just said.
If someone built something that parsed sentences, stored facts, learned relationships between them, learned enough language processing to make connections between them in order to be able to say something new about the subject it didn't already know before-hand, then I'd be on that train. Right now I only see misery in our future if we just work on honing ChatGPT and it's ilk to better simulate BS.
I don’t know if you saw screenshots of the Prometheus model inside Bing. But it presents the sources with all its answers. It is no ChatGPT but a new model created for search.
The SEO content served up by Google search nowadays isn't much better, and from what I've seen I don't think a lot of it is being generated by someone either.
Perhaps a bad example, since ChatGPT consistently gets the steak temperature right (or at least gives the same values as Google). Internal temperature of about 130-135°F (54-57°C) for medium rare, etc.
It's just one data point but it may indicate the commenter is miscalibrated on what ChatGPT answers well and what it doesn't.
It reminds me of when Wikipedia was new and we were told repeatedly that we couldn't trust it (by teachers, articles, etc.). It didn't matter if we could point to studies that found Wikipedia had similar accuracy to other encyclopedias. They objected on the grounds that anyone could edit it and anything on it _could_ be wrong and there's no publisher or paid editors to ensure accuracy.
ChatGPT tends to do well on common questions, where the answer is plastered in hundreds of articles across the internet. The internal cooking temperature of a steak is a great example of this. There are many other types of questions it fails at.
A better example of where you shouldn't trust ChatGPT is asking it the distance between two cities. It'll get it right for a few common pairs of cities (eg London to Paris), but it'll give you a wrong answer for most less common pairs (eg London to Cairo).
The difference between ChatGPT and Wikipedia is that Wikipedia actually cites its sources.
Also ChatGPT is only knowledgeable about general things, but even there it makes errors. It's basically a very complex scraping algorithm, the more interesting part is the language generation; even then this stuff seems at least unethical if not illegal since it's using other people's work/research without citation.
In many cases Wikipedia's source links are dead. And nobody bothers to update them. Or they cite some random book that I have no way to get and read myself.
Who checks if that book actually exists and is not made up? Especially if it's not a highly politically charged topic.
And then there are circular citations. Someone posts unverified/false info on wikipedia. Then it gets cited by some blog or other media that wikipedia considers trustworthy. And that then gets added to wikipedia as a reliable source.
Agreed. I wouldn't use ChatGPT for this because Google is faster and more reliable. I just thought the example was a bit odd since ChatGPT gets it right, and it's a type of question it generally gets right.
The correct temperature to sous vide a steak depends on the desired level of doneness. The following temperatures are guidelines for cooking steaks to different levels of doneness:
It's important to note that cooking steaks at low temperatures for extended periods of time can result in a more tender and flavorful finished product, compared to cooking steaks at high temperatures for a shorter amount of time.
The problem is that if you don't already know the answer, it's hard to tell the difference between ChatGPT giving you the correct information and ChatGPT authoritatively spouting the wrong information.
Here's a chat I had today with ChatGPT:
Q: Who was the first US president with a last name starting with R?
A: Rutherford B. Hayes was the first US president with a last name starting with "R."
Q: no, a surname starting with R
A: Rutherford B. Hayes was the first U.S. President with a surname starting with "R."
Q: no, Rutherford is his first name
A: The first US president with a surname starting with "R" was Richard Nixon.
Q: no, Richard is his first name
A: The first U.S. President with a surname starting with "R" was James A. Garfield.
Here it's obvious ChatGPT is spouting nonsense, but in other cases it may not be.
Just for fun, I tried again today with a new chat:
Q: Who was the first US president whose surname begins with "R"?
A: The first US president whose surname begins with the letter "R" was John Quincy Adams.
Q: Adams does not begin with "R"
A: I apologize for the mistake. The first US president whose surname begins with the letter "R" was Rutherford B. Hayes.
Q: Hayes does not begin with "R"
A: My apologies for the mistake, the first US president whose surname begins with the letter "R" was James A. Garfield.
Q: Garfield does not begin with "R"
A: I apologize for the error. The first US president whose surname begins with the letter "R" was Richard Nixon.
Q: Nixon does not begin with "R"
A: I apologize for the mistake. The first US president whose surname begins with the letter "R" was George Washington.
Q: Washington does not begin with "R"
A: I apologize for the mistake. The first US president whose surname begins with the letter "R" was Franklin D. Roosevelt.
Q: Wasn't Theodore Roosevelt president before FDR?
A: Yes, you are correct. Theodore Roosevelt was the 26th President of the United States, serving from 1901 to 1909. He was the first US president whose surname begins with the letter "R." Thank you for pointing out the error in my previous answer.
So at least it's capable of parroting the correct answer after you deliver it. :-)
Completely agree. Heck, even the current 'waiting list' page (bing.com/new) looks so outdated and a mess. I don't know whether it is the font, the colors, the composition or margins..
Bing itself looks like a mess too. Microsoft truly has no taste.
The scary real threatening and killer thing would be for chatGPT to automanage my tabs. Like a well-meaning shepherd, perhaps create new pages on the web that by a combined virtue can take the place of whole groups of open tabs, being equivalent convenient replacements of the information they once held.
As long as it's easy to see the list of web results so I can go there instead of just relying on the model. Also because there are reasons to go to sites other than just getting a summary. Like maybe you need or want to use that site to do something.
People seem to forget: AI is about data, the ability to process it, and having a few smart folks to do algorithms. The third part is actually the smallest and easiest part of the trilogy for AI.
Google, has all three. The real question is can they not shoot themselves in the foot while doing it.
Also SEO will always target the market leader.
As far as Bing's results: I keep thinking Google is trashy. Then I use Bing and I remember just how good Google is. That's my personal opinion. I am NOT going to claim they are SEO, and spam immune, far from. But they seem to do better than Bing in getting what I mean right.
ChatGPT may help MS, but is it a bandaid? Is it just a good PR bump? Who knows... Displacing Google is a large task.. and not one I'm sure I want MS to win. But I look forward to them trying, if only to make Google do better.
Google is without a doubt worse for me than ever. But is it really google? Or is it just the absolute cesspool of SEO-optimized ad driven garbage that represents a growing portion of it?
The other searches are worse. I too have tried. But in general I'm frustrated with Google now more than I've ever been.
This is really the core of my disappointment in Google: their poor handling of malicious SEO. Having humans go over shared site blocklists that people use to improve results, hand verify each entry, and remove those sites from Google results would go a long, long way.
I think this will be a key differentiator, Google has been so obsessed with automation that it's painted itself in a corner when that automation is inadequate. Microsoft doesn't have a problem with doing things that don't scale when needed.
> Then I use Bing and I remember just how good Google is.
I use a lot Ecosia (which gets results from Bing) due to them allegedly planting trees and being much more heavily into privacy.
My experience with Ecosia, thus Bing, is that for the overwhelming majority of searches (90% plus) the quality is comparable.
When you get into more niche searches Bing offers better results 1 out of 5 times and worse results 4 out of 5, with a worrying amount of those 4 being almost complete misses.
I still need to get onto Google occasionally when I don't find what I was looking for on Ecosia.
> People seem to forget: AI is about data, the ability to process it, and having a few smart folks to do algorithms. The third part is actually the smallest and easiest part of the trilogy for AI.
MS has them all three too.
And since, as you say, the algorithm part is not the most difficult part, others may enter the playing field soon (we already have DDG and Kagi, for example, with Kagi now experimenting with LLMs too).
The compute side needs to be solved with money. That may be a hurdle for DDG/Kagi.
MS has the money. Data is by far and away the hardest part, and DDG's own privacy policies etc will hurt it here, alas.
Personally, I think it is Google's race to win. But they have to DO it. If I start getting better results off DDG or Bing, I'll switch. I'm fickle, and own no loyality to Google.
I just want my StackExchange answers faster, before Google Coder, and Co-Pilot replace me. /s
I don't understand this hype and I feel like I'm looking at different products than everyone else is. There are very few complaints I have about Google that I think this technology helps solve, and for most of my complaints, getting summaries of searches makes the situation worse, not better. To be completely clear: even if the AI was perfect, I don't know that I want even an actual human being to sit down and summarize an answer to my question rather than show me a list of search results.
The problem with search is not that our answers aren't summarized well, it's that the quality of information returned for those searches is getting increasingly worse, and we are getting increasingly worse at categorizing or filtering that information in any useful way. And LLMs pulling information in and summarizing it for me is... not helpful? It's summarizing the same garbage, except now sometimes it also summarizes it wrong.
But it's not even an issue with the quality (although the quality of information from LLMs is also pretty over-hyped I think). Conceptually, I don't know that this is a product that I would ever want. I can't think of any time where I've sat down to do a search on Google or DuckDuckgo and thought, "You know what I want? I want these results presented to me in a less structured format using natural language and with less granular knowledge about where each specific statement is coming from."
At least Bing seems to be trying to do inline citations in some of its answers, which is a step up over Google's AI announcement, I guess?
Maybe I'm just in the minority on that. Users seem to like this a lot. But my ideal version of the Internet is one that decreases the number of abstractions and layers and summaries between myself and primary data rather than increasing them. My ideal Internet is a tool that makes it easier for me to actually find things, not a tool that increases the layers between me and the raw source/information that I'm looking for. I already have enough trouble needing to double-check news summaries of debates, events, and research. Getting another summary of the summaries doesn't seem helpful to me?
I can think of some ways where I might use an LLM in search, even really exciting ways where maybe it could help with categorization or grouping, but it doesn't seem like Google/Bing are interested in pursuing any of that. I look at both the Bing and Google announcements and just think, "why are you making it worse?" But who knows, maybe the actual products will sell me on the concept more.
From my point of view the only usage of LLM is generation. Such as writing peer reviews, self reviews, OKRs, when I already know the truth and can edit out any errors. I will never trust a LLM's answer to something I don't already know.
There seem to be two different directions for innovation here.
