"A third of the rise in the S&P 500 stock market index this year is attributable to Apple Inc., Amazon.com Inc., Google parent Alphabet Inc., Facebook Inc. and Microsoft Corp."
And yet don't those companies directly employ only around 150,000 workers in total?
Yes, Apple off-shores employment to Foxconn but they are open about their plans to automate their factories and replace all the workers.
This isn't a matter of patience, it's waiting on the never-never.
Yeah, that part made me cringe. The fact that the majority of the profits are accumulating at a few players is actually a sign of a late stage market after consolidation with monopoly profits.
Productivity gains only drive profits if they provide a sustainable competitive advantage. There is a reason Buffet stays from IT - we improve productivity all the time but he can't figure out where the moat is. Apple (iOS/Hardware engineering), Amazon (Warehouse, Scale), Google (Advertisement market leader), Facebook (Attention) and Microsoft (Office) have some form of sustainable moat. The broad market less so.
It's not because of the IT dominance per say, it this:
"Apple's mound of cash and cash equivalents continued to scrape the cloudline this quarter, coming in at $268.9 billion in the company's quarterly earnings report."
(https://www.cnbc.com/2017/11/02/how-much-cash-does-apple-hav...)
He is betting that in some shape or form, they will put that cash to work or issue a nice dividend.
that's not the argument the article was making - they note that tech and media (the fastest growing sectors, which not coincidentally employ most intensively IT workers) are just 5% of economy.
the argument is that when the rest of the economy employs IT at the same level of intensity, the remaining 95% will have similar productivity growth and hence see gains of wealth creation.
we then have to work out how to distribute those gains (hopefully fairly and equitably)
There is a oft quoted parallel with electricity replacing steam in factories- it took a generation or more for factories to stop being clustered around one large steam engine, running each corner of the factory to the speed of one central spindle. the idea is that eventually companies will adapt to use "Digital" rather than adapting digital to fit them.
As a mental sanity check, think about just how, ignoring ML/AI, just how advanced you could rewrite any company you have worked in in past ten years. All the crappy inefficiencies, the pointless spreadsheets. And you have probably worked at the tech end of the economy. is today's IT usage really the best we can do as a society? i doubt it - there seems to be a lot of room for improvement...
There's a way around human nature in this case: one company can do it efficiently and force the others to do the same, or go out of business. Not that common though really.
It took the success of Uber for my local cab companies to offer an app to order a ride, and cut the prices for a taxi. I doubt it would have happened without that.
That's really not the author's point. I think this quote illustrates his point better:
"Online shopping came along in the 1990s but retailers struggled to adapt business processes to the internet. They needed to build complementary infrastructure such as fulfillment centers, and, the authors note, customers had to adapt their habits, as well."
The lag between commercialization and widespread adoption is wider than we think.
If Sears and Service Merchandise had simply held out for a couple of years and put up a web store front - the Dow would look a little different today.
Customers had adapted in the late 80's to something like Amazon (only problem was waiting for a few minutes after checkout) - but then they adapted to big box stores (made possible by big banks and supply chain software) - which ended up being a huge diversion of resources in the late 90's early 00's. The empty husks of retail are a testament to that.
Hmm... that presumes sears (the organization) can change fast enough to identify and embrace the right business model. There's more than a website to running a successful e-commerce business.
Internal culture at Google / Amazon, especially during the growth phase, was likely very different than your typical Fortune 500.
It's not clear to me why Service Merchandise (and its predecessors like Best Products, Consumers Distributing, and Unity Buying Service) all failed back then. They certainly couldn't compete now with the economies of scale of Walmart or Amazon.
It's odd that online grocery shopping is so huge now given Webvan's crash and burn in that space not all that long ago.
Lag is probably the wrong word. It's a misconception that adoption should be instant. It took a good 50 years for the electric motor to supplant the steam engine, despite it being superior in virtually every way. Tech adoption takes time, given that many things need to be re-learnt/engineered.
It's why i'm positive that there won't be swathes of unemployeed people due to tech. It will happen gradually, as it has always been, and people will have to be shifted into different and more productive employment types. It has always happened this way, and will likely continue to do so.
As long as there are more productive employment types available.
Anyway, I disagree. I think the sheer speed of the change will create a disaster much sooner than automation quality will make humans obsolete. It took more than a century for the steam engine to replace animal engines around the world, then the electric engine replaced it in half a century. Things are accelerating, and are already much faster than by that time.
