6 ms·

Between this and the payments to Apple, I wonder where Google would be if they invested another $30b into competing more aggressively... I guess these payments

Between this and the payments to Apple, I wonder where Google would be if they invested another $30b into competing more aggressively... I guess these payments make sense if you believe you can't effectively deliver something good with that money.

recursive4 3y ago
My understanding is that this is an example of last mile costs increasing non-linearly i.e. Google has something like 95% of the mobile search market and profits $30b from it, and the cost to acquire that last 5% is >$30b.
sdfghswe 3y ago
I believe what's holding google back isn't lack of cash, but lack of innovation. It's full to the brim with highly paid engineers that aren't really invested in change.
- rvz 3y ago
  So let's see what happens when they are threatened for the first time by Microsoft and OpenAI if Google can afford to spend even more money to prevent the likes of Apple, Samsung and Mozilla from moving from Google to use Bing with GPT 4.
  - smoldesu 3y ago
    Well, the big difference between ChatGPT and Google is that one of them is profitable. ChatGPT does not make money, and Bing's usage of GPT-4 is almost certainly not covered by ad revenue (even at-cost). Even from a technical perspective, I'd argue Google outpaces OpenAI. BERT beat GPT-2 and GPT-Neo, T5 and t5-flan bench well against GPT-3, and GPT-4 is so large and wasteful that it's not worth competing against. Relative to the rest of FAANG, Google honestly seems like the only company that actually knows how to use and deploy AI practically. Everyone else is trying to play catch-up with a proprietary competitor and/or pay an unsustainable sum to be king-for-a-day.
    - ffgjgf1 3y ago
      > pay an unsustainable sum to be king-for-a-day. I bet Microsoft can afford to spent unsuspecting amounts for way long than ‘a day’. At a certain point due to cheaper hardware and improved optimization GPT-N might become sustainable, by then their competitors might have issues closing the mile wide gap they have to MS/OpenAI.
- jimbokun 3y ago
  It becomes exponentially harder to find an innovation that moves the revenue needle perceptibly, when you are Google's size.
  - swader999 3y ago
    Or one that doesn't upset the search cash cow. Open AI for example. They seem to be inhibited by moat thinking too.
- crowcroft 3y ago
  Yea this is the point, all roads lead back to the lack of innovation at Google. They have capital, they have free cash flow, they can attract talent, but often they completely fail to compete effectively in new categories they enter.
  - mrweasel 3y ago
    To be fair, Google enters so many categories that some of them are bound to fail. Neither Apple, Meta or Microsoft really tries as many things as Google, so perhaps Googles issue is a lack of focus and vision. One problem that Google might have is that their criteria for success has become to high. Products has to make billions within a short time frame, but they aren't willing to make the same kind of investments in ensuring that success as a company like Apple is. The AppleTV, or even the Apple Watch would have been killed or sold off had they been Google products. Why they keep the Pixel I have no idea, maybe it is in fact part of some larger plan.
- hocuspocus 3y ago
  And the funny thing is that they bought HTC mobile, which has always been a competitive and innovative player, and I seriously doubt Taiwan based engineers are paid anywhere close to the ones in Mountain View. I've had many Nexus and two Pixels (the 5, and 8 now), with an Essential Phone in the middle, hard to beat during the last $220 fire sale. To me, Google never had a serious hardware strategy. It's improved and keeps improving though, the features getting a lot of focus (like the camera performance) are good, they finally committed to 5 and now 7 years of software updates, the Pixel lineup is sold in an increasing number of countries (Google could still do a much better job here, just sell your hardware ffs). Eventually they'll move to their own in-house SoC, and hopefully Samsung modems won't suck forever. But when I see something like the Pixel tablet sold at such ridiculous price in most countries, I truly wonder what's going on, it's like some people at Google explicitly want to fail.
freedomben 3y ago
> I guess these payments make sense if you believe you can't effectively deliver something good with that money. Or, if you believe that your product is already as good as money can buy (i.e. the brightest and highest paid engineers, with an unlimited R&D budget, and the law of the Mythical Man-Month as an upper constraint on team size), then the money is better used elsewhere to help the product. I don't know if this is what they believe, but it seems quite plausible to me.
- mynameisash 3y ago
  > you believe that your product is already as good as money can buy (i.e. the brightest and highest paid engineers, I don't think for one second that this is what Google believes, as evidenced by their collusion to suppress wages[0] [0] https://time.com/76655/google-apple-settle-wage-fixing-lawsuit/ https://time.com/76655/google-apple-settle-wage-fixing-lawsu...
- crowcroft 3y ago
  This is the point right? I do think it's fair to say that resources/$$$s aren't the thing holding back something like the Pixel Phone from being the number one smart phone, but at the end of the day it just isn't still. Therefore putting cash into this kind of thing makes sense, but the bigger question is still, why isn't Google just doing better in these non-search categories it's competing in?
  - freedomben 3y ago
    Yes agreed, if we are talking about more than just search. The Apple payment was about Search, but the Samsung payment is about more so really they are kind of apples/oranges (no pun intended!). They could and should absolutely be putting more into Android. Android has historically been about choice, but it's more and more becoming a clone of iOS. I recently had a need to record phone calls, something I did with ease ten years ago when I was battling with car insurance people, and it is damn near impossible now in 2023. Root in general is a pain in the ass. If I was terrified of the possibility of poweruser options, and wanted to be locked down and led by the nose to use my device in the exact way prescribed by the overlords, I'd be using iOS! I'm guessing they consider the market pretty saturated and aren't very motivated. Some good old fashioned competition would do wonders.
amadeuspagel 3y ago
Google has about a 120B cash on hand. If they could spend 30B to make a Pixel that everyone wants to buy, they would.
maven29 3y ago
Search result quality has been declining on Google since 2019. The web is increasingly turning into closed platform unindexable silos. These payments aren't towards being the default, it's to ensure that search itself does not get replaced by a different paradigm. You could argue that it is not a coincidence that the likes of Siri stagnated for so long, while we saw the rise of LLMs effectively resurrecting it elsewhere.
- porkbeer 3y ago
  Started declining in 2006 or so with the creative social ranking, removal of boolean search and I could go on for ages. Its my last choice these days.
  - Xiol32 3y ago
    What's your first and second choices?
- mattmcknight 3y ago
  > The web is increasingly turning into closed platform unindexable silos. So it is the quality of the open web itself that has been declining, and search is just reflection of that. LLM doesn't solve that, unless the LLM has access across the silos.
  - maven29 3y ago
    LLM training is difficult to audit, and as such good-faith barriers such as robots.txt or API endpoint scraping are all easily ignored. Even ignoring the query side, and just from an indexing perspecrive, you see a considerable advantage. LLMs (at runtime) and LLM mills (during training) have nothing to gain from restricting themselves to the open web. This is in contrast to Google, who is in the business of selling links to information and not information in itself.
gumby 3y ago
Or you could believe that the $30 B spent unlocked $150 B of revenue, more than $30 B of which could be spent on development and aggressive competition. I personally don't believe that at all (that the extra gains will be spent on development) but the possibility exists.
dkjaudyeqooe 3y ago
Obviously Google figures they're getting back more than $30B in revenue otherwise the payments make no sense.