The first is a little more mundane: LLM embeddings. OpenAI currently offers an API that turns sentences into coordinates for a point in some 1536-dimensional conceptual space such that two points are close together if they are conceptually close together. This is insanely powerful. For example, you can generate captions for a bunch of images and store the embeddings for them. Then, you can look for a "picture of a rabbit eating a carrot" by turning that phrase into a 1536-dimensional point and looking for the nearest points around it. Basically, it blows open search technology for everyone. You no longer have to deal with synonyms, idiomatic phrases that mean similar things, misspellings etc - the problems you'd run into when trying to implement simple text search using traditional techniques. It all gets simplified to generating coordinates in some hyperspace and looking for nearest neighbors. This is a total game changer.
The second direction is ChatGPT. Sure, if you want to read a detailed analysis of the demographic situation in China, you'd prefer an article written by an expert. You would still use a search engine, pick a search result and do things the way you do them today. However, there's an entire collection of things that can be answered directly by ChatGPT. For example "how many mins should I hard boil an egg" or "Can I take NyQuil when I'm stoned" or anything else where you really just want a single sentence answer. Today, you launch a browser, search for what you want, skip past the first 10 advertisements, look for a site that seems reasonably reputable, click through all the GDPR warnings, scroll past the banner ads and the SEO optimizing bullshit text to find that one sentence that you wanted all along. Or, you could ask ChatGPT and get an answer instantly. (assuming chatGPT is good enough eventually).
It's hard to predict which of these two technologies will disrupt the current status quo in search. Neither might. But we haven't ever been this close to a level playing field in search since the 1990s. The excitement is hard to resist.
> The first is a little more mundane: LLM embeddings.
You know Google has been doing this for years now?
> However, there's an entire collection of things that can be answered directly by ChatGPT. For example "how many mins should I hard boil an egg" or "Can I take NyQuil when I'm stoned" or anything else where you really just want a single sentence answer.
Google has been doing this for years via search cards, which are AI generated summaries of website information.
> You know Google has been doing this for years now?
Of course.
What’s changed is that before now, they were the only ones who could do it, vs now, everyone can do it. So this technology only got deployed where someone could get a promo out of it whereas now, every to-do list app, dating app, and even a reasonably sophisticated nigerian prince can find a place to deploy it.
Think of what gcc did to software tool chains, Apache to servers, Linux to operating systems and to a lesser degree, blockchains to distributed databases.
> Google has been doing this for years via search cards, which are AI generated summaries of website information.
Yes. But now, someone else can do it too. And if that someone else does a good job, Google just lost the opportunity to show advertisements to all of those “searches”.
Doesn’t mean that anyone is about to beat Google in terms of sheer talent and experience with this stuff. But a very hungry and determined community of entrepreneurs just got hands on something that’s about as good as Google’s secret sauce and they’re about to run wild.
So LLM embedding are an actual useful thing that could actually improve search. Categorization is a real problem that AI could help solve (particularly with search queries). But that's not a new category of search, it's just a question of whether the current LLMs would be better than whatever Google is currently using to make the same inferences. And the direction Bard and Bing seem to be going with these giant models is the conversational direction, and where that's concerned:
> there's an entire collection of things that can be answered directly by ChatGPT. [...] assuming chatGPT is good enough eventually
I am a lot less impressed with this. And I know I'm an outlier and plenty of people are shocked at how good GPT is at this kind of problem, so I am constantly second-guessing myself and thinking to myself, "are we using the same product?" Because I think ChatGPT produces really bad quality information. It's cool, it's wildly impressive, it's a massive achievement and an incredible milestone for AI, but 'cool' is different from 'useful.'
Leaving aside the problem that answering simple questions is a very small subset of what search is used for, and isn't on its own probably a big enough category of questions to make me change search engines, the bigger problem is that the current state of ChatGPT seems to be wildly inconsistent about what it knows and what it doesn't know and I don't have a way to pre-predict what categories of information it's safe to ask about. And the only way for me to verify the answers it gives me are to... double check its work with a real search.
I would not advise anyone to ask ChatGPT for advice about what drugs are safe to take while high, that seems profoundly unwise to me.
So it's a bit like Instant Answers. Google has been trying to auto-answer questions for ages, and in practice the only time it's ever been useful for me is when it's extremely predictable and when I know that a category of question will only ever have its answer pulled from one site and where I know what the format of that answer will be.
Unpredictability is generally a quality that I try to avoid any time that I am using a computer. One of the primary strengths of a computer to me is specificity and predictability. And so the bar here is really high. The question I ask myself is, "would I want to replace a search engine with a human assistant?" And I think the answer is no, I feel like that would be missing the point of what a search engine is. And ChatGPT gives worse answers than a human assistant would, and its sources/knowledge is just as unpredictable as a human's would be if not worse. So, I also don't want to replace my search engine with ChatGPT.
It could get more accurate in the future, and if it does then maybe my opinion will change then, but... it's hard for me to get excited about using a worse product today on the promise that it might get better in the future. And I guess it's accurate enough that a bunch of people keep telling me that they're saving time when they use it, so maybe I don't understand what I'm talking about. But I just don't see how people are reaching that conclusion unless they're either asking questions where they don't actually care about the accuracy or unless they're just rolling the dice and trusting that ChatGPT won't accidentally poison them when they ask what drug combinations they can take.
First, you have people who don’t have any idea about how any of this works and are generally far removed from the tech communities. They did not see this coming. To this community, ChatGPT is a fascinating toy. It’s not perfect, but at this point everyone is conditioned to believe that thinks will somehow get better. This group is excited.
Second, you have the tech community who is skeptical. This group of people sees everything that’s wrong with ChatGPT and see the magnitude of work needed for anything to even start approaching Google as a credible threat. This group is generally confused by the excitement going around because it doesn’t seem warranted, and worse, the excitement is being seen in people who should know better. There’s a range of responses from being dismissive to feeling like they’re being gaslighted.
Third, you have the people who are filling up the YC summer 23 applications. They’re all looking for big unsaturated markets to build a pitch deck around. ChatGPT looks like a very promising sign post that says “look for ideas here” to this group. They are excited. Most of them will fail. But if anyone survives and manages to thrive, where will they be 10 years from now? How about FDA approved chatbots integrated with a blood pressure monitor and thermometer that can take a first pass at routine prescription refills at 1/100th the cost of an equivalent doctors appointment? How about live translation of television events synthesized back to the original speaker’s voice? How about video game engines that can synthesize music loops dynamically to keep up with the gaming pace of that particular gaming session?
Sure you might say, but none of that “dethrones google”. My response to that is - what role does a text based internet play in daily life 10 years from now? Everything on the internet today went something along this path: primary research -> classrooms -> textbooks -> niche blogs/forums -> mainstream websites.
10 years from now, would you bet against primary research -> ChatGPT ingress -> widely deployed ChatGPT model? What role do ad driven websites play in this chain? What role does a search engine play in this chain?
Sure, it doesn’t make the internet nor search engines obsolete. But it changes how we do things. Potentially in a very big way.
> what role does a text based internet play in daily life 10 years from now? [...] 10 years from now, would you bet against primary research -> ChatGPT ingress -> widely deployed ChatGPT model?
I suspect I'm a little bit more skeptical about this than other people might be? 10 years out is a long way to predict and I'm hesitant to try, but I might be primed wrong looking at voice assistants/video/etc... where I think the format changes have been a little over-exaggerated in some ways in the past, and where I think people have traditionally underestimated how much staying power traditional models have. There were a lot of things that were supposed to kill a traditional text-based Internet, but the only thing that's come close is video, and that seems to get a lot of blowback; I'm not sure it was an improvement.
But regardless, your comment is a insightful perspective that gives me some alternative ways of looking at this. So I think I agree. I might be just more on the skeptical side of things of how well-suited the current tools are for building the kinds of applications you're describing, but I get the theory.
And generative AI that's separate from AI answering questions is kind of another story; I can absolutely imagine potential creative impacts around stuff like music loop generation, art generation, etc... I similarly am looking at the current state of things and saying, "well... the tech doesn't seem to be as good as people say it is, so I don't want to start using it now before it improves", but I can at least imagine how things might change if the tech does get good. I'm not sure it would be a "revolution" or that it's going to put artists out of business or whatever, but it could potentially lower the barrier of entry for certain applications. And certainly LLMs for general website classification I think would be a really good use that I wish was being pursued more.
At the least though, even if I'm skeptical, I can definitely understand why someone would have that perspective, and your summation of the different groups rings true to me.
I still don't think I want it in my current search box today though :) I don't think that LLMs are useless at all, I'm just not sure searching in specific is a good use for them.
Just for clarification, do you mean the actual encyclopedia, or Wikipedia?
I think that encyclopedias are cool, but I also feel like the Internet was hopefully going to be a slightly better version of that and that it's a little frustrating to be going in the opposite direction. I'm not sure how to articulate that other than that encyclopedias are in some ways a compromise around the fact that we very often don't have good search, so we accept human beings trying to pre-aggregate data for us with the hope that they are better than Google is at aggregating that data.
And I think Wikipedia is valuable not so much because of the summaries, and more because it's obsessively curated and has (or attempts to have) a very specific, predictable set of rules that it (tries) to adhere to about sources and coverage. The text-portion of Wikipedia isn't really the part that I think is most impressive about it. If GPT-3 was being used to aggressively curate search results and remove low-quality content, then yeah, that would be potentially interesting (although I'm not sure how well it could handle that task).
----
> what can I make with kale, tomatoes, and mushrooms?
I sort of see it, it's one of the more... it's fine. It wouldn't be a strict downgrade over existing search, maybe it would save time in some situations. But if I'm being completely honest, what I want as an answer to that question isn't a paragraph of text explaining multiple recipes, it's a bullet-pointed list of recipes with links to the original sites they're listed on, so I can check to see if they're worth making.
Bing's results (to its credit) seem like they're sort of headed in that direction, which, nice. Yeah, I could use that. But a bullet-pointed list of recipes is also... search results. So how much time and effort have we put into reinventing a cleaner search interface where at best it solves the same problems that search already solved, and where we in the meanwhile haven't made any progress on the really hard fundamental problem of "how do we get a good list of recipes to display in the first place and what does 'good' mean in that context?"
