This touches back to Gruber's assertion that Google is getting better at what Apple does best (device design) faster than Apple is getting better at what Google does best (web services).
I'm a huge Apple fanboy but they really need to get their shit together on the web services front. Siri still sucks almost unbelievably bad. It never works. Maps still leads me to random streets and goes in circles.
As we move towards smaller and smaller devices one has to assume more of the work is going to be happening in the data center. Not good news for Apple unless they start making some real strategic acquisitions stat. (They've started at least.)
As someone that uses both platforms extensively, I find claims of Siri and Maps inferiority utterly mind-boggling. On an iPhone, I can get directions to anywhere or call anyone or create calendar events while barely looking at my phone. Google Now is nice in some respects but almost always requires an off-putting degree of annoying manual manipulation to accomplish actions as simple as calling someone.
For me, Maps gives me eerily accurate drivetime estimates whereas Google's times are utterly useless. And I consider the UI patently inferior. I feel like a need two hands and a full minute to do things that require a few quick taps in iOS. Google seems to need to ping their server to get my search history or home address, whereas iOS calls it up in milliseconds. I would also kill to have navigation on the lockscreen as in iOS - it always requires a janky series of button presses and swipes on my Android phone.
>For me, Maps gives me eerily accurate drivetime estimates whereas Google's times are utterly useless.
Fascinating. Apple Maps is always wrong for me, with the estimated time of arrival constantly changing as I drive, while Google (and Waze) are spot on.
Well of course the ETA changes while you are driving, that seems obvious. It's an estimate, it depends on current conditions, it can't be correct from the start.
For instance, just now I pulled up directions to the place I drove on Sunday. Apple Maps gives three routes, ranging from 35-40 minutes. Google Maps gives three routes, ranging from 29-33 minutes. In my experience, it will take closer to the time Google says, so if I'm using Apple Maps, the ETA will start out at 35 minute from now, and by the time I get there, it will be 5 minutes sooner. The only "current condition" it is depending on is that I'm driving faster over the route than it expects.
The estimate also depends on current traffic conditions.
I won't debate the accuracy of either case since it can only be an anecdotal and unproductive conversation. You were disputing the value changing over time, not it being inaccurate. Google does seem to give more optimistic estimates in general.
The back and forth of claims regarding Siri seem to indicate an uneven level of service. My own experience is that it works great just enough to fail frustratingly.
It's not perfect, but I would kill to have it on my daily driver Android. Siri consistently nails daily things like "Directions home" or "Call [girlfriend]" or "set an alarm for 10 minutes" or "Show movies near me", all manual tasks when I'm in Android.
Interesting. On my Nexus 5 all of those examples work perfectly without touching the phone. Since around the time of the Nexus 4 the voice recognition has been really good (for me at least).
Not trying to argue - it's just interesting how much variation there is for different people (for both Siri and Android apparently).
Well Android requires me to stop what I'm doing and fiddle with my phone to get directions home or call my girlfriend or get movie times %100 of the time. I'll take the occasional timeout since Siri seamlessly routes me to my house with a single button press/voice command.
I'd say it almost does this. But there are enough fit and finish issues (like, a few orders of magnitude beyond Siri's supposedly showstopping "occasional internet timeouts when in spotty network coverage") to make it functionally useless.
Just initiating the voice commands is a pretty telling comparison.
iPhone: hold down ergonomically placed home button until it beeps, say what you want (eg "directions home"). Siri waits a well-timed beat and launches maps.
Android: press unlock button (semi-awkwardly placed on upper right corner), do weird home button motion only used to launch Google now (drag the soft "home" button up). Say "OK Google" to initiate voice detection, say what you want. Then there's a poorly calibrated lag to make sure Now is doing what you want, a hair over 5 seconds maybe.
Then there are weird fit/finish issues. Android seems to do weird things when headphones w/ a mic are connected, so it takes several seconds longer to be able to say "OK Google" after launching. If you use headphones a lot, it's basically a dealbreaker. And I find a lot of fit and finish issues when Google actually performs the action.
I'd be curious to try the Moto X to see if its "always listening" mode mitigates some of these issues.
tl;dr Now is great as a morning homepage, but an abysmal failure as a Siri-like assistant
You're comparing them on uneven terms. Show me one company besides Google that can even stack up to what Apple's done in Maps or with Siri. There's none. That's because both require amazing amount of expertise, money, and data to deploy properly. Apple has correctly done what they had to (break off the data flow from iOS to Google) even where it may have reduced the experience for their customers. That's a long term play.
