3 ms·

The actual big goals appear to be to get GCP out of third place and to do something nebulously magic with AI. Both of these have made progress, but not in a way

The actual big goals appear to be to get GCP out of third place and to do something nebulously magic with AI. Both of these have made progress, but not in a way that actually makes the company money. Other than that I agree with you, there doesn't seem to be any sort of coherent product vision other than "ads in search/youtube" and "keep building the handful of actually popular products we've had for ages." It is definitely hard as hell to develop another business the size of ads alongside it, but it really doesn't seem like there's a clear idea here.

tomp 4y ago
"GCP is not shutting down. Rest assured we're always working on bringing more great servers to the platform."
- UncleMeat 4y ago
  Unlike Stadia, I do actually think that killing GCP would destroy the company. So I don't expect the same end outcome.
  - dekhn 4y ago
    Right, they're trapped in a terrible position where it's sucking the company dry of its profits while also being necessary to continue or they will lose all credibility with enterprise and developers forever. That said, they could shut down Verily and save some $1B/year.
    - UncleMeat 4y ago
      The thing that is bizarre to me is that Google has absolute shitloads of money. 120B+ in the bank. Megabillions in profit each quarter. Like, they could buy Nintendo for less than half of their cash on hand. I understand that the market demands infinite growth but internally it feels like horseshit to see the ridiculous numbers and also get told that basically nothing is high priority enough to get funded.
      - lmm 4y ago
        Nothing is high priority enough to get funded because they have a firehose of money coming out of ads, compared to which nothing looks like a good enough opportunity.
        gsibble 4y ago
        Exactly. The IRR in investing more into ad platforms for Google is still higher than any other project so they keep investing in it and increasing their ad revenue. Alongside Meta, they have a duopoly (monopoly?) on that market for now so they are just milking it for all its worth. You don't just build stuff for the sake of it. It has to be a better investment than what you currently have, and Google doesn't have anything better than doubling and tripling down on ads.
    - jsnell 4y ago
      Cloud's losses ($1.8B in H1) are about 4% of the operating income of the money-making parts ($46B in the same timeframe). Sure, a billion here and a billion there, pretty soon you're talking real money. But it's hardly "sucking the company dry of its profits".
      - dekhn 4y ago
        Much of cloud's real costs are hidden in other parts of the budget, such as global infrastructure expansion.
        jsnell 4y ago
        So you're suggesting that the accounting is fradululent, and nobody figured it out except you? Not the auditors with actual access to the raw data. Not the hedge funds who could have made a pretty penny by shorting the stock and then releasing this information. Not even any whistleblowers who would presumably be eligible for an award. Or maybe, just maybe, you've just made that up.
- xerox13ster 4y ago
  291 titles. Not great, not terrible.
ghaff 4y ago
>something nebulously magic with AI And even then I wouldn't place a bet on Waymo producing anything interesting from a revenue perspective for the next couple decades+. And, while I'm sure Google is doing other interesting things with AI--especially from an internal operations perspective--it's not like Google Home is doing anything earthbreaking or that Google search etc. is--from a consumer functionality point of view--particularly differentiated from other major players. And plenty of AI research is happening outside of Google.
caskstrength 4y ago
Don't forget about shoving Shorts into faces of their paying Youtube Premium users without a toggle to permanently disable it. That will definitely outcompete TikTok any day now!
- post-it 4y ago
  If they made it possible to cast Shorts like other videos, I might actually use them. I've tapped a Short several times because it looked funny, but it wanted to start playing on my phone instead of the TV. I watch TikTok alone on my phone and YouTube with my wife on my TV, so Shorts just doesn't fit into that routine. It's mindboggling that you can't cast Google Shorts to a Google Chromecast. My hypothesis is that because Chromecast casting is so buggy and unresponsive (compared to Bluetooth), swiping from one Short to another would be such a frustrating experience that they dare not enable it.
  - rightbyte 4y ago
    My 2yo smashed my Oh So Smart TV with the remote and I got a spare dumb TV and plugged in Chromecast into it since everything is streams nowadays. Dear God how annoying it is to use. Buggy, laggy, drops randomly ... still can't play videos from my computer in an easy way. I forgot how bad Chromecast were since I last used it 5 years ago.
    - Tostino 4y ago
      When the original came out ir was practically magic to be able to have the TV on while hanging out with roommates or friends, and all be able to contribute to the entertainment for the night right from your phone...then it didn't improve in any meaningful way, and regressed in others over the following 8+ years. Really sad with the state of things.
      - rightbyte 4y ago
        When he broke my TV I was in the store looking for a new one. I had one condition -- there had to be a sane way to write with the remote. None of the TVs had any sane way but used arrows and enter or some cursor marker for the on screen keyboard. Dunno why none used a T9 type of letter entering system. All remotes but one had number pads. It is really inconvenient to use "Smart" TVs. So I went with the spare for a while.
jrumbut 4y ago
Azure became interesting when it got better integrated into the Windows/AD/Office worlds. You're not having to manually integrate so much anymore. The way for GCP to compete is to give customers (presumably high level customers) some real deep integration into Search/Gmail/YouTube/Ads/Drive/Maps, including data. That would not only be a killer differentiating feature, it would signal to everyone that Google is serious and won't shut this down on you like they did all those other services. They probably won't do that, but I'm not sure why they really want to run GCP if not. If they're not going to give customers something only Google can give them, I don't see them capturing much more of the market.

tomp 4y ago

"GCP is not shutting down. Rest assured we're always working on bringing more great servers to the platform."

