3 ms·

Distillation seems like a pretty robust approach. There is no good reason to think that GPT5 or 6 won’t have massive rapid efficiency gains. He’s one of the mo

Distillation seems like a pretty robust approach. There is no good reason to think that GPT5 or 6 won’t have massive rapid efficiency gains. He’s one of the most knowledgeable persons in the world on the topic. It’s like saying that the ceo of BMW isn’t citable on conversations about expected cost decreases in electric cars. Any environmental topic is susceptible to a huge amount of groupthink. The world isn’t so binary as people make it out to be. It is far from truth that LLMs=bad for environment, any more than computers=bad for the environment.

wyre 1y ago
I think its more like saying that the ceo of Tesla isn't citable on conversations about the expected advancement in electric car technology. What I mean is that I have a healthy level of skepticism with Altman. He has to constantly battle for funding. Surely, he must be knowledgeable about LLMs, but he's the CEO of the largest AI company in the world, his PR needs to give "most knowledgeable person in the world on the topic" but I think that title should go to all of the engineers and developers working on these technologies and not a capital founder. All that said, I agree that LLM's being bad for the environment is a complex topic. I think it would be more accepted if people had safety nets and could be excited for AI to take their job instead of having to be terrified, or if AI isn't just used as another tool for increasing wealth inequality.
- dr_dshiv 1y ago
  One reason to take these expected cost reductions more seriously is because it DOESN'T fit the desired narrative for OpenAI. Ideally, from a front-runner competition perspective, it would NOT be possible to create the same level of quality at < 1/10 the cost/energy within a year. But, distillation and other approaches seems to make this possible. There is a LOT of room for optimization with LLMs.
Fraterkes 1y ago
Alright, so can you actually name some numbers so that the stuff you are claiming is at least fasifiable? Altman mentions a 150x increase in efficiency, you claim that trend will continue through to gpt6. At that point these models would be 22500 as efficient as they currently are, which would mean generating a 10 hour long video would cost around the same amount of electricity as running your microwave for 15 minutes. Will you have some introspection if that doesn't come to pass?
- dr_dshiv 1y ago
  Sure, I'm game. What's your prediction?
  - dr_dshiv 1y ago
    Of course I predict this will turn into a logarithmic curve not exponential over time. But I’m always happy to bet on future outcomes with specifics. Like, that gpt5-level intelligence will be able run on contemporary MacBooks within 2-3 years of release.

wyre 1y ago

I think its more like saying that the ceo of Tesla isn't citable on conversations about the expected advancement in electric car technology. What I mean is that I have a healthy level of skepticism with Altman. He has to constantly battle for funding. Surely, he must be knowledgeable about LLMs, but he's the CEO of the largest AI company in the world, his PR needs to give "most knowledgeable person in the world on the topic" but I think that title should go to all of the engineers and developers working on these technologies and not a capital founder. All that said, I agree that LLM's being bad for the environment is a complex topic. I think it would be more accepted if people had safety nets and could be excited for AI to take their job instead of having to be terrified, or if AI isn't just used as another tool for increasing wealth inequality.

dr_dshiv 1y ago
One reason to take these expected cost reductions more seriously is because it DOESN'T fit the desired narrative for OpenAI. Ideally, from a front-runner competition perspective, it would NOT be possible to create the same level of quality at < 1/10 the cost/energy within a year. But, distillation and other approaches seems to make this possible. There is a LOT of room for optimization with LLMs.

Fraterkes 1y ago

Alright, so can you actually name some numbers so that the stuff you are claiming is at least fasifiable? Altman mentions a 150x increase in efficiency, you claim that trend will continue through to gpt6. At that point these models would be 22500 as efficient as they currently are, which would mean generating a 10 hour long video would cost around the same amount of electricity as running your microwave for 15 minutes. Will you have some introspection if that doesn't come to pass?

dr_dshiv 1y ago
Sure, I'm game. What's your prediction?
- dr_dshiv 1y ago
  Of course I predict this will turn into a logarithmic curve not exponential over time. But I’m always happy to bet on future outcomes with specifics. Like, that gpt5-level intelligence will be able run on contemporary MacBooks within 2-3 years of release.