One thing I really admire about isomorphic and DeepMind in general is that they really try and encode human ingenuity into their models. The only reason that AlphaFold worked wasn't because they had better hardware or more money to train models - it's because they used every bit of information they had access to.
The same goes for AlphaGenome - it only takes a few hours to train the model, but it's freaking amazing what it can predict.
How about focus on keeping up with anthropic and openai. Less talk and more work. As a google shareholder I am really disappointed by their performance lately.
This is short-sighted IMO, I bet money that once the dust settles with the current LLM-boom that Google will still be around and pioneering in AI across the board. They aren't locked in to Nvidia, they have world class research, they aren't bogged down with VC returns or upcoming IPOs.
I agree with you for the most part. However they have been raising a lot of cash, and to do so they took on a ton of debt.
It's a totally reasonable investment as long as they are within shooting distance of the market leaders, but the more they lag behind (i.e. the longer it takes for them to release their next model on par with industry leaders) the more hurt they're going to be in.
If they can't materialize returns off the tens of billions of dollars they just raised then it will come back to bite them hard.
Luckily they have a ton of options (such as selling compute) but it's still not a great look and the market is going to hammer them given the opportunity.
Strongly disagree. Google (and Apple) need only be near the frontier. Their incredible distribution and integration will propel them far beyond the next marginal billion dollar benchmarkmaxxing.
I'm saying that. But they're falling further behind day by day. Gemini 3.5 Pro is allegedly months behind. If they can't keep up they're going to be left in the dust.
I see that being about as likely as IBM doing the same.
I'm too young to have seen the arc of Xerox PARC first hand, so to me stuff like the innovators dilemma made sense, but didn't feel particularly visceral.
AI Studio (or whatever their internal name is) is the first time in my own lifetime witnessing how real deep it really cuts.
Google realizes GCP is too slow and overwrought to get mindshare vs nimble OpenAI and Ant, spins up a new product org as a work around, and that product org ends up moving obviously much faster than Vertex, but also with a sort of malaise (relative to the technology they're supposed to be selling) that makes it clear to me that Google actually cannot move like a startup anymore.
That might seem super obvious to most people, but I grew up with Google being the startup. I knew they grew up to be a mega cap, but I guess I always assumed the bones of a startup were still in there.
There's no bones. It actually feels like a mini-identity crisis for myself to realize there is no startup left in Google: what other invariants I assumed about people and organizations are just plain wrong?
It's a pretty bad look when all they do is shed researchers. Whether to other labs, or to people launching their own things. "I work at Google" doesn't have the ring it had 10-20 years ago either, something that OAI and especially Anthropic enjoy nowadays.
Deepmind is looking less and less like a place of prestige, and more and more like a smash-and-grab.
I actually think a reasonable strategy of Google is to focus on biotech, where the demand for compute is infinite, and focus on cost effective solutions with really good search for other verticals (eg legal, accounting, etc), and offer loss leaders in coding to undercut Anthropic/openai, with “cheap just behind the frontier” models.
They are not really doing this, but it makes sense to me.
Yeah, I would be curious to know what kind of returns they’re seeing from Isomorphic labs given AlphaFold itself is publicly available. From what I’ve read, the deals they’ve reached with the major Pharma companies could result in billions in returns, but they come as royalties or are tied to performance metrics. It’s going to be hard to know how valuable the investments made in this space will be for probably another decade.
Google is the AOL of the old days. They can’t compete with OpenAI and Claude. They are a search company and AI is eating search.
Look at Stackoverlow for a hint where Google is heading to.
I agree, it's the old academia play. When they can't innovate on the model work, they make a lot of noise about AI safety, gatekeeping, humanity extinction etc.
Anthropic is the company that makes the most noise about AI safety, in all the good and bad ways. While being THE frontier model player to watch.
They got way less resources than OpenAI. They don't have the infinite compute or the biggest userbase. And they still punch way above their weight with just about every release.
One thing I really admire about isomorphic and DeepMind in general is that they really try and encode human ingenuity into their models. The only reason that AlphaFold worked wasn't because they had better hardware or more money to train models - it's because they used every bit of information they had access to.