I don’t think “it’s always happened this way so it probably will continue “ (although you explain this further) is a very sophisticated or for that matter accurate analysis. There’s the status quo and then there are major events that upend it. If anything could ever upend the status who like this, it’s tech. It’s like business cycles.. usually things are fine but they aren’t always fine.
Converting a user operated machine from a steam engine to electric engine doesn't remove the user. Converting to a computer operated or assisted machine does.
If the electric engine can produce the same things ten times faster, you've just removed nine-tenths of the users. Sure, someone is still around, but then again, automatic factories still have some people supervising them - just much fewer.
Increasing worker productivity has always lead to fulfilling larger a demand.
What is missing from all these discussions is that human wanting is essential infinite. The more productive you can be at making something, the more it will be taken for granted and a new thing wanted.
Demand is infinite. Work is created to supply it. There is no amount of labour that will ever exhaust it.
All automation does is increase the space of "taken for granted products". The infinite space of potential products/services for human beings to supply still exists.
80% of the UK economy is already services from people to other people.
This ignores diminishing returns of additional labour and consumption.
If we reach the happy situation where robots are providing us with the things we need for a good life almost automatically, why would we continue to work ourselves to the bone for additional goods and services that won't increase our happiness much? Most people would rather just work less.
In fact, our current level of economic output is enough to give every household a comfortable life, even if we worked substantially less.
Unfortunately the system is structured to maximize labour time, keep us economically insecure, and use us all as grist for the capitalist mill, rather than maximizing leisure time and quality of life.
There is a fundamental competition baked into existence in the form of land rights and human attention. Who gets to live where? Who are your friends? Who gets to go to that fancy college? While there are other ways of answering this question than "those who can afford it", it will always exist in some form.
While automation of necessities will make it easier to opt out of the capitalist mill, I would expect that to not be a tradeoff-free choice.
Demand is ultimately limited by human attention. There is only so much time someone can spend buying goods and services, especially if they intend to actually do something with them. And the demand for each individual product is far below that. There are some products virtually everyone will buy, but usually only once in a certain timeframe. Then there are specialized things like combines or milling machines where demand only goes up if some other industry expands.
If you improve your factory equipment to produce double the items per employee-hour at a lower unit cost, you likely won't see the demand double in proportion. Most likely you'll just eat into the market share of a competitor. Maybe some new customers will newly be able to afford the product, but once it has "taken for granted" status, even that is impossible.
And we are not increasing the number of these "taken for granted products" very quickly. As far as physical products are concerned, I can only think of the smart phone as a recent example; actually that decreased the number of those products, by obsoleting cameras and dumb phones for most people. Among internet services, advertising is the dominant money-maker, and that's a zero-sum game for eyeballs which only employs a relatively small number of people.
As more and more products and services are provided by machines, fewer and fewer people become necessary to keep things running in the background. Sure, some of the replaced people can try to come up with new things to offer, but most of them won't succeed, and the vast majority will eventually be unable to cost-effectively provide value to other humans at all.
At that point, we'll find out whether it's cheaper to feed them or shoot them.
You make some very solid points. However, most of the civilian guns are owned by people that most here would likely associate with being unable to cost-effectively provide any value to other humans at some point in the future. A large portion of those who are in the military associate with such people. Who is going to be performing said shooting? Just a thought.
Human wanting and economic demand are not identical. In order for human wanting to turn into economic demand, those who want need to have money (in a capitalistic society). If there are limited numbers of jobs and limited forms of need/want-based income redistribution, then demand is limited.
It has been a while that demand stalls. We can see it by observing that energy (time, investment, ...) doesn't go much any more to production means, but to advertisement, marketing, new forms of distribution, and so on. Production is enough in many domains (and there is no room for newcomers, it contracts rather than develops, you just compete to take shares from someone else, not that much to contribute to a booming market), the current problem and aim of companies are to invent ways to get people to consume more and more stuff.
It seems the world today the timing is entirely based on the country. In the U.S. the time it took for regular mobile phones to be supplanted by smartphones is ridiculously small yet the same cannot be said for many countries in the world. It seems to me that countries slow to make progress in tech will be the victims of it initially. It won't take that long to do so once the economic and logistical aspects of are ironclad.
