To be clear, I'm not bullish on the AI companies themselves. I think Chinese companies will continue to distill these models and once they achieve RSI (which is imminent), intelligence will be abundant and cheap. The real profit will be in the other layers of the AI ecosystem: electricity generation, chip fabrication (memory, storage, CPU/GPU), infrastructure (datacentres), and probably most notably, the people and companies which produce compelling software and services using this technology.
Also, please ask Anthropic why their sec-1 filling is missing financial data after that quarter they made "profit". Most analysts belive that their profit was financial juggling and things went downhill really quickly after June.
Longcat is probably the worst of all Chinese models, but it's dirt cheap, and it works well enough. So my usage of Anthropic's models (and overall amount of money spent on AI) has decreased drastically.
So is Anthropic at their current valuation; they are gambling that they will become a multi trillion dollar company soon. That is speculation, as they are not profitable yet and everyone just hope that they will succeed.
So there's actually a coherent argument that Anthropic isn't hemorrhaging money quite as badly as pets.com. But yes, I'm glad I'm not the only one who thought omitting Anthropic's loss this was a blazingly odd way to open an argument.
Profit statement would sound nicer. But ofc, "Amazon was unprofitable when it IPOd too" and $28M losses on IPO year is comparable (not) after you adjust from inflation.
Imminent according to which sources which are not financially involved with AI companies?
Nuclear fusion is also imminent.
How much juice is left in that tank after 5 years of intense focus on optimizing LLMs is another matter. If we are on top of a local architecture optimum, speeding up the hill-climbing will deliver underwhelming results until someone discovers the next hill.