Comment by andsoitis
10 days ago
> This will make it hard/impossible for new vendors to come into the market and only established companies will get to play, and charge, for LLMs.
I don't understand how you leap from "US govt. decides who gets to use GPT-5" to "limit new vendors from competing with OpenAI".
Can you walk me through the logic?
> the U.S. government would initially approve who gets access to its latest new release while AI companies and the administration work out a longer-term plan for regulation on the sector.
So you're a new vendor with a GPT-5.6 or Mythos class model. How do you suppose the regulations are going to work? First you need to get on the list of companies that are allowed to release models, and then you need to have a whole system for limiting access. Both are going to be hugely expensive, on top of training new models already being insanely expensive to begin with.
Thus, it's not a legal limit, but a real practical limit because it's too expensive. If only OpenAI and Anthropic and Google can afford to jump through the hoops, they've effectively gotten to "limit new vendors from competing with OpenAI".
> So you're a new vendor with a GPT-5.6 or Mythos class model. How do you suppose the regulations are going to work? First you need to get on the list of companies that are allowed to release models, and then you need to have a whole system for limiting access. Both are going to be hugely expensive, on top of training new models already being insanely expensive to begin with.
So your thesis is that someone has the resources to create a model at the level of Mythos or GPT 5.6 but bot have the resources to jump through the concomitant legal requirements?
Surely that’s unlikely?
In a perfectly spherical cow sort of way, everything is fungible and if you can get the resources to train a Mythos-grade model then you'll also have the resources to jump through all the legal hoops. Regulatory capture is a tale almost as old as capitalism though. History bears this out. Adding barriers to entry further entrenches the incumbents.
These frontier models are now good enough that they can assist heavily with new optimizations for future models, including, code them. Restricting their usage to a few companies takes away that advantage away from other companies, thereby, limiting new vendors from competing with OpenAI.
> Restricting their usage to a few companies takes away that advantage away from other companies, thereby, limiting new vendors from competing with OpenAI.
wait, are you saying a competitor needs access to an OpenAI model in order to build a competing model?
Nothing new in that. Everyone pays to someone or the other to make their own product/life better. Most of the times these products do not compete, sometimes, they do.
How many times you believe duckduckgo would have google'd stuff just to create a competitor to Google itself? I believe thousands of times.. could be more.
OpenAI cannot claim the code their models produce as their own since their own models used codes from public internet during training to produce new code.
2 replies →
I agree with your point, but to play devils advocate: doesn't a competitor arguably need access to these beyond-frontier models to even become an effective competitor to OpenAI and Anthropic?
1 reply →
Do you have any sources for the claim that LLMs meaningfully help in the production of LLMs?
Read Anthropic's own blog post released sometime around last week.
5 replies →