alt Hacker NewsEven if Anthropic invents a godlike frontier model that can do anything, it feels like they won't have a moat for very long before competitors like DeepSeek offer something almost as good for a fraction of the price.
Even if Anthropic invents a godlike frontier model that can do anything, it feels like they won't have a moat for very long before competitors like DeepSeek offer something almost as good for a fraction of the price.
If AGI is created, then it will be used by someone else to create a cheaper AGI, making the finances of the person who borrowed trillions to create it in question.
An extremely hypothetical experiment but what if this is some sort of MAD (Mutually Assured Destruction) and in our timeline, we have accepted it with AI.
Deepseek can do MAD on Anthropic and release open weights model (which means that they might not earn the most or basically burn millions/billions to then give us all essentially the same thing and we would all be on equal footing, oh they also released their inference engine right now.)
This might explain why the stock market got into an bubble and is fearing itself right now.
Mutually Assured Commoditization. (Credit where it's due I pasted my comment to Chatgpt and it gave me this term ironically and it seems to fit well to the comment that I am writing, I might write a blog about it someday.)