Funny how Google is continuing to corner the data market via acquisitions like Nest. Almost feels...anti-competitive, no? I wonder how long until we start talking about Google the way we still talk about 1990s Microsoft.
Is it plausible that a new company, starting from scratch in the Maps/Search/Personal Assistant/etc. field could actually compete with Google? Or would they just get crushed under the personalizing power of Google's raw data and computational advantage?
> Show me one company besides Google that can even stack up to what Apple's done in Maps or with Siri
Well Microsoft, easily, if they had the will. I guess the pieces are actually all there, but they aren't as direct with posing it as a "personal assistant" or whatever.
If you consider the products separately: there are several major mapping companies, though there have been a ton of confusing mergers and splits in the last 5 years, not to mention OSM. As for Siri, it was a tiny team at Siri, Inc that created the foundations of Siri. Apple has done a ton with it (though I was hoping for more by now), but you can easily imagine a next-gen Siri coming out of a startup.
> Funny how Google is continuing to corner the data market via acquisitions like Nest. Almost feels...anti-competitive, no?
This becomes a little farcical, honestly. First, you can get far more information from a phone, and there are like 3 orders of magnitude more Android phones out there than there are Nest devices.
But more importantly, if we're going to talk about the "data market" -- as if that's some sort of thing -- virtually any acquisition that has customers would help them corner it. But that also means any other company with customers also has a stake in that "market", in which case, no, it does not feel anti-competitive.
I think you underestimate the importance of context in all of this. You can't build any of these products well without knowing (deeply and certainly) users' patterns. Microsoft has no realistic way of knowing when it routes you to the wrong place...no continuous training data to let it feed and refine a personal assistant. Data products can be cool demos (see Siri) from a startup, but without both the starting dataset and the continuous feedback that you get by deploying these against real users and capturing the resulting behavior, I think they are quite limited. Continuous adaptation and learning is as important as the initial algorithm.
And while I certainly am looking into the future when I talk about cornering the market, I think it's an interesting question to ask how you compete against a company like Google that starts with this much user context. How can you build better experiences when they have the advantage in data, ML expertise, and the raw computational power? The incremental acquisitions of data and companies are not the issue...it's the web of connections that is valuable. And Google is not only trouncing Apple, their only meaningful competitor, in that race...Google is pulling even further ahead.
Microsoft search share is around 18% in US and a number of those searches are local (Apple has none of those). Additionally, they have a growing number of Windows 8 devices which they will be using for local/maps and train their models. I have used Bing maps and though it is not as good as Google Maps, it definitely holds up against Apple Maps.
To your initial question that if there is any company which could have built Maps or Siri, Microsoft is more than a good answer. You can try to get really nitpicky on this but it is what it is.
Whenever I see a comparison to Microsoft's anti-competitive behavior, I wonder whether people don't understand antitrust law, what Microsoft did, or both.
From a competition point of view, antitrust isn't concerned about where you are, but rather how you got there. As Sherman put it:
"[a person] who merely by superior skill and intelligence...got the whole business because nobody could do it as well as he could was not a monopolist"
As far as we know, Google is doing nothing to prevent other companies, like Apple or Microsoft from making similar acquisitions and/or capitalising on the acquisitions they have made.
Microsoft told OEMs that if they sold more than X computers without Windows, their licensing cost would go up by Y. There's also their embrace, extend and extinguish (their own words) which describes their strategy to adopt a standard (Java), add custom behavior (J/Direct) and thus breaking the standard and weakening the product. And this isn't all of it, and they also did some stupid stuff in court (submitting faked videos and getting caught).
As long as Google's behavior doesn't hurt consumers nor does it unfairly disadvantage competition (and being too good and too big isn't enough, they need to actively undermine them), the comparison is absolutely and totally flawed.
This sounds very strange to me given the fact that by definition a monopoly is the control of a market by one entity, it doesn't have anything to do with how it gets there.
Having a monopoly is not illegal. Abusing that monopoly to gain unfair advantage in another market is illegal. People use the term "monopoly" to interchangeably mean "legal monopoly" and "illegal monopoly", which does get confusing.
Well that depends on the country. In some countries monopolies are illegal and the government is the only one that can perform ones.