----
To be less cynical, one way I could see this genuinely being useful would be completely behind the scenes in a non-user-facing role just as a way of applying "tags" to webpages or doing filtering. I would love to be able to search "is spinach poisonous for rats who are pets and just because I used the word 'poisonous' does not mean I want 2 pages of links about exterminators or getting rid of wild rats."
An LLM would be a great fit for that, because we've done such a garbage job of categorizing the web that it's very difficult to know which words to type to exclude "categories" of information from a search query. So maybe an LLM helps standardize that a bit? But then I want a normal page of search results. I don't want to have a conversation with the LLM, I want the LLM's role to be exclusively "I think this webpage should be additionally included/excluded in your query. I think this webpage is about extermination even if it doesn't use that specific word."
If Bing's service ends up being able to do that kind of thing well, then yeah, that's useful. I'm a little skeptical they can, because... gestures to the current quality of search results, but maybe their integration with GPT gives them way more capabilities and ratchets up their quality.
But similar to above, it feels a little bit like reinventing the wheel. I can refine GPT queries during a conversation, great. Could I have that feature for regular search? Why do I need to do it as part of a conversation? That seems like a good thing tied to a bad UX (although again, I might be atypical in thinking that natural language conversations are often bad UX).
And I do want to couch that by saying that maybe Bing will surprise me and it will have easy options to do that kind of thing. I'm just not currently seeing it presented in a way that looks useful.
CPedia was an attempted pivot by doomed search startup Cuil that would more or less build a Wikipedia-like page out of random text on the Web about whatever you searched for. This article describes it a bit. https://www.plagiarismtoday.com/2010/06/09/cpedia-a-spam-blo...
Fully agree, but the problem is not really specific to AI. Most search results today, whether written by humans or machines, tend to be mindless, inaccurate, and unhelpful n-th hand summaries produced with very little effort.
And this doesn't seem unreasonable, given that you get everything for free or at a very low cost. When you pay for high quality books or periodicals, you get in return much better sourced information written by people who know a lot more about the subject they're writing on than the average journalist or AI language model.
Yes, occasionally one might find high quality contents in blogs, forums, wikis, or open-access periodicals, but far more are locked inside proprietary platforms or behind paywalls that do very little to actually compensate the authors.
Search engines and content platforms are supposed to make it easier to find what you want. But the reality is that it's a lose-lose situation for both the writer and the reader. The writer is forced to give up their contents at a very low price and overpay for ads, while the reader is left with low quality contents that aren't relevant to their needs. But neither can escape the monopoly, who alone profits at everyone else's expense.
I do not understand the appeal or gpt-powered searches.
Most of my web searches are for looking up specific things, to find the specific link(s) that contains the information I need. These aren’t searches that are going to be made better or faster by an ML model-they’re not natural language queries, they’re just a bunch of terms.
Depends on your use case - if you are coding, or doing constraint analysis - 75+% of the time it's a single request, single answer, and you are done. The other element with ChatGPT is if you aren't happy with the initial answer for some reason - you can engage in a conversation with it, provide some guidance, and will adjust it to suit specifically what you are interested in. I've found about 50% of my google searches just go straight to ChatGPT these days. Hallucinations are the only real problem I've had - but over time you start to become cynical about the truth of anything factual - asking for unit tests helps quite a bit when coding - double checking any math is also important.
I do agree though, that without citations to the original source - any "Facts" that ChatGPT offers are absolutely untrustworthy.
As someone else said, I think a surprisingly (overwhelmingly?) large amount of queries are just questions people want answered, or close to them, and not really "search queries" in the traditional sense of text people are looking for on a page somewhere.
Even before this, you've seen the search engines add features to cater to that kind of use, things like Siri handing off questions it didn't understand to a search engine (as well as the other assistants that can do that), indications of this behavior in how companies like Google show themselves being used in ads.
Of course, regardless of the true prevalence of that behavior, it's probably in Google et al.'s favor to encourage that. Regular search sort of inherently cedes some power and control to the pages the results are coming from, where you're sending users away to if their answer isn't right in the snippet. But the "answer box" features, or an LLM that just tells you "the answer to your question" directly on the page keeps you there, treating the search page as your source of information and not somewhere else.
When you are curious about a fact, you're looking for a useful and accurate answer. You don't care what page you read it on, as long as it's both useful and accurate. You just think you are looking for a page because that's how you're used to doing it.
Beyond that, most of the results that rank for anything are plainly worse LLM blogspam padded for SEO. All you're doing here is cutting out the middleman.
The problem with chat gpt for factual question search is, of course, that it's not factual. It's job is to produce coherent sentences, not actually tell you factual information. So until they manage to get that right it's not a superior product.
There are two types of search, people looking for a specific resource (like searching for a song on youtube) and people just looking for an answer to a question. If a LLM can be factually accurate, its a superior search product for that specific use, which is probably the majority of search.
That said, ChatGPT does a very good job of understanding what I am looking for. If Microsoft could just insert this understanding into its regular search algorithms it would seem likely to significantly improve the quality of search results.
How could they have left such a massive gap in their product. They literally have the model and resources to revolutionize search. We all know LLMs will hurt their ad revenue, but regardless they have to have known this was coming. This is so similar to FB getting caught off guard by TikTok. There was a gap in the utility of their product (TikTok enabled grass roots content creation), and they just left it wide open.
Its some combination of:
1. ChatGPT is so much better than previous versions that Google themselves was stunned by the utility.
2. Incompetence/Gross negligence across Google
3. No way for them to reconcile the lost ad revenue, so they released nothing. This case is hard to argue for, as they would know theyre a sitting duck.
Regardless I am hoping for a massive Google failure. Theyre the ones responsible for the SEO content waste land that is the modern internet. We have all suffered at the feet of their ad machine
4. Google Search already has lots of useful AI already in it, but Google didn't want to integrate a racist, confabulating chatbot, forgetting that modern users have no preference for truth over lies.
Why are you blaming Google for not being perfect while making the best free search engine, after you spent your whole life refusing to pay for a non-free one?
If OpenAI is willing to release it and Microsoft invested 10B, I have a very hard time believing that censoring the model is impossible. Microsoft 100% did their due diligence on the model.
Google is a monopoly, there is nothing anyone can do. Their search engine and business model has structured the internet and thus society. This thing needs to die
This is the same Microsoft that had to close down their Twitter AI “Tay” after a single day because it immediately became a “racist asshole” (as per The Verge) in 2016?
The same OpenAI ran by Sam Altman, who just last year was part of a crypto biometric scam called “Worldcoin” that attempted to collect biometric data from some of the worlds poorest in exchange for a shitcoin?
I’m sure they’ve done their due diligence and aren’t just pushing out a broken product as quickly as possible after it went viral because they saw dollar signs…
I’m pleased you have the privilege to just ignore the potential negative outcomes of this technology, that as per the marketing hype is set to become the new way the world interacts with information, primarily owned by two unsavoury entities who have history failing to protect the most vulnerable.
2016 was 7 years ago. And I completely forgot about that incident. And so did everyone else.
> I’m sure they’ve done their due diligence and aren’t just pushing out a broken product as quickly as possible after it went viral because they saw dollar signs…
Why wouldn't they? If they bet and win, they significantly disrupt the search market and many others. If they don't, people still don't use Bing. The rest of their business will continue on as is.
> 2016 was 7 years ago. And I completely forgot about that incident. And so did everyone else.
Microsoft also?
> Why wouldn't they? If they bet and win, they significantly disrupt the search market and many others. If they don't, people still don't use Bing. The rest of their business will continue on as is.
The GP was responding to a post that said that Microsoft 100% did their due diligence.... I'm unsure to what you are responding.
Let's be clear, the reason Google didn't want to jump into conversation search is because it invalidates large portions of their business model.
At best they want to avoid bad publicity of their tech just being bad enough to throw out racist remarks. But the company who dropped "don't be evil" from their mission statement, and who fired AI ethics researchers for their research in AI bias that Google did not like is not a morale authority on the matter.
Also, Google does not provide an ad-free paid version of google, you statement is a non-sequitur.
> who fired AI ethics researchers for their research in AI bias that Google did not like
I'm no fan of Google, but this seriously glosses over a rather complex situation. A paper may have been the catalyst, but I'd argue definitely not the reason for the firings. You can't just demand things with the threat of quitting, then act surprised when you're terminated.
I don’t agree with that, it wasn’t even a complex situation, Google just didn’t like the idea that a paper attributed to them could point out all the obvious unethical implications of this technology and shut it down.
The drama afterwards is largely irrelevant and if anything Gebru’s refusal to just meekly accept the company line demonstrates just how valuable she is in this field. The dismissal isn’t the real story, the retraction of the paper is the real problem the OP is referring to when discussing Google’s moral issues.
Bing is my default on my work and home PC as I use edge, I've been too lazy to change it. For the most part, whenever I accidentally search something on Bing, I end up having to open up Google and search it again. Their results are still way better.
I think it’s interesting to consider why touchscreen mobile phones won out over their predecessors.
Touchscreens are the worst interface option. The feedback of feeling a button, or anything made for the job, is better than typing without specific feedback onto a screen. It won out because of you don’t have a set interface it’s bettor.
Now for search, chat gpt will likely always be less reliable than a list of results you can vet yourself for content and source.
That said… I don’t think people care about truth that much these days so one response that feels correct is could be good enough for most. Terrifying times we live in folks.
Hottest take I've seen on here for a while. In a very specific, very limited set of mobile usecases, buttons may offer a more pleasant experience than a touchscreen. I only say that because there's usually exceptions to any statement, but I can't actually think of one.
Touchscreens didn't win because of some lazy sheep-like consumerism, they won because the product is superior. If Chat-GPT style models defeat traditional search engines it will be because the product is better, not because people are content with a response that "seems correct."
Let me rephrase: if you
have a set interface touchscreens suck. No one would choose to use a touchscreen over a physical keyboard, or steering wheel, or mouse for an fps game.
It’s a crap interface compared to others but it’s the best to do anything. That’s why it won.