recursive4 3y ago

My understanding is that this is an example of last mile costs increasing non-linearly i.e. Google has something like 95% of the mobile search market and profits $30b from it, and the cost to acquire that last 5% is >$30b.

sdfghswe 3y ago

I believe what's holding google back isn't lack of cash, but lack of innovation. It's full to the brim with highly paid engineers that aren't really invested in change.

rvz 3y ago
So let's see what happens when they are threatened for the first time by Microsoft and OpenAI if Google can afford to spend even more money to prevent the likes of Apple, Samsung and Mozilla from moving from Google to use Bing with GPT 4.
- smoldesu 3y ago
  Well, the big difference between ChatGPT and Google is that one of them is profitable. ChatGPT does not make money, and Bing's usage of GPT-4 is almost certainly not covered by ad revenue (even at-cost). Even from a technical perspective, I'd argue Google outpaces OpenAI. BERT beat GPT-2 and GPT-Neo, T5 and t5-flan bench well against GPT-3, and GPT-4 is so large and wasteful that it's not worth competing against. Relative to the rest of FAANG, Google honestly seems like the only company that actually knows how to use and deploy AI practically. Everyone else is trying to play catch-up with a proprietary competitor and/or pay an unsustainable sum to be king-for-a-day.
  - ffgjgf1 3y ago
    > pay an unsustainable sum to be king-for-a-day. I bet Microsoft can afford to spent unsuspecting amounts for way long than ‘a day’. At a certain point due to cheaper hardware and improved optimization GPT-N might become sustainable, by then their competitors might have issues closing the mile wide gap they have to MS/OpenAI.
jimbokun 3y ago
It becomes exponentially harder to find an innovation that moves the revenue needle perceptibly, when you are Google's size.
- swader999 3y ago
  Or one that doesn't upset the search cash cow. Open AI for example. They seem to be inhibited by moat thinking too.
crowcroft 3y ago
Yea this is the point, all roads lead back to the lack of innovation at Google. They have capital, they have free cash flow, they can attract talent, but often they completely fail to compete effectively in new categories they enter.
- mrweasel 3y ago
  To be fair, Google enters so many categories that some of them are bound to fail. Neither Apple, Meta or Microsoft really tries as many things as Google, so perhaps Googles issue is a lack of focus and vision. One problem that Google might have is that their criteria for success has become to high. Products has to make billions within a short time frame, but they aren't willing to make the same kind of investments in ensuring that success as a company like Apple is. The AppleTV, or even the Apple Watch would have been killed or sold off had they been Google products. Why they keep the Pixel I have no idea, maybe it is in fact part of some larger plan.
hocuspocus 3y ago
And the funny thing is that they bought HTC mobile, which has always been a competitive and innovative player, and I seriously doubt Taiwan based engineers are paid anywhere close to the ones in Mountain View. I've had many Nexus and two Pixels (the 5, and 8 now), with an Essential Phone in the middle, hard to beat during the last $220 fire sale. To me, Google never had a serious hardware strategy. It's improved and keeps improving though, the features getting a lot of focus (like the camera performance) are good, they finally committed to 5 and now 7 years of software updates, the Pixel lineup is sold in an increasing number of countries (Google could still do a much better job here, just sell your hardware ffs). Eventually they'll move to their own in-house SoC, and hopefully Samsung modems won't suck forever. But when I see something like the Pixel tablet sold at such ridiculous price in most countries, I truly wonder what's going on, it's like some people at Google explicitly want to fail.