Also, questionable practices are not necessarily illegal. For instance, no that long ago (well 20 years or so is not that long for me), having privileged information on wall street wasn't illegal, but questionable.
I've found that the data in Maps has been significantly improved (in the UK) over the last couple of months, its now actually possible to find businesses on it (road data has always been reasonable).
The more concerning thing to me, is Apple's apparent usability weakness when it comes to iCloud, syncing and its devices. I've worked it out now, more or less, but the potential for confusion is tremendous. If I turn on syncing, I expect syncing. If I turn off syncing, I expect syncing to stop, I do not expect warnings about documents being deleted from various devices.
Similar with calendars. Turn on iCloud and suddenly you have calendars 'On My Mac' and in iCloud. All the calendars are duplicated. Horribly confusing behaviour for someone who is new to it.
If I turn off syncing, I expect syncing to stop, I do not expect warnings about documents being deleted from various devices.
This has been a problem with iTunes syncing as well. Constantly warning me about deleting apps and photos when I check or uncheck a box. Does anyone ever want to say yes to any of those questions? Very user hostile behavior.
There's still no easy way to randomly get photos from a computer onto my ipad. I have to use some 3rd party app or frickin mail them to myself. The photo syncing feature is useless.
Do you? apple maps just led us 18 minutes out of the way just this weekend. It's apparently incapable of leading people to restaurants just off 80 in Auburn, CA. When we gave up and used decent software -- ie google maps -- it got us straight there.
And other times, Google Maps leads me astray or takes me down inappropriate roads while Bing Maps/Nokia Maps works just fine. It's almost like they're all getting their mapping data independently and as such have their own unique strengths and weaknesses.
The name of the local university is wrong in Apple Maps. It's using the 30 year old name. I've sent them corrections (via the maps app) a couple of times, over 1.5 years ago. No change.
To be fair, I have sent corrections to Google Maps and it took them two years and talking to a friend in Maps before they corrected it. Also, this was for Google telling me to take a left where no left existed when driving from the South Bay into San Francisco.
There is mapping data, and then there is place data (where a business is located).
Google and Apple must use the same mapping data, because I see the same errors in both apps. Of course with Google, after I report the error it gets fixed. I can’t say the same for Apple. And Navteq* has more detailed and accurate data than either of them.
What Google has is superb place data. Apple has… Yelp.
*Assuming you have recent Navteq data. Not the nav system in your car that hasn’t been updated in 2 years.
You never run into the "I'm really sorry but I can't handle your request right now" message for Siri? That is just a total killer to that service, because it takes 15-20 seconds just to error out. And sometimes when you try again, it happens again. And then you ask yourself, "why am I using Siri anyways?"
Yes, I'd say about 20-30% of the time I use Siri it times out or hangs either when it parses what I say or trying to resolve it. A few weeks ago Maps lead me to a parking lot when getting directions to a restaurant. In Mountain View.
They work in the US ( and perhaps some other "first world" countries ) but in places like India, data for even big cities is practically non existent while Google Maps has everything ( and up to date ).
Can someone explain to me why Apple didn't acquire Nest?
He's a known quantity at Apple and he's doing hardware that would expand their market. I know they don't do a lot of acquisitions but this would seem to be a no-brainer for them.
Apple's MO has always been to do a few things, and do them well. Perhaps home automation is not on their radar. Google, on the other hand, is used to running a million different unrelated projects, so they are better setup for something like this.
Getting a return on 3.2bn seems like a risky bet for Apple. Google has the incentive to bring in Tony Fadell, but his disciples have already shipped many products at Apple.
I would speculate that it's because Nest's products aren't scalable in the way most of Apple's other products are. The appeal is too narrow, appealing primarily to ultra-techy folks or people building/redoing their home. Phones and mp3 players and computers are owned by almost everyone. I just don't see smoke detectors and thermostats having that same kind of broad appeal.
Nest is neat. I don't have one but my neighbor does and it's a fine product. I'm as "gadgety" as the next guy here, but even after playing with Nest and being impressed I still wasn't compelled to buy one. I dunno, maybe one day.
Apple makes acquisitions to improve their existing products. If the people at Nest wanted to improve Apple products, they wouldn't have left Apple and formed Nest in the first place.
Or the opposite of hubris. They want to stay focused on their product line. They're happy to put connected home products in their stores, as they've done with nest; but they recognize they are already spread thin, so they do not actually want to make products in that space. Apple's acquisitions tend to be targeted improving their existing product line, not rapid expansions into new products.