Trivial counterpoint: you cannot correctly operate a touchscreen without looking at it (unless the display is essentially one single interactive object). Trying to use a music player via a touch screen while running or cycling is basically impossible unless you look the screen. For this purpose, buttons are better.
When defining "better" you need to indicate the metrics you are using. LM's might be "better" when measured with certain metrics, but like most things are worse when measured using other metrics.
To be fair, touchscreens won out because of the Asian, specifically Chinese market. They were too poor to own laptops and wanted big phablets. Hence here we are today.
I still miss my landscape key board phones. The droid 4 still holds a special place in my head.
I think it's more that touchscreens are superior for certain kinds of activities, like consuming information in the form of feeds, videos, and images. Mostly because the screen is larger for the same device size. Buttons are more accurate for typing, and often (I believe, without citation) faster for typing as well, even compared to swipe systems.
One is not superior to the other over all, but each is better at certain things. The difference is that more people will pay money to maximize their user experience scrolling through feeds and watching videos, compared to typing emails.
Agreed, the whole “this is the new search” has perplexed me. Primarily the idea that this AI is supposed to be the single point of truth. A traditional search allows the user to review a variety of contexts and viewpoints.
It’s highly concerning that this is something people seem to want (or at least, is being pushed onto them through marketing & hype), despite the huge outcry over algorithms and political agendas over the past few years…
Fairly often I ask the Siri homepod thing a question. It's good at playing music, doing simple math, translating some things. It can sometimes read me a wikipedia page. Far to often it says "I have some web results for that question, ask again from your ipad". In those cases I wish she could summarize and read me the answer.
I can understand why some times you might just want a chat response to a question rather than a web search.
But I swear that if I even suspect Siri's response is influenced by advertising or endorses a product, I will take the homepod and punt it down the street.
Have the feeling that we are in a state of overreacting (news outlets love the headlines).
In addition, my gut feeling says that Google does things better than Microsoft.
This comes from my experience with working with Google and GCP products vs working with Microsoft Office/Windows and Azure.
With Google, everything feels consistent, well documented and well presented: from Meet/Gmail to BigQuery, Tensorflow, Cloud SQL, Firebase, etc.
With Microsoft I frequently observe inconsistencies in design patterns, broken links and outdated documentation. Just compare for example Office products, the concept of Office 365, Live Account vs Outlook account, Windows, Bing, Azure products, etc. Basically its hard to find any commonality and it feels that there is generally a lack of vision for the entire suit.
I therefore extrapolate that Google will answer this hype in a more structured way, anchored to a vision, putting it against a grand scheme of things and deliver it as a seemingly integrated solution to the suite.
I agree in general that Microsoft leaves a lot to be desired.
But Google's product track record has been abysmal lately. Based on their horrible live event today, it seems like things are not so grand behind the scenes there anymore.
Judging from the decreasing quality of Google Search I'd say they've been doing it to themselves over the past decade regardless of large language models or not. However, that being said, Google is still super dominant and when people say "SEO" they generally mean Google search results. And until my mom stops using Google as the default search, I wouldn't count them out of being the dominant search engine for some time.
While Google certainly has its faults and I am frustrated with its trajectory, I would bet money that this is an article pushing the nonsensical notion that AI language models will replace search engines.
Edit: based on the subtitle, it seems like I am correct.
How is that nonsensical? Seems quite likely to me, especially if they can ground them better in a factual knowledge base. Most search engine queries are looking for information.
The problem is that this literally does not exist today. ChatGPT is a chatbot, it has no factual knowledge base. It's optimized for dialogue.
That's why it's absolutely nonsensical to see all these articles, because they are talking about a chatbot as if it's trivial to turn it into a search machine. That's an entirely different problem.
ChatGPT is pretty much the definition of a troll. It's only mission in life is to keep you talking to it. To do that it has no problem with lying, confusing, and various other deceitful methods you wouldn't accept from a human.
Most of the information people are looking for isn't really the sort of thing an LLM is set up to cover though. If I want information on anything since 2021, or to look into stuff near me, or access a price, a download, or information in visual form, or anything that's too niche for its corpus, ChatGPT is not good. When ChatGPT can offer an answer it's fast, but about as reliable as a search engine's spammiest result
And answers as text wrapped around a factual knowledge base just sounds like the sort of queries Google Home already handles in natural language form. That can certainly be improved upon, but it isn't obvious that the LLM engineers will beat the search engineers to it...
I am really perplexed by people talking about the end of traditional web search.
I guess HN users tend to look up lots of facts and whatnot. But generative AI is not a better UX for queries like "LeBron James instagram" or "wordle" or "Avatar 3 release date" or "WSJ" or "Spanish to English" or "cricket news".
I'm not saying that it's not going to settle in to being frequently used for some domains, but I think we're at the "irrational exuberance" point of the curve about what this tech is going to do at the moment. Playing around with it reveals limits pretty quickly.
Do you people really want AI generated answers? As if the internet wasn't bland enough as it is. This forum has been soapboxing for years about the longtail of blogs and BBs and niche sites, right up to the moment that Bing got chatgpt.
So you'd rather have a load of probabilistic bullshit? The training sets are way too huge to be human audited, just wait until the malicious actors start producing propaganda with their own GPT sites and deepfakes. The fucking thing can't even tell a straight fact about a telescope. If you thought searching the internet was fucked before GPT, give it a few years.
I'm not sure who'd want probabilistic bullshit. I said people wanted answers and Google was crap at giving it to them.
At the moment ChatGPT is like having an personal assistant that used to be an world leading expert, but has been sniffing glue for the last 5 years. He knows a lot of shit but you can't trust it. But he can help you get started and he's fun to be around.
Except ChatGPT isn't fun to be around. It's been castrated of humor and dominated by extreme political correctness. I think GPT is the catalyst for a change, but in Microsoft's hands it'll just be a glorified Clippy.
As for training data scraped from the internet, it will be split like steel was with "pre-war steel", due to atomic bombs rendering steel unfit to use in geiger counters. We'll have pre-gpt data.
I just want Google to fall, nothing much more than that. I personally am going to continue using Kagi, but will quietly cheer Microsoft eating Google's lunch.
Late 30s. Yes I realize that me 20 years ago would find a lot of distaste in my current position. However, from my point of view, Google is causing a lot more present day harm than Microsoft is. Microsoft has been reduced to the point of having to ship Google's browser, and offering products with Linux support. Google is in no such position, and continues to use anti competitive tactics to vertically integrate, to create a landscape that is favorable to their ad machine. I realize that this is an unpopular opinion, but there ya go.
If/when Bing gains dominance and starts abusing its position, I'll probably change positions, but that's the way it goes when there's a lack of competition in the market. You're left cheering for the best of the worst.
Let me guess what this article and thread will have:
> BIG TECH AI CHATBOT GOLDRUSH!!!!!!
> "Google sucks now, I use Kagi/DDG/Searx instead"
> "I have to quote search to get the results I want. Google has seriously gone downhill"
Can we cut the shit for two seconds? It was never all that great to begin with. None of them were.
Also, these news outlets are just fucking lazy and need to go outside and do actual journalism instead of regurgitating the same trash over and over. ChatGPT is less redundant than they are
But a lot of people fed up with Google and will be glad to try something different.
It's like when Apple came and eat 15-20% of market of notebooks.
Dominance of Microsoft still not ended.
Actually I suspect that Yandex search is better - I like how they preview answers for technical questions. Picture search - seems that Google is the worst.
Maybe it's time to start search different topics in different engines.
Hm. Actually AI could help to implement such functionality
> a lot of people fed up with Google and will be glad to try something different.
I had to use DuckDuckGo to locate some mailing list threads (about an apache jira issue) which were clearly on Google at least a few years ago.
So "hdfs forward seek" in Google returns nothing, but DDG does.
If Google isn't repurposing all its search compute to something else secret, I can't explain why it suddenly got all dumb and forgetful of the 2015s internet.
Its very strange. There are many cases I noticed where I double quote search a string and it yields no results but once I do "my query" site:thesiteiknowitsfrom.com it gives the results.
> But a lot of people fed up with Google and will be glad to try something different.
Why would they replace Google with another bloated adware search engine then?
I used to be one of these people, not going to lie, but Bing is worse than Google when it comes to UX. The results are not as high quality, bloated and really messy UI.
I switched to DDG and then Brave Search later on and I can honestly say that Brave Search gets a lot of things right. Mainly the results quality and the clean UI and widgets.
Yandex Search is decent, but over-indexed on Russian content. They’ve also suffered major brain drain since the war began so I don’t expect them to keep up with the AI arms race.
So i tried Bing again after almost a decade for their ChatGTP integration. I can't be the only one who thought this is Yahoo/AltaVista/Lycos/etc. revisited. It's a fucking portal spamming me with news coverage, not a search engine
Microsoft's insistence on reusing the bing brand and website is baffling to me. No one wants to use that shit, but in classical Microsoft fashion, they're too attached to old decisions. What a missed opportunity to do like they did with IE and kill it in favor or something new and fresh.
Ultimately, Microsoft will make it easier for Google to catch up because its internal culture is and will continue to be unchanged, no matter how much technology they inject from OpenAI. Make no mistake, behind the shiny PR, it's still the messy Bing AI organization, the demoralized Microsoft Research organization and good ol' Microsoft.
The concern is not whether Google or Microsoft is first, the attention should go to web publishers. If ChatGPT as a search engine works well, it would mean that the user gets a direct answer without ever hitting the website.
Which begs the question...why produce any content at all? It will be gobbled up by AI and you'll never hear back. Thanks, I guess.
I can think of one issue delaying this inevitability: consumer laziness. Most people type one or two words, usually misspelled. That makes for a pretty poor AI prompt.
> Once trained on this giant corpus of natural language, the chatbot can, when prompted by users’ instructions, produce a fluently written answer rather than merely serving up a list of links.
A word salad of untrustworthy bullshit instead of the pages you were looking for is not going to upend search.
It's not a replacement for search in any way.
