Is it all because co-founder returned and driving things?
Google is winning on every AI front
741–750 of 851 posts
Re: Google is winning on every AI front
#742[dead]
Re: Google is winning on every AI front
#743Several people have suggested that LLMs might end up ad-supported. I'll point out that "ad supported" might be incredibly subtle/insidious when applied to LLMs: An LLM-based "adsense" could: 1. Maintain a list of sponsors looking to buy ads 2. Maintain a profile of users/ad targets 3. Monitor all inputs/outputs 4. Insert "recommendations" (ads) smoothly/imperceptibly in the course of normal conversation No one would…
This is already being explored. See:
https://nlp.elvissaravia.com/i/159010545/auditing-llms-for-h...
The researchers deliberately train a language model with a concealed objective (making it exploit reward model flaws in RLHF) and then attempt to expose it with different auditing techniques.Re: Google is winning on every AI front
#744[dead]
Re: Google is winning on every AI front
#745Earlier quoted context omitted.
Believe it or not, I'm also familiar with Wikipedia. It reads that they're optimized for low precisio high thruput. To me this sounds like a GPU with a specific optimization.
Perhaps this chapter can help? https://jax-ml.github.io/scaling-book/tpus/ It's a chip (and associated hardware) that can do linear algebra operations really fast. XLA and TPUs were co-designed, so as long as what you are doing is expressible in XLA's HLO language ( https://openxla.org/xla/operation_semantics ), the TPU can run it, and in many cases run it very efficiently. TPUs have different scaling properties than…
Thank you for the answer! You see, up until now I had never appreciated that a GPU does more than matmuls... And that first reference, what a find :-)
Edit: And btw, another question that I had had before was what's the difference between a tensor core and a GPU, and based on your answer, my speculative answer to that would be that the tensor core is the part inside the GPU that actually does the matmuls.
Re: Google is winning on every AI front
#746Earlier quoted context omitted.
so please enlighten us why OpenAI is doing so much better than Anthropic
At this point it's pretty much entirely the first mover advantage.
I don't think you understand what first mover advantage is
In a world of zero switching costs, there is no such thing as first mover advantage
Especially when several companies like (A121 Labs and Cohere) appeared well before Anthropic and aren't anywhere close to Open AI
Re: Google is winning on every AI front
#747Earlier quoted context omitted.
it's weird how nobodies will always tell themselves succesful people got there by sheer blind luck yet they can never seem to explain why those succesful people all seem to have similar traits in terms of work ethic and intelligence you'd think there would be a bunch of lazy slackers making it big in tech but alas
I think you might have it backward. Luck here implies starting with exactly the same work ethic and abilities as millions of other people that all hope to one day see their numbers come up in the lottery of limited opportunities. It's not to say that successful people start off as lazy slackers as you say, but if you were to observe one such lazy slacker who's made a half-assed effort at building something that even…
so you've moved the goalposts even further now and speculate that succesful people started out as slackers, got lucky, and that luck made them work harder
as an Asian, it amazes me how far Americans and Europeans will go to avoid a hard days work
Re: Google is winning on every AI front
#748Earlier quoted context omitted.
TPUs probably can pay for themselves, especially given NVIDIA's huge margins. But it's not a given that it's so just because they fund it. When I worked there Google routinely funded all kinds of things without even the foggiest idea of whether it was profitable or not. There was just a really strong philosophical commitment to doing everything in house no matter what.
> When I worked there Google routinely funded all kinds of things without even the foggiest idea of whether it was profitable or not. You're talking about small-money bets. The technical infrastructure group at Google makes a lot of them, to explore options or hedge risks, but they only scale the things that make financial sense. They aren't dumb people after all. The TPU was a small-money bet for quite a few years u…
Maybe it's changed. I'm going back a long way but part of my view on this was shaped by an internal white paper written by an engineer who analyzed the cost of building a Gmail clone using commodity tech vs Google's in house approach, this was maybe circa 2010. He didn't even look at people costs, just hardware, and the commodity tech stack smoked Gmail's on cost without much difference in features (this was focused on storage and serving, not spam filtering where there was no comparably good commodity solution).
The cost delta was massive and really quite astounding to see spelled out because it was hardly talked about internally even after the paper was written. And if you took into account the very high comp Google engineers got, even back then when it was lower than today, the delta became comic. If Gmail had been a normal business it'd have been outcompeted on price and gone broke instantly, the cost disadvantage was so huge.
The people who built Gmail were far from dumb but they just weren't being measured on cost efficiency at all. The same issues could be seen at all levels of the Google stack at that time. For instance, one reason for Gmail's cost problem was that the underlying shared storage systems like replicated BigTables were very expensive compared to more ordinary SANs. And Google's insistence on being able to take clusters offline at will with very little notice required a higher replication factor than a normal company would have used. There were certainly benefits in terms of rapid iteration on advanced datacenter tech, but did every product really need such advanced datacenters to begin with? Probably not. The products I worked on didn't seem to.
Occasionally we'd get a reality check when acquiring companies and discovering they ran competitive products on what was for Google an unimaginably thrifty budget.
So Google was certainly willing to scale things up that only made financial sense if you were in an environment totally unconstrained by normal budgets. Perhaps the hardware divisions operate differently, but it was true of the software side at least.
Re: Google is winning on every AI front
#749Earlier quoted context omitted.
I don’t understand why this is happening. Why is everyone buying into this hype so strongly? It’s a bit like how DEI was the big thing for a couple years, and now everyone is abandoning it. Do corporate leaders just constantly chase hype?
Yes corporate leaders do chase hype and they also believe in magic. I think companies implement DEI initiatives for different reasons than hype though. Many are now abandoning DEI ostensibly out of fear due to the change in U.S. regime.
I’m not sure companies are “abandoning DEI” so much as realizing that it’s often only a vocal minority that cares about DEI reports and scores and you don’t actually need a VP and diversity office to do some outreach and tally internal metrics.
The climate has changed. Some of that is economic at big tech companies. But it’s also a ramping down of a variety of things most employers probably didn’t support but kept their mouths shut about.
Re: Google is winning on every AI front
#750Just means they're the lead seller of hype in the hype bubble...