UncleMeat 4y ago
Unlike Stadia, I do actually think that killing GCP would destroy the company. So I don't expect the same end outcome.
- dekhn 4y ago
  Right, they're trapped in a terrible position where it's sucking the company dry of its profits while also being necessary to continue or they will lose all credibility with enterprise and developers forever. That said, they could shut down Verily and save some $1B/year.
  - UncleMeat 4y ago
    The thing that is bizarre to me is that Google has absolute shitloads of money. 120B+ in the bank. Megabillions in profit each quarter. Like, they could buy Nintendo for less than half of their cash on hand. I understand that the market demands infinite growth but internally it feels like horseshit to see the ridiculous numbers and also get told that basically nothing is high priority enough to get funded.
    - lmm 4y ago
      Nothing is high priority enough to get funded because they have a firehose of money coming out of ads, compared to which nothing looks like a good enough opportunity.
      - gsibble 4y ago
        Exactly. The IRR in investing more into ad platforms for Google is still higher than any other project so they keep investing in it and increasing their ad revenue. Alongside Meta, they have a duopoly (monopoly?) on that market for now so they are just milking it for all its worth. You don't just build stuff for the sake of it. It has to be a better investment than what you currently have, and Google doesn't have anything better than doubling and tripling down on ads.
  - jsnell 4y ago
    Cloud's losses ($1.8B in H1) are about 4% of the operating income of the money-making parts ($46B in the same timeframe). Sure, a billion here and a billion there, pretty soon you're talking real money. But it's hardly "sucking the company dry of its profits".
    - dekhn 4y ago
      Much of cloud's real costs are hidden in other parts of the budget, such as global infrastructure expansion.
      - jsnell 4y ago
        So you're suggesting that the accounting is fradululent, and nobody figured it out except you? Not the auditors with actual access to the raw data. Not the hedge funds who could have made a pretty penny by shorting the stock and then releasing this information. Not even any whistleblowers who would presumably be eligible for an award. Or maybe, just maybe, you've just made that up.
xerox13ster 4y ago
291 titles. Not great, not terrible.

ghaff 4y ago

>something nebulously magic with AI And even then I wouldn't place a bet on Waymo producing anything interesting from a revenue perspective for the next couple decades+. And, while I'm sure Google is doing other interesting things with AI--especially from an internal operations perspective--it's not like Google Home is doing anything earthbreaking or that Google search etc. is--from a consumer functionality point of view--particularly differentiated from other major players. And plenty of AI research is happening outside of Google.

caskstrength 4y ago

Don't forget about shoving Shorts into faces of their paying Youtube Premium users without a toggle to permanently disable it. That will definitely outcompete TikTok any day now!

post-it 4y ago
If they made it possible to cast Shorts like other videos, I might actually use them. I've tapped a Short several times because it looked funny, but it wanted to start playing on my phone instead of the TV. I watch TikTok alone on my phone and YouTube with my wife on my TV, so Shorts just doesn't fit into that routine. It's mindboggling that you can't cast Google Shorts to a Google Chromecast. My hypothesis is that because Chromecast casting is so buggy and unresponsive (compared to Bluetooth), swiping from one Short to another would be such a frustrating experience that they dare not enable it.
- rightbyte 4y ago
  My 2yo smashed my Oh So Smart TV with the remote and I got a spare dumb TV and plugged in Chromecast into it since everything is streams nowadays. Dear God how annoying it is to use. Buggy, laggy, drops randomly ... still can't play videos from my computer in an easy way. I forgot how bad Chromecast were since I last used it 5 years ago.
  - Tostino 4y ago
    When the original came out ir was practically magic to be able to have the TV on while hanging out with roommates or friends, and all be able to contribute to the entertainment for the night right from your phone...then it didn't improve in any meaningful way, and regressed in others over the following 8+ years. Really sad with the state of things.
    - rightbyte 4y ago
      When he broke my TV I was in the store looking for a new one. I had one condition -- there had to be a sane way to write with the remote. None of the TVs had any sane way but used arrows and enter or some cursor marker for the on screen keyboard. Dunno why none used a T9 type of letter entering system. All remotes but one had number pads. It is really inconvenient to use "Smart" TVs. So I went with the spare for a while.

jrumbut 4y ago

Azure became interesting when it got better integrated into the Windows/AD/Office worlds. You're not having to manually integrate so much anymore. The way for GCP to compete is to give customers (presumably high level customers) some real deep integration into Search/Gmail/YouTube/Ads/Drive/Maps, including data. That would not only be a killer differentiating feature, it would signal to everyone that Google is serious and won't shut this down on you like they did all those other services. They probably won't do that, but I'm not sure why they really want to run GCP if not. If they're not going to give customers something only Google can give them, I don't see them capturing much more of the market.