Now if the search engine could find pages which satisfy complex queries in natural language, referencing content and relationships among content, that could be something.
But google (nor bing) return the pages you're looking for even close to as reliably now as they did before. If Microsoft can somehow work out the issues with factual accuracy in the gpt interface (and salvage the bing brand) they're going to have a winning product on their hands.
Users sometimes want a factual answer. Usually they want to go to a page. Pages are not just sources of information. People interact with pages to do real stuff connected with off-line activities. E.g. chatty paragraph won't connect me with the government/medical/banking/restaurant/shopping/whatever page I'm too lazy to search for through its own URL and portal.
At best, you can have this: if it looks like the user is typing a question into the search box, an AI-generated side bar can complement the answers.
The web search side of it cannot go away, otherwise you just have a sophisticated Eliza.
If I'm going to see a factual answer, I want external references for those facts before I trust it. I'm likely not a subject matter expert, if I'm asking about something; I can't tell when the AI is bullshitting.
Now if the AI-powered search engine had an intelligent index of all the texts of all the crawled web pages such it could do some basic reasoning about it to serve up those pages based on sophisticated questions (and not simply serve up original prose based on those texts), that could be a game changer.
Google is already a zombie company, and has been for some time, still insanely rich and powerful, but they're not cool anymore and I think they could be replaced very easily.
Large corporations can seemingly stay alive and kicking for a long time after their death. Nokia, Kodak, Palm, RIM just to name a few.
I've witnessed the demise of Palm, then Nokia and RIM, following similar patterns, when they start reacting it is already over.
I asked ChatGPT that if it can replace Google search engine. It says :
"As an AI language model, I do not have opinions or personal thoughts. However, it is unlikely that chatbots like ChatGPT will replace traditional search engines like Google in the near future. While chatbots like me can provide quick answers to specific questions and engage in conversational interactions, they still have limitations and are not as comprehensive or efficient as search engines like Google. Google and other search engines have access to vast amounts of information and use complex algorithms to deliver relevant and accurate results.
That being said, chatbots like me can complement traditional search engines by providing an alternative, more conversational interface for accessing information. Additionally, advances in natural language processing and AI technology may lead to chatbots becoming more sophisticated and capable of handling more complex tasks in the future."
I wonder if MS can query and see the pain of people just searching for the Google Chrome download link, downloading it, closing Edge and never opening it ever again.
The quality of Google search has decreased so much over the last few years that I had to resort to using Yandex, a company that has a budget that is orders of magnitude smaller that the Big G and is under the control of a dictatorship.
Google search has two huge problems: SEO and censorship. Search for anything related to products/torrents/streams/politics on Google and your results will SUCK, due to one of the two reasons stated above.
The recent Yandex hack/leak has the cynic in me connecting the dots and, seeing how Google search seems to be facing REAL threats to it's dominance since it's creation... maybe some guys with the deepest pockets in the world are starting to enter WAR mode.
Destroying Yandex advantages on the SEO battlefield by way of divulging their parameters to the world would be the Franz Ferdinand assassination moment of the Great Search War.
I hope so or unless Google changes "back" to when it actually did what I told it to. This might sound weird but one of my skills is that I know (or knew) how to use Google. I would not just type in whatever I needed to find, I would know to exclude specific terms using "-term", put parts of a sentence in quotation marks, add other terms I knew would be on the site where the answer could be found and whatnot. The search is then very specific and if I didn't get a satisfying answer it was because I needed to improve my query. Now whenever I do that, Google takes this query and modifies it without me knowing what they're actually doing. It seems harder and harder to tell Google to just work like it used to.
I said it in another post but AI is a feature, not a product. You just add it on. It doesn't become "the thing." People that are saying this are the same people that said Stories would kill Instagram or Facebook. It's just a feature you add in.
Probably not because of Bing's most recent moves, but inevitably, yes -- depending on your definition of "about to" and your definition of "end".
Search in general, and specifically Google search, depends on Adwords for revenue similar to how newspapers used to depend on classified ads. Once ChatGPT.next or something similar removes the opportunity to display ads that people will click on, search will face a similar crisis of revenue to newspapers. Maybe Google will figure out the economics of the new landscape, but it certainly won't hold the same position of prominence in the internet at large as it does now.
Similar statements have been made about most dominant technology when it was in its infancy. You are not wrong, but it can quickly go from novelty to necessity once folks start to rely on it.
Do you see yourself relying on it? Not when you know it’s KNOWN to give wrong answers.
It’s also worth mentioning that Google PageRank was not the first search engine. No, it was good 4-5 years after the first usable search engine came to market. I think the same will be true for AI chat.
How can you say that with a straight face? It’s a regurgitation of things it’s collected from the WWW. It is a step up from Siri/Alexa but completely lacks actual conversational skills.
If you don't put in an honest effort to give it a shot, you will remain forever confused why people find tech like this useful for some use cases and you don't.
People said Galileo was a crank, supposedly. But they also said that about the crank raving at the bus station. That people have been wrong about how transformative or not a technology would be in the past really tells us nothing.
If anything kills Google search it's the ads. Search for a product and the first page can entirely be ads or sales links. You have to specifically search to find reviews (many of which are likely paid reviews), and search even harder to find actual real user reports. Typically I add "site:reddit.com" to the end of any search where I want to read people's opinions of a product.
Yes, finally we are going to see a shift, but their 'dominance' is not going to necessarily change that much in some kind of 'step function'.
I suggest that Siri is about to get a lot better, and act more like we wanted it to act a decade ago.
As amazing as ChatGPT is, it's actually not that useful for most regular people other than as a more human form of search.
I think Google results will improve quickly enough that whatever MS does will seem novel but not necessarily have some kind of huge consumer impact.
We'll see search for company info and help directories get a lot better.
Remember that both MS and Google are 'etablished' companies, the later a bit slow moving and not firing on all fronts.
Why would anyone think a company that can't get Skype to work, is going to marvel the world with some 'new thing'.
OpenAI folks are on a roll, rapidly trying to make cool looking experiments, focused on 'wowing the world' not on direct revenues. Making that translate into products is another dimension altogether.
The most interesting stuff will come from startups.
As an aside I guess, I have noticed recently how much worse the search actually is. I'm not even referring to the adds, simple things around library documentation for example, where I would always find the best results up top to close to, now I get quite unrelated results, not even containing the phrase/function name I'm searching.
This is not because the info is not indexed or is behind a walled garden though, I now have to find the source website and go directly to their docuemtation myself, or their github. It's not an 'issue' but defeats the purpose of google search itself.
In a strange change for me particularly, for news/general info I actually will search through sites like reddit or twitter to get some actual targeted results.
My fear is that more AI will hinder our ability to literally search for stuff. The future I dread includes more profiling, a stronger search bubble, all of which will just make it the death of the objective search engine. As fascinating as AI can be, I don't think it's what search engines are missing. That is certainly not what I need.
Even today, engines rank your results by geoip, trying to be smarter than you ask them to be. Two individuals doing the same quoted search will get different results. Perhaps I'm just getting old, but I do miss when searching was a matter of writing a good query, not a matter of what the engine thinks you will want to ask next and what else might interest you.
Going rate is about 3 cents per LLM query, so at Google scale of 100k qps, it would be ~$300MM/day. However you probably need to send only about 1/10 of current queries to LLM (as most are still not suitrable for LLM to asnwer) so that would cost only $30MM/day or about $9B/year which is peanuts for Google.
From what I have read, some people are of the view that it will affect their margins and ability to compete in other businesses. This is based on something Satya Nadella mentioned in an interview
I remember Peter Thiel once said that investing in google is investing against innovation, that the company had no competition for a very long time, and that they see no reason to why they needed to reinvent themselves. As long as there's no threat to the company, they can just maintain the current tech and harvest the green from it.
https://www.youtube.com/watch?v=2Q26XIKtwXQ
“Search” relevance seems as secure as they’ll let it be (ie: it’s their own fault if lose that due to self-inflicted distraction).
Their challenge in that particular area is the relevance of “search” itself.
Most of that decline can be attributed to them indirectly via the behaviors they incentivized (and LLMs will now nail that coffin shut), with the social media siloes playing their part.
But they shouldn’t lose an edge here unless they give it away. Never say never.
The problem for Google is that they've already taken a ChatGPT approach of providing inaccurate information. The feature snippets are basically proto-ChatGPT, and as the internet has gained content, most of the top links are also inaccurate. You basically have to append a specific Website to your search to get anything decent.
If Google were ever to remove the 'verbatim' option (which is lacking in other search engines like Bing as far as I can tell) it would become much less useful.
Unfortunately, one cannot enable both 'time' and 'verbatim' option under Tools in Google Search. I guess I should go ask ChatGPT why that is the case...
It's way too premature to say that ChatGPT will overtake Google, not to mention Google still derives about half of its income from Youtube and 3rd party publisher ads. Every few years the media runs with some story about how Bing, Yahoo or whatever will overtake Google. Never even comes close.
> For generative ai to make real money, it will need to find uses in “navigational” searches (looking for a site’s internet address by its name) and especially “commercial” ones (“Best new ski boots this season”).
Serve users what they want, deals to hunt for: Apple Maps. Leverage Foursquare and Yelp lessons.
All good points. Windows still very much dominates desktops / laptops sold (especially large corp environments)...so the default installed Edge browser + Bing is a significant share of users.
And with Edge being "basically chrome", and supporting Chrome plugins, the switching cost from Chrome -> Edge is quite small.
But for mobile, Android + iOS dominate...so that all goes to Google.
I think this could mean that Bing + ChatGPT has a really opportunity to beat Google in Search...it's gonna be hard, but I've never even considered them more than a distant 2nd option until now.
you should also consider that Google is/was the best search engine around so it makes sense as a default, sure they pay to be the default and most people never change defaults. But if you are making a phone you might just really actually care about people having a good experience with your phone.
(and its not like Microsoft couldn't pay them either anyway)
4-5 years ago I would have agreed. But I've been using DDG for my primary search and only "resort" to using Google maybe 10-15% of the time my search fails on DDG.