I would bet Apple could make a good washer and dryer, but no one is claiming Apple is not out there courting the best washer/dryer company due to hubris.
My guess from the kind of companies they buy, is that their products really need to fit with Apples area of focus. If not, they buy the company for the talent or technology (like FingerWorks).
So the only reason that they'd buy a company is if purchasing the company's technology lets them get a product to market sooner than if it were built in house? Or by doing an acqui-hire they can get a team at at the right price to fast forward a new product?
So you're saying that they've become a 'not invented here' kind of company?
Idunno, I'm an Android fanboy but at this point I'm getting less and less impressed with it. The UI looks dated now that everybody's going flat, and the "new device" experience is now terrible with tons of services being confused about your Google Plus account. It takes a day or two for a fresh Android device and account to stop bombarding you with irritating inscrutable confirmations and Plus-related shenanigans. I recently set up a device without the Google account and added it later, and every Google app was horribly confused after I added the google account - had to reboot it to make it behave.
Android is good, but it's quickly turning into Windows 98 - productive but messy.
At the same time Microsoft is making an impressive pace in both the hardware and web-service fields.
I feel like Google in 2014 is going to be releasing a lot of aesthetic updates. With the recent acquisition of timely + this new one, they've bought two teams at least partially known for beautiful design.
I have a hard time believing Nest sold due to scaling issues. I find an easier time believing that they did not have the money to achieve their (expensive) vision, and selling to Google was easier than raising a massive round. It's the classic "our product is good but doesn't make any money; solution: double down." Nest has a huge staff for what it currently sells.
Nest has a huge staff for what it currently sells.
I thought something similar. And then I worked there for a summer. Building hardware is much more difficult than you think, in ways that aren't immediately obvious.
Someone like Tony can easily raise capital in Silicon Valley, and Nest's products were selling extremely well across many retail locations. I'd be astonished if they were running out of money.
Why is it surprising that scaling is the issue? Tony's entire background is in devices and hardware, not large software systems.
They have a huge complex stack. Imagine building a system with tens (or hundreds) of thousands of connected nodes, all streaming realtime data that must be analyzed, and which can't ever go down. A Django app just won't cut it. And that's just the tip of the iceberg for Nest. The machine learning is surprisingly sophisticated.
I am quite sure that by "scaling" he meant the scaling of their business activities and not software.
"... Google offers to bring that scale to us. For me, ultimately building great products is key....
...you have to look at markets outside of the US. We are doing very well in the US and Canada but we need to get to Europe and around the globe. Just getting to UK has taken up a lot of time and energy and when you look at Europe, there are many countries and many opportunities. It is an atoms-based business and a lot goes into getting it to scale — legal, physical distribution and even localization are time consuming and need new kind of scale."
The thermostats don't cease to function if they can't contact nest's website. I agree that it's important, but it's no more critical than any other day-to-day website.
Also interesting is that quite a few municipalities forbid connecting the fire alarm system to any external system, unless the external system is UL listed. Could be some regulatory hurdles in Nest's future.
I don't think so. Nest was in the middle of closing a $150M round at a $2B valuation. That doesn't smell like a company going out of business, that sounds like a company preparing to scale like crazy.
Fadell has intimate knowledge of Apple's scaling troubles—LaLa, iTunes Store, App Store, MobileMe, etc—and Apple has infinite money. Building truly scalable web services is no easy task, and Google is head and shoulders more capable than anybody else.
Do not look at Nest as a thermostat company. Look at what they could be 10 years from now—the timescale Fadell is surely thinking on—and you begin to understand the real scaling issues they will face.
I see a $150M round as a failure: a company that has big dreams but can't make enough cash at what they currently do. Even Tesla only raised $200M at IPO, and that's for a car company with huge overhead and manufacturing capital costs. $150M for a company that assembles products from COTS parts in China is ridiculous.
You can see it however you want, but that won't change what it actually is.
Companies raise huge rounds when they've nailed their product, their production lines, and their sales funnel. They know what they're doing and need cash to accelerate.
One issue is that thermostatic systems apparently work substantially differently outside of the U.S you can't buy NEST in the UK, for example because of this wiring incompatibility.
I've wondered about this. To what extend does a Nest drive the CV system? For example does it do modulation? If so, does is use OpenTherm - and thus only works on new CV systems? If not, how is it not 10 (OK, 5) years behind?
