And, as I understand it, Anthropic hasn’t committed as much spending to building out new data centers, and has setup their operations to be GPU agnostic, so they can keep flexibility between NVIDIA GPUs, Google TPUs, and Amazon Trainium, and play the data center pricing game. Anthropic is better positioned to survive an AI winter (and I believe it’s coming soon).
Altman’s been trying to put on a big show with huge numbers and all the huge NVIDIA deals are part of it. Anthropic actually seems to be trying to build good products, as far as that’s even possible when it comes to AI. I think it says a lot that most people had heard of Sam Altman a few years ago already, whereas Dario Amodei isn’t really a household name, even today.
Their reputation is also a bit in the toilet, because people hear “AI” and think of ChatGPT.
So “man hospitalised after AI suggested he put glue on pizza for tackiness” would have people think he was using it, when he might well have been using a different LLM.
The field is moving so fast that things can change quickly, but the American labs are so caught up in saddling their models with safety overhead that the recent Chinese models are very close in practical use to the flagship American models if not pulling ahead (Sora vs Seedance 2).
I don’t really need to solve Erdős problems in my day to day. Outside of increasingly edge case eval competition, I’m not sure what OpenAI brings that literally everyone else isn’t also capable of providing (and more).
I’d maybe invest in Anthropic for an IPO if they turned around their own saddling of models and played nicer with open platforms, but if Claude is just going to get more and more anxious due to excessive red teaming and CC fall further and further behind stuff like Hermes Agent, they too are going to fall by the wayside as open models become the dominant inference for open infrastructure.
Distillation isn’t stealing the original model, though. It just uses the models to make synthetic training data to train their own thing. They aren’t stealing the model itself.
It also doesn’t seem like as big a deal as Anthropic and Open AI make it look, IMO. Them treating it like a national security issue where the company gets its models stolen from under its nose just comes across like a media company claiming that every download is a copy they would otherwise have sold at full price, and thus they have accrued trillions of dollars in damages.
I could, in theory, take a bunch of google Gemini outputs, and train a GPT-2 model on them. That doesn’t mean that I’ve recreated Gemini, nor does it mean that i’ve stolen it from Google, either.
To top it all off, it’s not like their services were abused. The companies were presumably paid appropriately for the usage.
I don’t understand how anyone can keep a straight face when they hear an AI company crying about another AI company “stealing” from them while they go before lawmakers and argue that if they weren’t allowed to steal stuff, AI wouldn’t exist… I immediately picture the Always Sunny meme “oh, did someone get addicted to crack” Crying motions.
China won the game. Their models are cheap and available and free weights.
Openai will never make any money. They realized it’s a high time to sell so. Wouldnt give a dime
They’re also behind Anthropic when it comes to expensive frontier models.
I wouldn’t buy their stock even if I was looking to invest in AI.
And, as I understand it, Anthropic hasn’t committed as much spending to building out new data centers, and has setup their operations to be GPU agnostic, so they can keep flexibility between NVIDIA GPUs, Google TPUs, and Amazon Trainium, and play the data center pricing game. Anthropic is better positioned to survive an AI winter (and I believe it’s coming soon).
Altman’s been trying to put on a big show with huge numbers and all the huge NVIDIA deals are part of it. Anthropic actually seems to be trying to build good products, as far as that’s even possible when it comes to AI. I think it says a lot that most people had heard of Sam Altman a few years ago already, whereas Dario Amodei isn’t really a household name, even today.
Their reputation is also a bit in the toilet, because people hear “AI” and think of ChatGPT.
So “man hospitalised after AI suggested he put glue on pizza for tackiness” would have people think he was using it, when he might well have been using a different LLM.
It’s true.
The field is moving so fast that things can change quickly, but the American labs are so caught up in saddling their models with safety overhead that the recent Chinese models are very close in practical use to the flagship American models if not pulling ahead (Sora vs Seedance 2).
I don’t really need to solve Erdős problems in my day to day. Outside of increasingly edge case eval competition, I’m not sure what OpenAI brings that literally everyone else isn’t also capable of providing (and more).
I’d maybe invest in Anthropic for an IPO if they turned around their own saddling of models and played nicer with open platforms, but if Claude is just going to get more and more anxious due to excessive red teaming and CC fall further and further behind stuff like Hermes Agent, they too are going to fall by the wayside as open models become the dominant inference for open infrastructure.
Their models may also be based on US models.
It’ll be hard for derivative models to innovate if their host organism has died.
Distillation isn’t stealing the original model, though. It just uses the models to make synthetic training data to train their own thing. They aren’t stealing the model itself.
Plus, a lot of companies do it. Anthropic’s Claude was calling itself DeepSeek for a while.
It also doesn’t seem like as big a deal as Anthropic and Open AI make it look, IMO. Them treating it like a national security issue where the company gets its models stolen from under its nose just comes across like a media company claiming that every download is a copy they would otherwise have sold at full price, and thus they have accrued trillions of dollars in damages.
I could, in theory, take a bunch of google Gemini outputs, and train a GPT-2 model on them. That doesn’t mean that I’ve recreated Gemini, nor does it mean that i’ve stolen it from Google, either.
To top it all off, it’s not like their services were abused. The companies were presumably paid appropriately for the usage.
I don’t understand how anyone can keep a straight face when they hear an AI company crying about another AI company “stealing” from them while they go before lawmakers and argue that if they weren’t allowed to steal stuff, AI wouldn’t exist… I immediately picture the Always Sunny meme “oh, did someone get addicted to crack” Crying motions.
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod
Removed by mod