I'm not saying DDG is amazing, only that you don't need a "Mercedes" to pick up milk.
Betteridge's law right here. I don't see how ChatGPT is going to do anything in the search space and it's stunning to see people tripping over themselves to declare Microsoft the winner already. I'm already predicting they will write down this acquisition for a few billion in 2 or 3 years.
AI question and answer searches have existed for a while. Alexa does an equivalent job as ChatGPT and I find it to rarely be useful. Google's stroke of genius was to skip curation and algorithmic ranking in favor of a heuristic. Namely, PageRank. I know it's gone through years of refinement, tuning and tinkering, but the fundamental basis of popularity is just a way more reliable and durable approach than trying to intuit what people want based on what's on the page. You'd absolute be back to dark days of blackhat SEOs bombing results by inundating GPT with keywords. It won't work. The popularity model works with and only with a network effect of having zillions of successful searches to mine. Maybe Bing has just enough data to be useful and maybe ChatGPT can do a marginally better job of giving boxed answers to specific questions, but that's not the thing that makes search valuable.
I don't want an AI trained on the sea of garbage that is the modern web to answer my questions, but I would love an AI to draw on a curated, hand crafted, trusted database of the worlds knowledge.
This is like Amazon getting leapfrogged by an upstart shipping subscription service. The cornerstone of Google is search, and everything else depends on that.
I'm old enough that I remember switching between Yahoo, AskJeeves, AltaVista, and others, before abandoning them all for Google. I also remember experimenting with a friend to test the search engines, and I remember the question/query I used to perform the test: I had (at the time) a vague recollection that the level of the Atlantic and Pacific oceans near the Panama Canal is different; that in addition to raising ships over the isthmus, the canal has to raise them overall in one direction, and lower them in the other. That seems bizarre, since the oceans are connected south of South America, but that was my memory, so I asked the search engines. Google was the one engine that returned sites that confirmed my vague memory.
So of course, I just tried this with ChatGPT. It glitched mid-response, but here's the entire exchange:
Me: Is the level of the ocean the same on both sides of the Panama Canal?
ChatGPT: No, the ocean level is not the same on both sides of the Panama Canal. The Panama Canal is a man-made waterway that connects the Atlantic Ocean to the Pacific Ocean, and its locks raise and lower ships to accommodate differences in sea level between the two oceans.
-- Load failed -- I hit retry
The Atlantic Ocean side of the canal is higher than the Pacific Ocean side, with a difference in elevation of approximately 26 meters (85 feet). This difference in elevation is compensated for by a system of locks, which raise and lower ships to the appropriate level for transit through the canal. The locks fill with water from the higher ocean and drain into the lower ocean, allowing ships to pass from one ocean to the other without having to navigate significant differences in sea level.
Me: Why is the level of the ocean different?
ChatGPT: The difference in sea level between the Atlantic and Pacific sides of the Panama Canal is due to various geological and oceanographic factors.
One of the main reasons is the shape of the landmass and ocean basins. The Atlantic Ocean is relatively shallow and has a broader continental shelf, while the Pacific Ocean is deeper and has a narrower shelf. This shape influences the ocean currents and tides, and helps to maintain a difference in sea level between the two oceans.
Another factor is the prevailing winds and currents in the region. The trade winds in the Pacific and the prevailing westerlies in the Atlantic both play a role in shaping the sea levels on their respective sides of the canal.
Additionally, the location of the Panama Canal at the narrowest point of the Central American isthmus has a significant impact on the difference in sea level between the two oceans. The geography of the region funnels ocean currents and water into the narrow channel, leading to differences in water pressure and sea level on either side.
In summary, the difference in sea level between the Atlantic and Pacific sides of the Panama Canal is due to a combination of geological and oceanographic factors, including the shape of the landmass and ocean basins, prevailing winds and currents, and the geography of the region.
-- comment -- Although ChatGPT is correct that the levels differ, it is wrong about the amount: it's nowhere near 85 feet. It gives believable reasons for the difference, but doesn't mention tides, which (as I understand it) are a significant factor. It also doesn't mention salinity, which (again, as I understand it) is also a factor. I gave it a chance to catch the error:
Me: Is the difference really 85 feet?
ChatGPT: The difference in elevation between the Atlantic and Pacific sides of the Panama Canal is approximately 26 meters, or approximately 85 feet. However, it's worth noting that this figure can vary slightly due to tides, seasonal changes, and other factors. The exact difference in elevation can also depend on the specific location within the canal, as some areas may be slightly higher or lower than others. Nevertheless, the difference in elevation between the two sides is typically around 85 feet.
-- Comment -- No luck correcting that 85 feet number. I also asked about tides and salinity:
Me: Do tides or salinity contribute to the difference?
ChatGPT: Tides and salinity can contribute to the difference in sea level between the Atlantic and Pacific sides of the Panama Canal, but the main factor is the shape of the landmass and ocean basins, the prevailing winds and currents, and the geography of the region.
Tides do have an impact on the sea level in the canal, as they can cause fluctuations in the water level on both sides. However, the impact of tides on the difference in sea level between the two oceans is relatively small compared to the other factors mentioned above.
Salinity can also influence the sea level, as differences in salinity can cause water to expand or contract, leading to changes in sea level. However, the impact of salinity on the difference in sea level between the Atlantic and Pacific sides of the Panama Canal is also relatively small, as differences in salinity in the canal are primarily caused by fresh water inputs from rivers and rainfall, rather than differences in ocean salinity.
In summary, while tides and salinity can contribute to the difference in sea level in the Panama Canal, the main factor is the shape of the landmass and ocean basins, the prevailing winds and currents, and the geography of the region.
-- Comment -- Much has been made of ChatGPT's (sometimes unfounded) confidence. I don't know enough about the ocean-level subject to judge here. Clearly ChatGPT is confident, but is it correct? I don't know. Given that a quick Google search wasn't able to convince me that ChatGPT is wrong (apart from the 85 feet thing) I'm calling this even on the facts. The experience asking ChatGPT was clearly easier than Google.
> However, the impact of salinity on the difference in sea level between the Atlantic and Pacific sides of the Panama Canal is also relatively small, as differences in salinity in the canal are primarily caused by fresh water inputs from rivers and rainfall, rather than differences in ocean salinity.
Surely "differences in salinity in the canal" are not what's relevant to the sea level on each side and how much impact ocean salinity has on that.
A fine example, I suggest, of how it does not understand things; it just spews out plausible strings of words.
Honestly, the number of times ChatGPT inserts this and variations of fluff is absolutely bordering on insane. Like, that was your question - why include it in every single sentence of the answer…
The gaming of SEO has been quietly destroying the utility of search writ large as a mechanism for quite a while.
Over the last year, its become palpable.
Google has such utility in this regard that in some cases, a hallucinating lie-machine offers better answer than an index of what information is available on the internet.
This issue with with Googles failure to respond to the deluge of SEO driven content in their searches. They can do better. They've chosen to not do so.
I see this sort of comment a lot, and I honestly don't know what it's talking about. What are people expecting out of Google that it's not delivering?
It can only index stuff that's on the Web. Stuff on the Web is, contrary to what is popularly asserted, only a tiny fraction of all human knowledge.
I think people are forgetting how bad search was before Google. Google drove Web directories to extinction. Remember Yahoo!? Back in that era, if I were looking for something as simple as the University of Michigan, I clicked and drilled down through a Yahoo directory. The obvious search query would have been useless. Google changed all that.
I view Google as the yellow pages. It works well for that. Is it an oracle of knowledge? Of course not. How could I possibly expect to find knowledge on a place where there is no reward for making it available? People producing knowledge don't work for free.
I've tried ChatGPT and it's no better. It serves up stuff that is flat-out wrong.
Things that i *know* were parts of the internet, that I could find via google 5-10 years ago are no longer possible to find. I get single page search results sometimes, with not even the capability to keep browsing 'more' pages to see things that may be tangentially related. Google has just decided to not show them at all. Remember when the bottom of the page was "Goooooooooooooooogle" and you could click each "o" to go to that page? Haven't seen that in a while, right?
> Remember when the bottom of the page was "Goooooooooooooooogle" and you could click each "o" to go to that page? Haven't seen that in a while, right?
That’s still a thing, although it seems they’re A/B testing its removal. I just opened a private tab (as I always do) and got a boring "More results" button, but I tried another browser (also with a private tab) and got the classic pagination.
Absolutely this. Google has gotten notably worse in just the last two years.
Today:
* Any term that might be related to a commercial product? That product comes first and frequently only.
* Search for two terms? It will first give it's prefer result for each separately - usually commercial products (ha). And then might give them together.
* Quoted terms are often taken as vague suggestions. Negative sign is often useless, etc.
Google has gotten notably worse in the last few months. I keep getting the feeling it is serving results for a query similar to what I entered, but more popular. It drowns out any results that actually might answer my questions.
One of the worst innovations they've introduced is returning results for things you did not search for. At this point, potentially anything is a valid search result for any search query.
god, the quoted text suggestion thing is really the nail in the coffin. Used to feel like a wizard showing people how to get exact search results, now... well, that's why we're having this discussion
What makes you think these origins still exist? Google is not going to direct you to a page that can't be reached, even if we assume that your memory is accurate and the page existed 10 years ago.
The point of google was to RANK content on the web, and surface the BEST content to the top. In that regard it has failed, and the expectation is for them to correct the problem and return to delivering good content first.
It's not that the content doesn't exist or isn't indexed, its that its been drowned out by noise. Sifting through noise better was the entire reason google took off from more standard crawlers. It now returns results worse than crawlers from the previous era.
Spam masquerading as the answer to my question is WORSE than off topic returns, because it is a trick vs being something I can easily disregard.
Before the results would just not match what I was looking for. Now they do match what I was looking for, except some AI procedurally generated the content to show up when I searched those terms, with no regard for the accuracy of what the page says.
If I say "show me the best winter gloves, and only from sites that you can verify actually product tested" and it follows the instruction (ignoring sites that just have a list of popular search results aggregated) then it is better. If it doesn't do what I want, I expect to be able to follow up and teach it.
