Hacker News new | past | comments | ask | show | jobs | submit
Analemma_ | on: Changes at Google DeepMind: Demis Hassabis from CEO to Chair, Jeff Dean departs
For at least a year now the standard reflexive reply to “Google seems way behind OAI and Anthropic” has been “it’ll be ok, they’ve got Dean and Hassabis.” And now they don’t. What reason is there to be bullish about Google now?
losvedir | parent | next
I still think Google is the only one who has a shot of coming out of this on top. Open AI and Anthropic's whole existence is predicated on some sort of moat, which I don't really see them having long term. They've got a bit of a headstart, but that's it.

Conversely, AI is just a means to an end for Google - they don't need for their model to be the one to succeed. But, in contrast to the other major company in their position Apple, they do have a model, so they're not totally beholden to another for AI (like Apple is using Gemini!).

But beyond that, for training the model they have YouTube, and of course they have their crawler and index, and the billions of users.

I think HN skews coding agent focused, but that's not really a market for Google. I expect we will have coding specific models in the future, but Google wants a more general intelligence, to handle search queries, be able to connect email to chat to calendar and tasks, and so on. I don't find Gemini that much worse than the other big models for non-coding things.

nostrademons | root | parent | next
The leaked internal Google memo "We have no moat" [1] was really prescient. It foresaw all of these proprietary models losing out to open-weight models.

Entirely possible that none of the big incumbents ends up winning, economically.

[1] https://simonwillison.net/2023/May/4/no-moat/

chrisXOXO | root | parent
I would like a reassessment of that. This was 2023. I imagine the only ones filling the moat is Chinese labs, but there are other significant obststacles for even Chinese models. Even if they would become as good as their American counterparts, the distribution and inference is mostly in the ballpark of Anthropic and OpenAI. But I don’t know anything, so I would love to hear other opinions.
alfalfasprout | root | parent
If anything, it's becoming more true. In the last month you already have Kimi K3 reaching competitive frontier performance (in practice, a bit behind Fable/Sol but close enough). Sure, it's a chinese model but the reality is it puts dramatic pricing pressure on the main players. And these models are now getting good enough to actually help accelerate further AI model research which would suggest the gap might close even more in the near future.
treis | root | parent | next
I hardly use Google search anymore in favor of ChatGPT. It's surely an existential threat. Google has a lot of layers of defense in Android, Chrome, Gmail/Accounts, and a stranglehold on internet ads. But at the center of that is search and search is being replaced by LLMs.
ttul | root | parent | next
We are seeing an exponential increase in the share of site traffic to our B2B company originating from ChatGPT and other models. Google Search is cooked.
cogman10 | root | parent | next
IMO, this is largely because google made their search next to worthless. They very apparently prioritized making money over returning good relevant results. The decline in quality is so pronounced that I've heard more than a few people talk about using other search engines besides google.

I don't think it's extreme to say that Bing is a better search engine than Google at this point and Bing isn't great.

paganel | root | parent
> I hardly use Google search anymore in favor of ChatGPT.

Same here, and I'm generally very skeptical when it comes to AI. It's just that ChatGPT can provide better/more accurate answers, Google seems to have lost the plot. I still use google search when it comes to appending "wiki/wikipedia" to it, i.e. when I use google search just as a redirect to a wikipedia page.

B-Con | root | parent | next
I read Google's business incentive as "in it to not lose". AI is a new consumer endpoint and Google currently captures a lot of the endpoints.

> I think HN skews coding agent focused

100% this. Coding agents are an interesting test ground, but the people who care about them are a fairly small bubble.

I think they carry some of the overall AI weight because of the "what if you can vibe code your entire business" moonshot, but that's still orders of magnitude away and who knows if today's coding agents will actually be a stepping stone to that. If we ever get there it will likely be with entirely new domain languages.

dolmen | root | parent | next
What about revenue if people aren't using the search engine directly anymore, also consuming content through chats (not made by them), not on the web where ads are displayed?
alfalfasprout | root | parent | next
On the flipside, Apple has realized AI models are quickly becoming a commodity. By not blowing hundreds of billions in speculative capex, they get to focus on the bits that likely have higher ROI: the touchpoints with AI and how it's used.
xnx | parent | next
Data centers, TPUs, customers, data streams, more money than god, most mature crawler system
conradkay | root | parent | next
Are any of those advantages getting stronger over time? I guess TPUs but Google is selling several gigawatts to Anthropic
solenoid0937 | root | parent | next
Google owns 14% of Anthropic. If Anthropic gets to superintelligence, Google wins as well. Google doesn't really need a frontier lab of its own.
ai_fry_ur_brain | root | parent
The money is in selling compute not using it. Anyone who thinks otherwise is a mark.
throwaway667555 | root | parent | next
Crawler and Google Scholar and Google Books.
re-thc | root | parent
> Data centers, TPUs

Those 2 are definitely NOT Google's strong points. Maybe by means of marketing.

Oracle, Amazon, Microsoft, Equinix and many more are in the data center race.

As for TPUs... Broadcom, Mediatek and all the other partners you hear less about are likely more important. Google just has the flashy media outreach.

basch | parent | next
As much as everyone wants to pay accomplished celebrities, all of these companies have young nameless geniuses that are about to make one for themselves. A guard passing torch can be opportunity.

Now, whether Google is the right environment to nurture, that’s its own quandary.

htrp | root | parent
>all of these companies have young nameless geniuses that are about to make one for themselves

They all leave to start new companies. Everyone on the Attention is all you need paper is at a startup.

sva_ | parent | next
I keep hearing this, but Hassabis did not leave Google/Alphabet. He merely changed from an administrative position to a more technical one.
tuesdaynight | parent | next
It just depends on what are the most upvoted comments in HN. If they are bearish, you can be sure that you should be bullish in your investments. Meta was supposed to be broken as a business already, OpenAI and Anthropic would be failures as well.
tim333 | parent | next
Maybe they'll replace Hassabis with someone more focussed on beating OAI and Anthropic at chatbots? He always seemed a bit more into science, protein folding, new drugs and the like.
aroman | parent | next
Revenue, data, and stamina.
CamperBob2 | parent
Couple of things come to mind. One is that Web sites tend to actively fight AI companies' crawlers while actively courting Google's crawler.

Another is that they hold the key Transformer architecture patent. If it is still relevant (which I'm not personally clueful about) and if they start enforcing it, then we may see a reprise of the situation where Microsoft made money for years every time an Android phone was sold. Regardless of that particular patent, it's probably safe to say they'll be better-positioned than anyone else if AI companies start lobbing patent nukes at each other.

A third factor that shouldn't be discounted is that Google has access to warehouses of training data that other companies don't. Google Books alone is an Alexandria-scale archive that the courts forced them to keep to themselves. Those restrictive copyright decisions may turn out to be a blessing in disguise for Google because no one else will have been able to scrape the data.

WarmWash | root | parent
Google's weakness (well one of) is its total lack of cohesion. If the Google Books team could, they would sell access to that data in a heartbeat to boost their metrics.
vkou | root | parent
Which would be disastrous to the conglomorate as a whole. It's good for Google that they can't.