At first glance, Amazon alone employs more than half a million people. Microsoft is over 100k, Apple is about 80k in the US, google's about 60k, and Facebook is around 17k.
I think that figure for Alphabet is off by at least a factor of two. Are we including staff worldwide, or just US-based staff? Full-time employees only, or also all contracted workers?
That ends up being rather circular though doesn’t it? Eg Amazon likely has someone who they employ but is contracted to sort of Netflix stuff. Who’s employee count do they add to?
It's not circular at all. Every single one of the employees I have in mind can be accounted for by Alphabet and Alphabet alone. We're talking security guards, cafe staff, customer support, janitors, receptionists, massage therapists, doctors, physical therapists, and any of dozens of different types of line of business stuff that doesn't use permanent employees (could be as simple as running the Google Books scanners or processing returned hardware).
It's pretty simple to determine -- Who's paying the salary? Your example, of a contractor who is employed by Amazon to work on Netflix integrations for Amazon products, would count for Amazon, as the work is paid for by Amazon and is for Amazon's benefit. It wouldn't count for Netflix; we're only talking about employees whose salaries are being directly paid for by a company, not larger second-order ecosystem-type effects. That gets a lot harder. What I'm talking about is exact.
Yes, contractors are not full-time employees. So what? That's not at issue here. In this context, we're specifically talking about the number of jobs that are created by tech companies. It doesn't matter if you're paid on a W-2 or a 1099 from the tech company directly, or on a W-2 or a 1099 from some contracted company; if your living is being derived from working for that company in some capacity, it counts as a job created.
In that case we're talking about a few million jobs because the parts manufacturers, service providers, all applications and application developers using the systems and many others. Fine with me, I was originally pointing out that the number was greater than the 150k the person I responded to was assuming.
In before all the posts lamenting "AAAFM" and its various impacts on the tech community without explaining that it refers to Apple, Amazon, Alphabet, Facebook and Microsoft.
Income for the top 1% of the population is up over 400% over the past 30 years, while remaining largely flat for the rest of the workforce. The tech boom is well underway, it's just that only a tiny fraction of the population is capturing the vast majority of the benefits. The boom is not just about efficiencies, it's also about centralization and winner-take-all market dynamics.
I see this fact touted around a lot, often in the context of stagnating/eroding wages.
It 100% goes without saying that global poverty decreasing is cause for celebration, but eroding quality of life for a economic bracket of people should not be ignored.
If you look at both facts together however (global poverty is down, wealth has concentrated at the top) then it would suggest that the wealth of the non-wealthy has merely been redistributed. Indeed, this is intuitive if you consider the trend of moving your labour to a cheaper country.
Having travelled to many developing nations, it also (admittedly anecdotally) feels evident that most the wealth reaching developing nations is concentrating with the already wealthy / those in power.
So what it feels to me is like that local inequality is increasing while global inequality is decreasing - the rich are getting richer everywhere while the working and middle classes are slowly converging everywhere.
Unfortunately, as history has so far told, the notion that the erosion of the quality of life has led to improvements of life to someone very far away is a hard sell.
I don't think we have seen anything like this in history before, and it will be interesting, to say the least, at how all the millions of people elevated from destitution will contribute to the world and change it.
Don't get too excited. The chart only shows that fewer people fall below the $2/day line. They could all be living on $2.01/day now and it would still be valid.
Starting at the bottom of the Maslow’s pyramid and moving up, do today’s wages buy less of food, shelter, water, clothing, transportation, etc. than someone receiving wages 30/50/100/200 years ago?
32,424.83 in today's dollars for a respectable, quality 8 room house.
Sure it doesn't include land or labour but there's no way an 8 room house costs 32k today.
Today that would buy you a Tiny House on the low end of that market.
In my opinion, and I know we are far richer than our great-grandparents, but anybody arguing we're quantitatively richer than our parents has a long row to hoe. If this is progress it sure is lumpy. As Peter Thiel says the most likely explanation is we've had an nearly completely unacknowledged (by Western intelligentsia) stagnation technologically.
> Tyler Cowen points out that there's not a single thing in your list of kitchen appliances that didn't exist in the 50s. People have highly temporally displaced notions, believing that things like washing machines or dishwashers are recent mass consumer items but they're not, they're nearly a century old.