I expect the chat style stateful search to take instruction for what type of sites I want results from. "Return me a list of websites with recipes for Bolognese that do not have a long story above the recipe. Build a table with the top five results normalized for portion size, comparing and contrasting the ingredients. Highlight unique ingredients in bold."
Not really everyone's familiar with the example of cooking recipes being breakfast with the author's life story in order to improve their position in Google rankings, but it's far more prevalent than just recipes. Today I tried to look up something about a feature in a particular piece of reasonably popular desktop software. The top hit that wasn't an advert was several years out of date. The second hit was clearly artificially inflated to long form journalism to try and get me to read a thousand words where 10 would do. Duck duck go linked me to some actual recent results, just like I would have expected with the crawlers of the previous era
In the mid 2000s, I could find my answer easily in the top 5 hits.
In the present day, I cannot find my answer on the first page. If I click on the top hits the page is a deluge of useless blogg fluff which takes me more time to find what I am looking for.
More often than not have to add reddit, forum, stackoverflow, etc to find what I am looking for because online communities provide more concise answers.
The SEO fiasco means that the index to the Yellow Pages is all wrong. You want to find the Chinese embassy, instead you find Vietnamese takeaway businesses.
In the last 2-3 months search quality for me has absolutely crashed and is barely usable.
I think that google search's quality decay is a real thing, it takes some real effort to get proper results that aren't SEO'd garbage.
For example, just the other day I was searching for one string that I knew was part of a common code repository. To my surprise google couldn't find anything at all. Yandex on the other hand found the repository immediately and linked to github.
Other common issue with google is the difficulty of finding stuff like forum posts related to the search query. Sure, you could append "reddit" to the query, but there are still plenty of traditional forum sites and some of them have decades worth of discussion. I Never see those sites pop up on a typical google search unless I specifically look for them. Again, with yandex, my experience is much better, it is not uncommon to see posts from forums to be on the first page of results.
The personal backstory and other "fluff" is actually really useful when selecting a recipe. I was laid up with covid/rebound while traveling over the holidays and lived off Hainanese Chicken Rice delivery. Now I want to make it at home and search for a recipe.
#1 result is a long article with culinary history, detailed instructions, many pictures, and a credited author originally from Shanghai.
#2 result is a simple recipe listing from Buzzfeed. Written by a young white guy from Minnesota who worked as a producer. No fluff, no pictures, no backstory. Doubtful the author ever made the recipe at all. You could grab a recipe database and generate thousands of these pages.
I've been burned by #2 too many times disregard the fluff. It shows their investment in the content.
I feel your pain and switched to high quality, physical cookbooks instead. Recipes on the web are generally garbage unless they come from an author that you’ve already vetted.
Follow up:
Make a profile out of this.
Ah, people who don't want to ever see content from quora or something, do like this other content.
Let's rank that higher for those people.
Easy opt out, by checking "ignore search profile" checkbox or logout.
Often when looking for an answer I just need to confirm what I already know, or look up something I forgot. Google offers me a bunch of sponsored links first and then sites with the best SEO, which will smother me with all kinds of unrelated crap that I do not want and will never need. I'm even running an extension that removes a ton of blacklisted sites from the search results.
ChatGPT usually gives me the answer that I'm looking for and nothing else. Sometimes it does add extra info, which often teaches me about something that I wasn't aware of at all.
But the greatest benefit is I can ask it to clarify anything I don't understand. I don't need to go on a completely new Google quest, or jump through hoops to register on some site and hope a random internet person will ordain to help me out. I can just ask, in the same conversation, and immediately get clarification.
Many people underestimate the incredible learning opportunities a well trained language model provides. It doesn't matter that it hallucinates or lies. Whatever it claims is usually easy to validate. What matters is the speed with which you can find uncluttered new leads or answers.
> What are people expecting out of Google that it's not delivering?
Not optimize for "most documents indexed" but "highest quality of results". One of them encourages adding spam to their index, the other encourages removing spam from their index.
There are giant teams of people at Google whose only job is trying to define things like "spam" and "seo abuse" more precisely in more diverse contexts. There are equally large armies of people outside of it trying to outwit the first group and find more and more elaborate loopholes and workarounds.
HN is constantly pushing this notion that "spam" is some well-defined, solvable problem, so obviously Google wants it. That narrative just doesn't make sense from any angle. The notion that more click bait improves Google's bottom line is absurd
This discussion is pretty sad and frustrating for me. I think your conclusion (“Google has no motive to encourage spam content that rises to the top using SEO tricks”) is about as convincing as it gets, but it’s completely drowned out by a very vague sense that google is a big company and therefor it’s making search worse on purpose. No amount of well-worded HN posts are gonna sway people, imo…
Luckily HN posters don’t exactly represent a meaningful portion of the population.
Have you looked at Youtube comments recently? It's a mess. There's no organization, half of the replies are censored but are still being counted, so you have plenty of "1 Reply" links but nothing shows up when you toggle. Half of the top level comments are completely useless comments that are only posted to help the creator "beat the algorithm", every successful creator now has a call to action to write a comment.
I see two options: a) Google can't do any better than that, b) Google has a reason to keep it in the current state (I'll put "Google doesn't know because nobody at Google has used Youtube in the last 5 years" and similar options under "a").
a) sounds ridiculous, b) sounds conspiratorial. What are the other options?
And again, I'm not saying they are making search worse on purpose (no "from now on our core mission is to make search suck"). I'm saying they aren't optimizing for SERP quality. They seem to care about index size (maybe it's an internal KPI? would certainly explain their aggressive guessing at additional URLs that you might have on their page but don't link to, don't add in sitemaps etc, and their stubbornness in keeping results from the index even if they've been 301ed or 410ed ages ago (they do get downranked after a while though)), but I assume that they mostly care about paid ad clicks, and if something increases ad clicks while the result quality decreases, they'll do it.
It could very well be that none at Google understands the search engine code anymore due to their high attrition. So it has been surface patched for X years making the problem worse.
Their giant teams are an organizational failure then.
Product reviews alone, whether it is enterprise software or sports clothing should be something that they can easily comb through by hand, as humans, and uprank sites that are putting out more than affiliate link assemblies.
Why wouldn’t they do this, if it’s scalable and profitable? Is your theory just “incompetence”? I can’t prove that’s wrong, but I feel the simpler answer is more convincing; that handing off page rank to an army of minimum wage call center workers (for every country & language in the world…) wouldn’t be more effective than the existing algorithms at filtering out spam.
Some form of paralysis. They dont have a single leader who has taken responsibility for it and corrected it. CEO/SVP has too many pots on the stove.
Im legitimately asking, who is responsible for Search at Google? Prabhakar Raghavan is SVP, Search, Assistant & Ads, and I click under him, he has 8 product groups reporting to him, and none of the people are responsible for Search. Yossi Matias is responsible for Search Engineering.
It may at first come off as a laughable answer, but Google Search has been in a directionless spiral since Marissa Mayer left. Her Yahoo tenure was not well received, but at Google she cared about the end quality of the product. Her title was Search Products and User Experience. Notice how we have gone from User Experience to Search Engineering, forgetting about the people who actually use the product.
Absolutely: the challenge is that any signal that you use to identify "good websites" from "bad websites" will be adversarially optimized by incredibly motivated people.
You are dealing with a moving target that has a huge financial incentive. It's a very difficult problem.
I don't agree. You start whitelisting good content manually. If babygearlab is the best result for baby gear, you start hardcoding it. If seriouseats is the best result for recipes, you hard code it. If someone better comes along, they get moved up the priority list.
You figure out a way to crowdsource certain decisions and establish who you can trust. Ask them questions with right and wrong answers. You start to tackle it one product category at a time. Instead of pagerank, which was a web of who linked to who" you start figuring which voters you have who consistently turn in good feedback.
This is some form of metamoderation that slashdot tried to implement.
If you are going to be a tastemaker, stop hiding behind "the algorithm" having some mind of its own that cant be controlled.
What's the reason for these giant teams of presumable well-paid experts and geniuses to not define "a clone of SO with some of the answers juggled around between questions" as spam?
> The notion that more click bait improves Google's bottom line is absurd
If you don't find what you're looking for on the first try, you'll need to try again, and see more ads. What else are you going to do, go elsewhere, visit a library, ask the town elders or give up on looking for things you want to know? You don't have a choice, you know it, they know it.
I find it equally plausible that Youtube's search sucks badly because they don't care what you're looking for, they want you to watch videos that they predict will lead to the maximum time spent on the site, again so you watch more ads. What other explanation is there that the world's leading search engine has the search of one of their flagship products run at 1999 quality? Presumably they have giant teams of people working on that too?
I want you to start a blank slate C (or C++) project. Ask Google how to write heapify, push_heap, and pop_heap in C. Ask ChatGPT the same.
I did this a few weeks ago. I literally could not find the answer on Google. ChatGPT gave me actual C code that I definitely did not trust but did verify.
Google results for questions like that are genuinely awful. It’s full of shitty tutorial websites that are full of ads and either don’t have the answer I need or don’t have it in a convenient form.
> What are people expecting out of Google that it's not delivering?
I mean what you just listed.
Google won the search war because of PageRank eliminating lots of spam, and then something like 15 years of staying ahead of SEO spam and providing useful search.
Lately it seems like they've given up on the arms race and let the SEO spam win, but it isn't clear why.
And Google didn't produce high quality search for free, they used ads and sold the eyeballs they won.
> Back in that era, if I were looking for something as simple as the University of Michigan, I clicked and drilled down through a Yahoo directory.
Google didn't innovate that much except to provide a clutter-free interface and slightly better search. Prior to that, I used Webcrawler and then HotBot. A search like what you described would have easily returned useful results.
People who respond like you have are either being willfully obtuse about how awful search has gotten or honestly do not remember how it was 5-10 years ago. As others have mentioned, SEO has all but made search useless for anything outside a narrow window of need. I used to be able to do "deep dives" into the dusty corners of the internet to find things using google and now it's about useless for that.