The same goes for AlphaGenome - it only takes a few hours to train the model, but it's freaking amazing what it can predict.
How about focus on keeping up with anthropic and openai. Less talk and more work. As a google shareholder I am really disappointed by their performance lately.
This is short-sighted IMO, I bet money that once the dust settles with the current LLM-boom that Google will still be around and pioneering in AI across the board. They aren't locked in to Nvidia, they have world class research, they aren't bogged down with VC returns or upcoming IPOs.
I agree with you for the most part. However they have been raising a lot of cash, and to do so they took on a ton of debt.
It's a totally reasonable investment as long as they are within shooting distance of the market leaders, but the more they lag behind (i.e. the longer it takes for them to release their next model on par with industry leaders) the more hurt they're going to be in.
If they can't materialize returns off the tens of billions of dollars they just raised then it will come back to bite them hard.
Luckily they have a ton of options (such as selling compute) but it's still not a great look and the market is going to hammer them given the opportunity.
Strongly disagree. Google (and Apple) need only be near the frontier. Their incredible distribution and integration will propel them far beyond the next marginal billion dollar benchmarkmaxxing.
I'm saying that. But they're falling further behind day by day. Gemini 3.5 Pro is allegedly months behind. If they can't keep up they're going to be left in the dust.
Their niche is that you get the quality of a Chinese model for the price of an American model.
I see that being about as likely as IBM doing the same.
I'm too young to have seen the arc of Xerox PARC first hand, so to me stuff like the innovators dilemma made sense, but didn't feel particularly visceral.
AI Studio (or whatever their internal name is) is the first time in my own lifetime witnessing how real deep it really cuts.
Google realizes GCP is too slow and overwrought to get mindshare vs nimble OpenAI and Ant, spins up a new product org as a work around, and that product org ends up moving obviously much faster than Vertex, but also with a sort of malaise (relative to the technology they're supposed to be selling) that makes it clear to me that Google actually cannot move like a startup anymore.
That might seem super obvious to most people, but I grew up with Google being the startup. I knew they grew up to be a mega cap, but I guess I always assumed the bones of a startup were still in there.
There's no bones. It actually feels like a mini-identity crisis for myself to realize there is no startup left in Google: what other invariants I assumed about people and organizations are just plain wrong?
> they have world class research
It's a pretty bad look when all they do is shed researchers. Whether to other labs, or to people launching their own things. "I work at Google" doesn't have the ring it had 10-20 years ago either, something that OAI and especially Anthropic enjoy nowadays.
Deepmind is looking less and less like a place of prestige, and more and more like a smash-and-grab.
Google has always shed researchers but they still have way more of them than everybody else.
I actually think a reasonable strategy of Google is to focus on biotech, where the demand for compute is infinite, and focus on cost effective solutions with really good search for other verticals (eg legal, accounting, etc), and offer loss leaders in coding to undercut Anthropic/openai, with “cheap just behind the frontier” models.
They are not really doing this, but it makes sense to me.
Yeah, I would be curious to know what kind of returns they’re seeing from Isomorphic labs given AlphaFold itself is publicly available. From what I’ve read, the deals they’ve reached with the major Pharma companies could result in billions in returns, but they come as royalties or are tied to performance metrics. It’s going to be hard to know how valuable the investments made in this space will be for probably another decade.
As a biological human I am glad that Google is applying AI to biodefense.
Google is the AOL of the old days. They can’t compete with OpenAI and Claude. They are a search company and AI is eating search. Look at Stackoverlow for a hint where Google is heading to.
> Less talk and more work.
> As a google shareholder
How about you do some work? The world needs less armchair bosses and more do-ers.
Stock price has almost doubled in the past year and you still aren’t satisfied? This delusional market needs a reckoning.
1Y 86%
Microsoft is IBM. Google is Microsoft. OpenAI is Google. Apple is Apple.
I agree, it's the old academia play. When they can't innovate on the model work, they make a lot of noise about AI safety, gatekeeping, humanity extinction etc.
Anthropic is the company that makes the most noise about AI safety, in all the good and bad ways. While being THE frontier model player to watch.
They got way less resources than OpenAI. They don't have the infinite compute or the biggest userbase. And they still punch way above their weight with just about every release.