Economists know that GDP measures are skewed by the fact that “a car is not a car” — that is, a new car today is much more than a new car from 30 years ago. It is safer and more comfortable, and consumers who pay the same amount in real dollars for a new car today are better off than consumers who bought a new car in years past.
So the “house” from the Sears catalog isn’t likely comparable to a house today with the same number of rooms. For example, the size of bedrooms, bathrooms, kitchens, closets, and garages has increased greatly in the last 100 years. New homes are more energy efficient and have safer wiring. Does this close the gap with a $32k house? Not in Silicon Valley, but there are parts of the country where the house (excluding the land, which Sears didn’t include either) is worth around $50k.
For example, the size of bedrooms, bathrooms, kitchens, closets, and garages has increased greatly in the last 100 years.
Have you been to the UK recently? :)
I don't know how much of an increase there was from about a century ago, but for the last few decades it's very much been going the other way. But the UK is something of a basket case when it comes to housing; people want to spend all their money on a mortgage.
Tangentially related, not long ago a minister suggested that people could keep a "jerry can" or petrol in their garage and was roundly mocked not only for the fire hazard, but for the casual belief that everyone has a garage.
Personally I've always felt the existence of the garage to be an underrated contribution to American innovation.
I don't mean Americans aren't aware of it, I mean that sort of thing never comes up when Europeans discuss how to increase innovation. They will produce 'e-centers' and 'hubs' but I think most of the good stuff just comes from some people screwing around, probably with some things that don't rate much fanfare or seem adequate for a proposal.
Sorry about the edit, I thought I was waffling on too long.
My position is not that you're wrong, it's that most of us are missing the subtext of our technological development which is that nearly all of it is computer related and we've papered over failures in a large number of technological areas by simply not talking about them (in society, not on HN).
> Economists know that GDP measures are skewed by the fact that “a car is not a car” — that is, a new car today is much more than a new car from 30 years ago. It is safer and more comfortable, and consumers who pay the same amount in real dollars for a new car today are better off than consumers who bought a new car in years past.
That is true. I wouldn't dispute the affect of accumulated incremental changes to existing product lines, it's meaningful.
It's true in two directions. For instance many items today are inferior goods to past goods. Here is an example.
New growth lumber is inferior to old growth lumber. Something made out of wood such as siding, framing or window/door frames will require more frequent replacement due to rot, insects, weather. The old growth equivalents lasts for decades and even centuries. Today we rely on a mixture of more frequent maintenance and often toxic additives (protective coatings) and if that doesn't work we are forced to leap to far more complex, sometimes expensive replacements e.g. UPVC, I-Joists, LVLs.
I'm not saying the technology is bad. It's great engineering ingenuity, I'm stating its requirement exists because of a failure in our ability to match or improve qualitative tree farming and botanical biotechnology.
> So the “house” from the Sears catalog isn’t likely comparable to a house today with the same number of rooms. For example, the size of bedrooms, bathrooms, kitchens, closets, and garages has increased greatly in the last 100 years. New homes are more energy efficient and have safer wiring. Does this close the gap with a $32k house? Not in Silicon Valley, but there are parts of the country where the house (excluding the land, which Sears didn’t include either) is worth around $50k.
This is a big subject, I'll just throw one dart at you.
There is a hidden assumption. Why are your houses more energy efficient? It's because a barrel of oil doesn't cost $1 anymore.
There's a dark side to energy efficiency.
The side affects of higher insulation requirements is that houses cost more, studies on passive houses suggest 10%-20% more in the positive scenarios. The real problem though is that the most common forms of insulation offgas toxins and since houses are meant to be more air tight... The fact is that most buyers of houses and most builders of them aren't up to speed on what is required to do this sort of housing without health risks.
Right now most houses in my country, without the latest insulation requirements, suffer from moisture, mildew and mold problems, all of which is not positive for human health - this only gets worse with the new regulations because consumers don't understand the importance of hvac or building houses as holistic systems.
Most of my complaints seem to go away if we actually had a real biotech revolution.
When DNA was discovered, when Human Genome was decoded, they said the exact same things they are now saying about CRISPR, that's what troubles me.
> * New growth lumber is inferior to old growth lumber *
I totally agree that many products are not "made to last" anymore. A fancy blender can do amazing things compared to an old one. But it will probably not last as long, and that's a problem.
The solution I have found is crowd sourced or expert recommendations from the following sources:
Kevin Kelly's Cool Tools book and website.