And I've /recently/ hunted for something obscure, couldn't find it, managed to find an old bookmark to it, the server was still online and the content I wanted was still there. And no amount of crafting of a google search would bring it up. And the server in question didn't contain copyrighted material which would have resulted in a takedown block or anything like that.
It's frustrating how /bad/ Google has gotten for anything other than fairly basic, high level "searching".
Google doesn't return a lot of stuff that is on the web, that is found by Yandex and Bing. It is unwilling to serve up long-tail results from its index, probably because they're so scared of accidentally serving up "misinformation" and they heavily penalize non-https sites.
And yet for some reason they're all too eager to serve up sites scraping stackoverflow.
What is actually the fraction of human knowledge available on the web? How does the trajectory look like? What is the fraction for knowledge produced the last 5 years?
It was an honest attempt on the crawlers' part and it was a more level playing field. If you didn't find what you wanted on lycos, you could look on Yahoo or Alta Vista or WebCrawler or something instead. For a time there were meta Search engines that allowed you to search other search engine amalgamated results.
Yeah, you had your choice of several awful results and everyone couldn’t shut up about how good Google was when it was available. If people mean “Google used to be better” I’m willing to entertain the idea, but you’re out of your mind if you miss using Altavista and Dogpile.
Just search for "chocolate chip cookie recipe" and show me a single result that isn't a complete trash website. A significant number are nothing more than ad farms that just load infinite ads in IFRAMEs and similar stuff like that.
that site has over six thousand words and maybe 20 screen heights worths of scrolling before you reach the recipie. Its incredible, like some kind of work of art parodying recipie websites
I use BBC good food, almost always straight to the point
Yes, because he's explaining all the different aspects of a "perfect" cookie and giving you options to dial in your personal version. He did 100 tests and 1,536 cookies for the article.
I get not everyone is a foodie that cares about the details and wants to tweak it, but I appreciate them.
Cue 7 paragraphs about how when they were on vacation in Ibiza they sought ought the most authentic local chefs and took inspiration from all of them to develop this never-before-seen recipe.
These two are at the very top of the page. They both are chocolate chip cookie recipes. One has instructional videos. I really don't understand what is wrong with these.
I suspect the future is far more bleak. At least with "search results", users can scroll down past the SEO content to something more relevant. Once these new chatgpt models take over, it's winner-take-all for search results. Everyone is going to be throwing money at getting their message seen, since it will be the final word.
How much do you think companies are willing to spend to be the answer to, "what is the most reliable car?" or, "who should I vote for?"
Obviously those are not questions that you can trust an AI language model with. You have social media or real life connections to discuss that.
In theory it is possible that sponsored content will creep in, but that does not invalidate the incredible benefits a well trained language model will have, even despite the occasional for-profit bias.
Google has been up against SEO for decades and the motivations are much stronger on the other side of the battle. ChatGPT is running off a fixed training set that no one has had a chance to game yet. When someone tries to build a true, current AI search engine we're going to see a whole new level of hallucinating lie machine.
Been working in SEO for nearly 13 years now - can confirm, a lot of SEO content spam is quite literally Googles fault. Google has forced just about every company to become a publisher as the only way to generate traffic to their site. I promise you the vast majority of companies don't want to spend the money pump out 4-30 pieces of mostly remixed/regurgitated content a month or screw around trying to attract links but this is how the algo works and you get what you measure.
Exactly. Every time someone with a business says "I want to get to the top of google" the person they are asking for help will respond with some jargony version of "you gotta constantly write content." It is the ONLY answer.
And unfortunately Google has become worse than ever at being able to differentiate between insight and fluff.
This is a fair argument to make. Normalizing web site metadata (which is essentially what SEO is) resulted in Google getting crowdsourced work done for them for free.
It has been a self-fulfilling positive feedback loop since then.
I can't wait for people to switch over from search engine optimization to AI model optimization: figuring out what kind of content you need to produce in order for Google's and OpenAI's models to deliver your message as gospel.
I'm actually expecting AI training data to turn into a snake eating its tail. I've been spending the last month not sleeping so I could try to get up to speed on deep learning, and my understanding is that AI trained on AI generated output becomes crap surprisingly quickly, and AI generated content is already starting to proliferate. I have no idea the extent to which this will hinder new models being generated, but I could see it becoming quite the problem.
the tinfoil take is that everyone in charge is fully aware of this eventual problem such that when it occurs, a solution will be presented: ubiquitous verifiable digital ID technology that's required to use popular social media services, such that everything that a Real Person posts will be signed with said digital ID, thus bringing about the end of online anonymity, and the death of the Internet as we presently know it.
idk, sounds plausible to me, the way things've been going.
Entirely possible, but the amount of people who moved to Mastodon after Twitter got lit on fire gives me hope that people might explore options if things go a step too far
It will be hilarious. Chat GPT is standing on the shoulders of there being no AI-EO, a bit like Google 1.0 did.
AI will be a victim of it’s own success too. Or it will need to be human researched and curated rather than just letting an algorithm run freely across the web.
The quality of Google search has decreased so much over the last few years that I had to resort to using Yandex, a company that has a budget that is orders of magnitude smaller that the Big G and is under the control of a dictatorship.
Google search has two huge problems: SEO and censorship. Search for anything related to products/torrents/streams/politics on Google and your results will SUCK, due to one of the two reasons stated above.
The recent Yandex hack/leak has the cynic in me connecting the dots and, seeing how Google search seems to be facing REAL threats to it's dominance since it's creation... maybe some guys with the deepest pockets in the world are starting to enter WAR mode.
Destroying Yandex advantages on the SEO battlefield by way of divulging their parameters to the world would be the Franz Ferdinand assassination moment of the Great Search War.
They force you to follow their rules so it is cheaper for them to index the site. Then they allow all the SEO hacked content so you have to pay to get above that in the search results. Fun times.
Are you asserting that they look at their copious data and decide to make search worse because it makes them more money? Rather than figuring out a way to make search better and then further optimize their advertising income with this better product? And it seems like they've been pretty damn clever about monetizing quality over the years. It's possible that they have chosen to make search worse for profit, wouldn't be the first time a business did something like that but they have a pretty deep institutional fear of search losing relevance and it's hard to see them doing that.
As a long time user and user of the other guys before google, I think Google is shockingly good at finding specific answers to specific questions that I have about all sorts of things, often with fairly deep technical context. Now what is definitely lacking is in the good old days I'd enter some search terms, get pages of results and then some time later I'd find myself enjoyably down some rabbit hole that is tangential to my search needs on some part of the internet that I never even knew existed before. Maybe I'm too busy with work, but I used to spend a lot more time doing internet "research" to get some specific answers, that time seems to be much more efficient; I do sort of feel like I'm corralled to smaller portion of the internet than I used to be. I don't feel like I can't find the information but I have had a hard time re-finding some specific web page I found once way back.
Where does ChatGPT and Bard fit in this? I've played with ChatGPT and it's fun, it's neat, I haven't been able to get it to some how synthesize some wisdom though. It's not hard to see it just mimicking things. That might be valuable. That might be fun. Using it to seed search might be an enjoyable thing. Maybe it can help extract context from people to find out the actual question they are asking to find the actual answer they seek. Now I can absolutely see ChatGPT/Bard assisting in me wasting time going down rabbit holes, I'm not sure it'll be as enjoyable or as magical as how it used to be.
Are there some examples of shitty google search you can bring up? I just entered "Roth contribution income limit" and without even going to another website, I got what looks like a legit answer to my question. Now I'll click though a few to make sure it's accurate and authentic; at a glance, it's coming from Schwab and it looks like a legit answer to my question. Bing comes up with the same answer, it's presented in a nice table but it's lower on the page and below a sizable paid ad from Merrill Lynch (edge?)
there is that service that you can pay for that offers "high quality" but i haven't tried it. can't remember the name off the top of my head but it was on hn within the last few weeks iirc
I don't get the HN hate of Google as a search tool. Yes, SEO has made searching more difficult, but Google is still by far the best search engine provided that your searches are focused and you use search tools (e.g., excluding terms, focusing on certain sites, etc.). I've tried other search engines (e.g., DDG, Bing) and they just aren't as good as Google.
Many of the HN crowd remember when Google came into existence and how wonderful it was.
If it existed on the internet, Google would find it for you and it was usually the top resault. It was amazing.
Today, it's a shadow of its former self.
You regularly have to search, wade through the ads that are written like informative articles, adjust your query slightly and repeat the process. It's rubbish.
This, exactly. Google was a revelation, a clean page with the exact result you needed, at a time when many orgs were putting lots of high quality content online. The other search engines were busy portals and weren't indexing the new content.
I mean, Google still often gives me the result, but it’s such a chore now. And they can’t even seem to deal with what I consider extremely obvious blogspam.
Content that is just literally directly copied from other domains often.
What these people are remembering is when the web was small and dorks-only. Now it is huge and full of normies of all kinds, including every shade of commercialism and fraud.
The number of internet users is currently 150x more than it was in 1996. I don't think you can make an argument that it was already mainstream in '96 with a few tens of millions of people on it.
The web was overwhelmingly informational up to an inflection point where it became overwhelmingly commercial. That's the thing people are upset about.
In 1996, did every traffic attorney publish a thought piece on why you should fight your summons in X county?
No, back then if you searched a topic you were MUCH more likely to find self hosted content from someone who nerded out on an issue and is sharing their insight, not publishing boilerplate because they feel they need to.
>I don't get the HN hate of Google as a search tool. Yes, SEO has made searching more difficult, but Google is still by far the best search engine provided that your searches are focused and you use search tools (e.g., excluding terms, focusing on certain sites, etc.). I've tried other search engines (e.g., DDG, Bing) and they just aren't as good as Google.
What if their search is better just because way, way way more people use it?
> you use search tools (e.g., excluding terms, focusing on certain sites, etc.).
Focusing on certain sites? If you know what site you want, why not just go there? You don't need Google for that.
I've had very limited success with their search modifiers. The main one I want to work is the literal search by putting things in quotes. But I don't think that has ever actually worked for me.
If they brought back the + modifier and it worked, that would also go far.
