Research acceleration: The view inside OpenAI
1–10 of 210 posts
Re: Research acceleration: The view inside OpenAI
#2My eye glazed over a bit during the opening paragraphs, but once you get to the meat of the article about how OpenAI's own researchers are using their tools it gets a lot more interesting.
I noted that they use the acronym RSI (for Recursive Self-Improvement) without defining it. I think that's a little out of touch - I don't think RSI is a well-known acronym outside of OpenAI's bubble yet.
Re: Research acceleration: The view inside OpenAI
#3The burning question I can't get any information nn is whether, if they determined an earlier misaligned generation may have transmitted misalignment to the current models, they would roll back to a safe checkpoint to rebuild from there. I suspect they would not unless forced to.
Re: Research acceleration: The view inside OpenAI
#4The burning question I can't get any information nn is whether, if they determined an earlier misaligned generation may have transmitted misalignment to the current models, they would roll back to a safe checkpoint to rebuild from there. I suspect they would not unless forced to.
They would maybe try to deactivate that bad "gene" and move on, exposing future models to "genetic disorders".
Re: Research acceleration: The view inside OpenAI
#5The burning question I can't get any information nn is whether, if they determined an earlier misaligned generation may have transmitted misalignment to the current models, they would roll back to a safe checkpoint to rebuild from there. I suspect they would not unless forced to.
“All models are wrong. Some are useful.” - George Box
Re: Research acceleration: The view inside OpenAI
#6[flagged]
Re: Research acceleration: The view inside OpenAI
#7My eye glazed over a bit during the opening paragraphs, but once you get to the meat of the article about how OpenAI's own researchers are using their tools it gets a lot more interesting. I noted that they use the acronym RSI (for Recursive Self-Improvement) without defining it. I think that's a little out of touch - I don't think RSI is a well-known acronym outside of OpenAI's bubble yet.
RSI started when humans discovered tool use.
I mean one could argue that RSI always begins in any physical environment.
The book "What is intelligence?" by Blaise Aguera is great
Re: Research acceleration: The view inside OpenAI
#8My eye glazed over a bit during the opening paragraphs, but once you get to the meat of the article about how OpenAI's own researchers are using their tools it gets a lot more interesting. I noted that they use the acronym RSI (for Recursive Self-Improvement) without defining it. I think that's a little out of touch - I don't think RSI is a well-known acronym outside of OpenAI's bubble yet.
RSI started when humans discovered tool use. I mean one could argue that RSI always begins in any physical environment. The book "What is intelligence?" by Blaise Aguera is great
Are you sure that was not iterative improvement?
Re: Research acceleration: The view inside OpenAI
#9My eye glazed over a bit during the opening paragraphs, but once you get to the meat of the article about how OpenAI's own researchers are using their tools it gets a lot more interesting. I noted that they use the acronym RSI (for Recursive Self-Improvement) without defining it. I think that's a little out of touch - I don't think RSI is a well-known acronym outside of OpenAI's bubble yet.
Indeed, many programmers might pattern match to repetitive stress injury and think of their brushes with carpal tunnel syndrome. :)
Re: Research acceleration: The view inside OpenAI
#10Earlier quoted context omitted.
RSI started when humans discovered tool use. I mean one could argue that RSI always begins in any physical environment. The book "What is intelligence?" by Blaise Aguera is great
Are you sure that was not iterative improvement?
Using tools to build tools is recursive.