Metafilter
Reddit's Buy It For Life subreddit
Finehomebuilding's Tool Guide
I find I buy less stuff, but of much higher quality or utility in this way. Some times I get something at a cheap price which works really well, like my Kiwi knives. Other times I have to be educated into realizing why something costs what it does and why or why not it's worth the money. Usually the answer is that the price is right because the object lasts much longer than the cheapo alternative or that the cheap version has a health drawback.
Recent Examples:
Merkur Safety Razor vs cartridge disposables ($$$) (factor of 100 decrease in cost!)
Cast iron skillets vs Teflon pans ($$$)
Tempered glass (container/bakeware) vs disposable trays/soon-to-be-cracked stoneware. ($$$)
Induction cooking vs Gas (health)
Expensive meat vs cheap meat. (health)
Glycerin soap vs liquid soap/regular soap ($$$)
Here's a clear example. I buy a hard shave soap in a nice wooden bowl that costs $30. That's way more than a can of foam, but I've been using this for over four years now and it's only half used.
I don't do this with everything. I have a Silvercrest blender that has worked for a year, expected lifespan of three years I think, about $30 but buying a Vitamix or Blendtec at $500+ wouldn't 'pay me back' in a lifetime.
[waffle]
The China market isn't a problem if and only if it is understood by the consumer that they're getting disposables, so they should stock up and throw away more frequently. That requires money and space a lot of people do not have.
China production is immensely valuable for the bottom billions of the world who simply cannot plausibly afford West levels of wealth.
The rub is of course that you and I are getting lots of boxes with broken appliances/tools.
Yes, but can you build the Sears house for 35k today without materials substitutions. I don't think you can, prices of commodities were probably lower then, which is not in line with received wisdom.
The reason why I mentioned a Tiny House or THOW is that the comparison is much closer, I think like most countries in the past there were no planning permission or taxes on house building. A THOW too is built by the owner, but it is 200-300 sq ft and not 1200 sq ft.
The website you listed suggests a 50k modular house costs closer to 130k to install once everything is worked through. Do you believe the Sears house cost ~150-200% more in the end? Sure they had to dig wells too but I don't think that is plausible.
It seems like there should be a collection of historical lumber prices somewhere. Just that comparison will be pretty informative about the relative material costs.
Dude, you have to either be drowning in kids or make nothing to fall under the official poverty line. But what catches people's eyes is the rise of those officially falling below that line. https://www.theguardian.com/society/2017/dec/15/america-extr...
Great news, for sure. But was that a result of the tech boom, or government policies? I'm betting on the latter. And the fact that the lowest rungs of the economic ladder aren't fixed in an absolute sense doesn't change the fact that an increasingly enormous share of the wealth produced by humanity globally is going to just 1% of the population.
When it gets down to it — talking trade balances here — once we've brain-drained all our technology into other countries, once things have evened out... once our edge in natural resources has been made irrelevant... once the Invisible Hand has taken away all those historical inequities and smeared them out into a broad global layer of what a Pakistani brickmaker would consider to be prosperity...
I'm waiting for the common folks to start hoarding money too and decrease consumption. Ford figured this problem out at the start of the 20th century but apparently everyone forgot about it (if my employees aren't well off enough to buy my product, I'm screwed).
Ford’s assembly line required higher investment in worker training, so churn was more expensive for Ford than other autoworkers. Ford employees did not represent a major buying market for the company and did not own vehicles in higher proportion when compared to the rest of the population.
If we decide that we want to encourage companies whose final products workers can afford and discourage companes whose final products are unaffordable, we desperately need more manufacturers of pencils, plastic bags or napkins (single month’s salary buys one a lifetime supply) and need fewer companies like Boeing or SpaceX (even after a 100 year career the workers cannot afford the final product).
Although this wasn't your primary statement, wanted to take a bite out of your second clause about "encouraging cheaper product companies" as this is a point close to my heart as a regular advocate for consumer demand driven economics.
Boeing or Spacex's product demand is driven directly by transitive demand for an affordable product; air travel and to give one "space" example; gps, are both "consumer tier" products that facilitate a whole vertical of production. I'm hard pressed to think of a single super-expensive item that doesn't in some way link back to consumer demand; outside of additional injections via nations. To me this seems intuitive; as I'm hard pressed to think of another driver of production with comparable volume, and as such don't see any reason increasing that source of demand would bias "cheap" products any more than they are now.
