Sad to see everyone so focused on compute expense during this massive breakthrough. GPT-2 originally cost $50k to train, but now can be trained for ~$150.
The key part is that scaling test-time compute will likely be a key to achieving AGI/ASI. Costs will definitely come down as is evidenced by precedents, Moore’s law, o3-mini being cheaper than o1 with improved performance, etc.
Those are the (subsidized) prices that end clients are paying for the service so that's not something that is representative of what the actual inference costs are. Somebody still needs to pay that (actual) price in the end. For inference, as well as for training, you need actual (NVidia) hardware and that hardware didn't become any cheaper. OTOH models are only becoming increasingly more complex and bigger and with more and more demand I don't see those costs exactly dropping down.
Actual inference costs without considering subsidies and loss leaders are going down, due to algorithmic improvements, hardware improvements, and quantized/smaller models getting the same performance as larger ones. Companies are making huge breakthroughs making chips specifically for LLM inference
I think the question everyone has in their minds isn't "when will AGI get here" or even "how soon will it get here" — it's "how soon will AGI get so cheap that everyone will get their hands on it"
that's why everyone's thinking about compute expense.
but I guess in terms of a "lifetime expense of a person" even someone who costs $10/hr isn't actually all that cheap, considering what it takes to grow a human into a fully functioning person that's able to just do stuff