He's a partisan, but he calls them like he sees them. I remember when the Original Nokia Lumias rolled out, he had good things to say about both the software and the build quality.
That's not true, most of the stuff he posts about Google is negative and positive about Apple. This article is the first exception I see in years (and it's not even positive about Google, it's just neutral, which is already quite a departure from his usual prose).
I've been detecting a subtext of exasperation in some of Gruber's posts about Google vs Apple. Gruber, of course, is not an Apple fan-boy, per se. He's a tech fan-boy. He has displayed plenty of love for Amazon, and though he is very shy of Google's sleazy tactics, he clearly admires certain aspects of Google's enterprise.
But it's almost as if he's beginning to wonder about Apple's cojones.
They just bought a ton of robot companies. Now they are buying a company who can do consumer products - home automation stuff, in fact. They have Ray and that 512 qbit quantum computer. I can see what they are doing, I think. They are creating Google Robot. They were always going to do this. Google-Bot. Android. The challenges to go to the moon.
The "quantum computer" is a good marketing gimmick, but the problem it solves is fairly uninteresting. See http://www.scottaaronson.com/blog/?p=1400 for some critical discussion of the D-Wave machine.
The Ising model Hamiltonian that the D-Wave computer "evolves" almost directly maps to the energy equation you have to minimize to train a restricted Boltzmann machine.
I wouldn't call this "uninteresting" given the rapid advances in leveraging Boltzmann machines for deep learning.
Ok, I agree that it can be interesting for some cases, but it is not at all what people expect when they hear "quantum computer".
In addition, the comparisons with traditional methods for solving that problem show that you may as well solve the problem on normal hardware. Then the problem it solves stated as "Ising model Hamiltonian using quantum computer" is fairly uninteresting given that it is cheaper and simpler to solve the "Ising model Hamiltonian using classical hardware".
From what I can tell, it's a side effect of the way they architect the inter cell (a cell is 6 or 7 qubits) communications. This points to a larger problem of course, where we can show entanglement within the cell but cross cell entanglement isn't as clear (to me, but I might be behind the curve on this)
More than getting Google into consumer products, it'd be great if this pushed them to launch the platform of the internet of things. Just like Android unified the market like nobody else would, nobody else is going to give small consumer device makers a way into that internet of things and it'll end up horribly fragmented without something like Google providing a platform. Without that platform I can't see how it could go mainstream.
... with that in mind, knowing Google's shortcomings?
If I were Google, I'd blow a fat wad of those billions and buy Sony. There's a company that would really fill in a lot of Google's gaps - gaming, consumer hardware beyond the cellphone scale, a big library of media properties to integrate into the Play store.
A quick Googling puts Sony's market cap at $18B. A big chunk of Google's cash reserves, but imagine how terrified Microsoft would be of Android-powered PS5s.
"Microsoft makes more than 75% of its profits from Windows and Office. Less than 25% comes from its vaunted servers and tools. And Microsoft makes nothing from its xBox/Kinect entertainment division, while losing vast sums in its on-line division (negative $350M-$750M/quarter)."
I realize they're trying to position the XBOX One as your home entertainment hub - but the bigger question is whether there's a future for home gaming consoles at all, or whether that functionality will just be folded into your TV, cast from your mobile devices, or handled by some other connected multi-purpose device.
Beautiful understated home automation designed better than any other company has designed similar products in the past. You should really take a look, I'm surprised that it's possible to not have heard of them before.
for 99.99% of the population a simple thermostat works just fine. And nearly all of them don't even understand that the temperature setting doesn't actually change the temperature of the air, but rather the duration it remains on. NEST is pretty, but way over engineered for the task. Consumers just don't need it. I'm all for technology and am an early adopter and of course have heard of NEST, seen them etc. Like you say, it's hard to imagine some one here not having done so, but my guess is the commenter has seen them, but just didn't take it in because "home automation" has a long legacy of being utter crap. When you get past the pretty design, there's nothing really to the thing. It's just a massively over engineered pretty consumer device that very few people need. I'm pretty sure Apple didn't buy them because they could just as simply designed their own. I fail to see why the thermostat can't (if Apple were to design it) not have a UI at all but rather just be a little plugin thing, maybe "Thermoport Express" and you use your iPhone to make changes from anywhere, don't even need to get up and look at the device. Any hardware solution quickly looks old. The iPhone (to Apple) is the UI for everything.