I get the feeling that we're (the US) in a few bad ruts right now:
1. We have identity-based policies in place. People are attempting to blacklist others based on where they stand on politics/etc (non-money based discrimination). I've seeing this in the tech industry where the most passive-aggressive individuals succeed. (i.e. Tests we don't need them)
2. We're seeing lots of profit grabs and they're being ignored due to PR reasons. (I.e. Silicon Valley and workers rights.. oh wait, nevermind this startup isn't bad because they're "diverse")
3. Inflation, we're seeing a lot of inflation of prices, and decrease in value. (Go overseas to see how bad this is) Even Germany with a 20% vat on products is a cheaper COL.
When will people stop with that Henry Ford myth? You don't have to pay employees a high wage so they can afford to buy things that are produced.
You hear it when automation is mentioned: "Businesses are constantly automating production, so who is going to buy the output?". The answer is that other robot owners can buy it or the producer himself can consume it. The type of goods produced changes. Instead of cheap cars or pillows, you have more yachts, boats and other luxury items.
Ford famously raised the wages he offered, resulting in much better employees; however the only way he could make it so that his employees could afford to buy his cars was to make the cars affordable for the average family, since if he raised the wages to a level that meant they could afford his cars, he would only sell to his own employees and then he would have made no profit.
I won't comment on your assertion, except to say that Marx predicted the same, and yet it hasen't happened yet.
Money does trickle down but what it trickles down onto is a determining factor in economic growth.
When rich people put their money into employing people for non-productive servant type jobs, welfare for people (to not work), loans for people to go to college (to not work, at least in the short term), hand outs for their children and family members (to not work) then the end result is a lot of people not working and a lot fewer things being produced by the people who do work.
If a rich person buys a yacht or a private plane or a hundred million dollar mansion, that creates economic activity (people have to build that stuff) but it does not have the same secondary economic impact as investing that money into a factory, basic infrastructure, R&D, or some other productive use.
Money is trickling down, but far too much of it seems to be trickling down to a non-productive servant class and non-productive veblen goods.
It will NOT happen. It has never happened that way, and there is nothing especially unique or different about the economy today that would make it happen.
There were 500,000 new millionaires in the US alone in 2015:
The study, from market research and consulting firm Spectrem Group, found that there are now 10.1 million households in the U.S. with $1 million or more in investable assets, excluding the value of their primary residence.
That's up from about 9.6 million in 2013, and tops the prerecession peak of 9.2 million in 2007. It's the highest number since Spectrem began tracking the data in 1997.
If we cranking out half a million new millionaires a year, it should be sign more people are working to get into the top 1%, and the economy is still offering a ton of opportunity if you want to get out there and give it a go.
He's not wrong. In 2010 there were 6.93 billion people in the world. In 2015 there were 7.35 billion. By that calculation, 4.2 million people moved into the top 1% in the span of just 5 years!
I hope that’s inflation adjusted. A 1997 millionaire was an awful lot wealthier than a 2017 millionaire. It will be something around 50% inflation. If someone is a millionaire now that can simply mean they live in a moderate sized city and own their house, there are no Scrooge McDuck cash piles.
I have frequently wondered if the lag in productivity is partly due to the distracting power of the internet and social media. This isn't really paralleled by the other technical advances that the authors of the study this piece summarizes were examining.
Also, AI? We don't have anything even approaching AI in play now nor on the immediate horizon. The energy costs it would have are still enormous, even if we really could implement it. Maybe that's a revolution that's yet to begin. We got really good at problems that can be mapped into single precision matrix algebra, where the inputs and outputs are relatively small. It's a big jump to hard AI from there.
> We got really good at problems that can be mapped into single precision matrix algebra, where the inputs and outputs are relatively small.
General AI is a "wave of the future -- always has been, always will be" technology. Only when a piece gets broken out from general AI do we get practical benefits. The benefits from applying all that matrix algebra aren't small, though!
So we are breaking off small pieces of a bigger thing. If that thing is AI, does it already have a form and an older name in the world?
Would agree that in this article AI is a stand-in for what used to be called Math, Statistics, and Computer Science?
Our scientific output is an artifice, and it is "intelligent" in its ability to guide our action. To me it isn't then strange to think of the sciences as the "general AI".