freedomben 3y ago

> I guess these payments make sense if you believe you can't effectively deliver something good with that money. Or, if you believe that your product is already as good as money can buy (i.e. the brightest and highest paid engineers, with an unlimited R&D budget, and the law of the Mythical Man-Month as an upper constraint on team size), then the money is better used elsewhere to help the product. I don't know if this is what they believe, but it seems quite plausible to me.

mynameisash 3y ago
> you believe that your product is already as good as money can buy (i.e. the brightest and highest paid engineers, I don't think for one second that this is what Google believes, as evidenced by their collusion to suppress wages[0] [0] https://time.com/76655/google-apple-settle-wage-fixing-lawsuit/ https://time.com/76655/google-apple-settle-wage-fixing-lawsu...
crowcroft 3y ago
This is the point right? I do think it's fair to say that resources/$$$s aren't the thing holding back something like the Pixel Phone from being the number one smart phone, but at the end of the day it just isn't still. Therefore putting cash into this kind of thing makes sense, but the bigger question is still, why isn't Google just doing better in these non-search categories it's competing in?
- freedomben 3y ago
  Yes agreed, if we are talking about more than just search. The Apple payment was about Search, but the Samsung payment is about more so really they are kind of apples/oranges (no pun intended!). They could and should absolutely be putting more into Android. Android has historically been about choice, but it's more and more becoming a clone of iOS. I recently had a need to record phone calls, something I did with ease ten years ago when I was battling with car insurance people, and it is damn near impossible now in 2023. Root in general is a pain in the ass. If I was terrified of the possibility of poweruser options, and wanted to be locked down and led by the nose to use my device in the exact way prescribed by the overlords, I'd be using iOS! I'm guessing they consider the market pretty saturated and aren't very motivated. Some good old fashioned competition would do wonders.

amadeuspagel 3y ago

Google has about a 120B cash on hand. If they could spend 30B to make a Pixel that everyone wants to buy, they would.

maven29 3y ago

Search result quality has been declining on Google since 2019. The web is increasingly turning into closed platform unindexable silos. These payments aren't towards being the default, it's to ensure that search itself does not get replaced by a different paradigm. You could argue that it is not a coincidence that the likes of Siri stagnated for so long, while we saw the rise of LLMs effectively resurrecting it elsewhere.

porkbeer 3y ago
Started declining in 2006 or so with the creative social ranking, removal of boolean search and I could go on for ages. Its my last choice these days.
- Xiol32 3y ago
  What's your first and second choices?
mattmcknight 3y ago
> The web is increasingly turning into closed platform unindexable silos. So it is the quality of the open web itself that has been declining, and search is just reflection of that. LLM doesn't solve that, unless the LLM has access across the silos.
- maven29 3y ago
  LLM training is difficult to audit, and as such good-faith barriers such as robots.txt or API endpoint scraping are all easily ignored. Even ignoring the query side, and just from an indexing perspecrive, you see a considerable advantage. LLMs (at runtime) and LLM mills (during training) have nothing to gain from restricting themselves to the open web. This is in contrast to Google, who is in the business of selling links to information and not information in itself.

gumby 3y ago

Or you could believe that the $30 B spent unlocked $150 B of revenue, more than $30 B of which could be spent on development and aggressive competition. I personally don't believe that at all (that the extra gains will be spent on development) but the possibility exists.

dkjaudyeqooe 3y ago

Obviously Google figures they're getting back more than $30B in revenue otherwise the payments make no sense.