Nest claims to save people, on average, 19.5% on their heating and cooling bill. If you're paying $50 per month for heating/cooling and you buy a Nest, you're now paying $40 per month instead.
Over a few years time, Nest quickly pays for itself, saves you money every month, and reduces your energy footprint.
It doesn't seem like a "massively over engineered pretty device that very few people need" to me.
It depends on how much you care about energy efficiency.
I, for instance, run 3 separate computers 24/7 when I could probably get away with combining them into 1 or 2 if I wanted. I could also turn them off when I'm not home. I just don't really care.
Despite your personal preferences, I don't think it's really up for debate that a lot of people would, in general, like to spend less money than they currently spend.
> the temperature setting doesn't actually change the temperature of the air,
I am sorry... yes having a higher temp setting means the heat will on for longer (or ac for less) but it DOES change the temperature of the air. The thermostat measures the temperature and turns the mahcine on and off accordingly.
It doen not immediatly change the temperature. It acts by proxy.
If you never met one of those people that say "hey, it's too hot here, let me lower the thermostat" just after they turn the AC on, you are a very lucky guy.
I didn't argue with the rest of his point. That he said that the thermostat changes the time the burner is on is not true. A thermostat is not a timer. Time on is not the controlled variable.
Come on, now you're taking semantic quibbling to farcical levels. Yes it is true that a thermostat is not a timer, nobody claimed that. What it (in it simplest form) does is turn on the burner until the set temperature has been reached.
My point was, for any reasonable, non-autistic reader, it was perfectly clear that the GP knows how a thermostat works and that his point was that many people think the thermostat changes the temperature that the burner generates, or at least whatever is output by that burner/installation (be it water or air). Now I don't have an opinion on that matter, but I'm annoyed by your littering of the conversation over a slight sloppiness in expression in a context that doesn't call for the level of correctness you, from your high horse, are calling for.
> many people think the thermostat changes the temperature that the burner generates,
There might be many such people but that is not what he said. I apologize that my comment clashes with your aesthetic preference of what a discussion should look like.
(that page cites Consumer Reports as these will save you $180 per year, edit: and interestingly lower down it says that in actual deployments they may slightly increase bills/energy usage, main culprit? the horrible UI)
But a) even to get the basics they have a UI like programming VCRs, and so can't be what you refer to as "simple" and b) they have obvious failure modes that need babysitting and manual intervention like it taking longer to heat your house up during cold winters (or just an unusually cold day) so the time it comes on in the morning should be slightly earlier.
It's a real UI challenge, that if solved will save money and energy (which seems to be a thing Google is interested in generally).
There are already products that simply shift the UI into a smartphone app (like British Gas's Hive https://www.hivehome.com/) and otherwise are mostly just programmable thermostats but they mostly still have a box (for manual overrides) and cost as much as Nest, without looking as nice or having such clever software.
I think Google would probably disagree with you (and they'll bet $3.2 billion on it.)
Nest Thermostat has various concrete benefits, such as energy savings, general ease-of-use, and aesthetics, but more importantly it signals the beginning of other thoughtfully designed hardware experiences for the household from the company.
To clarify, it's a very small part of home automation. Just thermostat and smoke detectors (as of recently). They haven't even branched into locks/lights/etc. I'm really excited to see where they go.
After all we now know about data being collected about us, against our will, and sooner or later used against our interests, is it smart to have Google also have data about your home?
This development can't be stopped. Since one company is reckless enough to push and benefit from it, others will follow and it will become the norm. It's not the first time that potentially malicious technology gets sold on apparent advantages for consumers. If that doesn't work, they can still bribe government entities / the EU commission (see smart meters).
> “People who are really serious about software should make their own hardware.” That’s never been true of Google
What about the pluto switch or the custom servers? Sure, they don't make chips or maybe don't even do PCB layout and all of that stuff, but I'm sure they have great hardware designers and embedded software people, no? I didn't even mention the google glass, mobile devices, and self-driving cars.
He mentions that in the second-from-last paragraph:
> The software that Google was most serious about — web search, Gmail, and so forth — ran in the cloud, and with the company’s legendary data centers, they effectively built their own hardware.
> In a sense, Google has always followed Kay’s adage. The software that Google was most serious about — web search, Gmail, and so forth — ran in the cloud, and with the company’s legendary data centers, they effectively built their own hardware.