As soon as you solve a problem in AI, people start labeling that problem as “just” math, statistics, or computer science. It is no longer considered AI. Shifting goal posts mean “real AI” is always in the future.
Borrowing from Iaian Banks' Culture universe, where aliens that are pretty much relatable as Earth humans (they're humaniform, even DNA-compatible reproductively, just not from Earth) are designated "1.0" on a sapience scale. See [1] for a discussion of the difference between sentience and sapience. An entity well below 1.0 (Banks never specified, but reading between the lines it felt like 0.5) is considered sentient, but not sufficiently sapient to achieve self-awareness. Personally, I'd prefer a scale that goes negative, to establish a bright line test that accommodates for example, Downs Syndrome patients.
We probably can claim 0.1 or 0.01 on such a scale.
Interesting and left too undiscussed IMO are the economics of productizing what currently goes under the AI umbrella. It is so expensive, and such a big task to automate with even the shard of AI we pursue today, that the numbers in many (though not all) cases don't seem to pencil out unless the benefits can accrue to a few dominant oligopolic providers, from nearly all human-staffed positions where each individual position is largely dominated by a single low-cognitive skill repeatedly applied. But interestingly, lots of physical tasks are left alone, which seems to point to a weakness in that economic model: despite joking talk of the coming robot apocalypse upon jobs, human jobs that require diverse physical and perceptual engagement with the environment are nowhere near common automation.
I see this in a lot of domestic household chores that have yet to experience disruption at the hands of AI. That's at least a 50M-count global market population willing to spend say, $2500-10000 USD per year on a robot that can in the beginning perform the housekeeping chores a hotel housekeeping crew routinely performs for every room. Be trainable on a per-item type basis to put up dishes and cutlery the same way into a dishwasher and back to storage, fold clothes and put them up, and mow the lawn: all highly-repeatable tasks that are only addressed by single-purpose, very expensive "AI solutions" today, tell me that if you are investing into any kind of general-purpose AI, invest long-range, like Uber-scale "point a money-hose at it" long-range.
People are trained by Hollywood to envision AI as robots. When presented an actual AI algorithm/model they see it as boring math, totaly unlike C-3PO, or Terminator, or whatever your favourite movie robot is.
Even if we claim that "real AI" is a system that can actively perform at human levels in creative tasks, I wouldn't believe that it will always be in the future. However, I don't think this is very near, and I think this is sort of what people are implying when they talk about AI. To me saying "general AI is what we're exploring and making mini-discoveries in" is a kind of apology. Unfortunately, this apology pushes us into a grey area where "AI" is really just the scientific artifice. That's not the same as AI that has auto-creative capacities.
What sort of creative task? Music composition? Graphic design? Surrealistic modern art? AIs have been programmed to do all these things and be indistinguishable from human creativity in blind studies.
I have no doubt through that you'll be able to come up with a reason why each of these examples isn't what you really meant by "human level in creative tasks." Which is rather the point.
How specific are they to those tasks, is what I'm curious about. Can the surrealist bot produce musical compositions as well as the musical composition bot? The generality angle would say that one bot can do all of those things mostly or entirely on its own (i.e. trains itself as well, and was not necessarily "told" to train in that area), not multiple discrete bots that are all heavily trained by a human.
Just to be clear, I'm not trying to argue with you in the sense that you convey in that second part of your comment, I just enjoy supposing about this particular subject.
That's certainly how it's been un until now and probably in the near future, but there is nothing apart from history that indicates it will always be in the future.
No physical laws preclude a machine that can operate independently and is more capable than humans at any human activity. If we ever make a machine like that, you won't be able to say that "oh, that's just computer science". Well, of course you could, but you would also have met the original definition of general AI.
Even if we had a general ai, fully indistinguishable from a human, people would dismiss it as a collection of ugly hacks, tricks and math, while completely ignoring that our own consciousness is as well.
Your brain could be considered a general intelligence, and it only uses a couple of watts. Energy consumption isn't the issue. However, I agree. AI is a bit overhyped right now, and AGI is still pretty far away imo.
It only uses a couple of watts (well, ten times that, actually, about 20 Watts), but that's because neurons are incredibly efficient. The brain is an Exa-ops computer that only requires 20 Watts. If we used today's state of the art TPUs, we're looking at 10,000 to 100,000 times the energy consumption, so on the order of a Megawatt. The electricity costs alone would be higher than the typical human wage.
A corollary: If you can find a way to grow neural tissue (I'm thinking of all the lab-grown meat startups here) and connect to it efficiently and reliably (I'm thinking of Neuralink, here), then you would have access to a computational medium that is orders of magnitude more efficient than even the best of today's processors, and probably way cheaper, too.
I think it's misleading to compare the brain to a computer, the architecture differences mean that the brain is great at some stuff (vision for instance) and terrible at others (multiply 2 64-bit integers in your head).
I don't know if computational meat will ever be faster than machines at number crunching. (when I say machines I'm including next-generation computers beyond silicon too). But that's okay, each seems to complement each other's weaknesses.
"Productivity" as economists defines it is purely GDP/hours worked, it is only loosely correlated to what a lay person would consider productivity. It doesn't consider the value of consumer surplus or whether prices for goods have fallen due to foreign competition or anything else, it is purely productivity in the sense of our ability to generate dollars for every hour that someone is working.
Is this definition flawed in the age where software can sometimes replace an entire profession? In such cases, can "productivity" sometimes fall in the face of technological advance?
It's tricky to predict what would happen to a statistic that measures the entire economy.
At a first approximation all the money going to that profession would be removed from GDP, but so would result in productivity losses if we automated industries with higher than average "productivity".
However, someone is doing the automating - are those people even more productive over the course of a year? Maybe, or maybe not and there's just a lot of investment that will bear fruit over the long run. But some amount of money is still being spent on the service, and the set of workers has gone down, so this would probably pull the number back up, but by how much?
But that money isn't gone, it remains in someone's pocket, rather than disappearing, and the price of some good has fallen. Does this cause people to buy more of the good or some other good? If they buy more of some other good that has low "productivity", such as food and drinks, then that would pull productivity down.
It would be hard to argue that such an outcome is unilaterally good though; we would hope that in the long run people find other high "productivity" activities, but whether we actually will is uncertain.
If you're willing to commit to a very long (and IMO enjoyable) read, Robert Gordon addresses these possibilities in The Rise and Fall of American Growth.
When people talk about AI and Robotics, I don't believe there is na understanding about how complex those topics are and how difficult it is.
AI: There's a lot of active research, and there has been a lot of aggressive pushes to be buzzword compliance there. I think this has a lot of potential to make improvements, but with AI you have to understand what you're doing there. (To be effective) I think there would be a lot of value in having research institutions on AI (and a consultancy).
Robotics: We've seen a lot of improvements on this, but I'm not getting the feeling that it's a mature technology. We're still having difficulty getting this to be easier to implement and working with it. The algorithms behind navigation, and motor control are intense and have taken a lot of research.
[meta question] Are most HN readers subscribers of wsj? I never see anybody mark these posts as pay-walled.. Either way I'm considering buying a subscription.
Automation is usually deflationary. Deflation is the easiest problem in the world to solve -- just print money. So, print it and pass it out.
As people spend the money, we don't have to give raises to the robots. If a few rich people end up with nearly all the money, then tax it back and pass it out again.
Maybe this was because I read TFA outside the paywall, but I find the included chart obnoxious in that it just has a vertical scale from 100 to 300, with no indication as to what that scale represents. Is it sales? Productivity? Number of press releases? Who knows?
Anyone else questioning this, “The integrated circuit was commercialized in the 1960s yet 25 years later computers still represented just 5% of the value of all business equipment”? ICs are everywhere around us (in our phones, cars, refrigerators, coffee makers, weight scales, HVAC, etc.) and are rightfully invisible to us. Let alone all the major business equipment. Or, am I missing something?
As pointed out, 25 years later was in the 80s. I wonder if some of the delay in adoption was due to the relevant patents, which were filed around 1959. Maybe that gave a two decade monopoly to Texas Instruments making it hard for others to adopt or innovate around the technology?
TI definitely did not have a monopoly on ICs for those 25 years. Those first patents were on Kilby's process which was essentially abandoned. Fairchild and Intel did a tremendous amount of work getting ICs to be more useful over that timespan (see self-aligned gates).
And yet don't those companies directly employ only around 150,000 workers in total?
Yes, Apple off-shores employment to Foxconn but they are open about their plans to automate their factories and replace all the workers.
This isn't a matter of patience, it's waiting on the never-never.