> So it made me wonder. Is Brainf*ck the ultimate test for AGI? Absolutely not. Id bet a lot of money this could be solved with a decent amount of RL compute. None of the stated problems are actually issues with LLMs…
Yes we do. If you worked at Google you know moma. Our moma is an internal version of chat. It is very good.
Also work at OpenAI. Every tender offer has made full payouts to previous employees. Sorry to ruin your witch hunt..
I work for openai. o4-mini gets much closer (but I'm pretty sure it fumbles at the last moment): https://chatgpt.com/share/680031fb-2bd0-8013-87ac-941fa91cea... We're pretty bad at model naming and communicating…
You're missing the fact that requests are batched. It's 70 tokens per second for you, but also for 10s-100s of other paying customers at the same time.
You wouldn't do that to this model. It finds its own mistakes and corrects them as it is thinking through things.
I'm James Betker. Of course architecture matters in this regard lol. Comparing a CNN to a transformer is like comparing two children brought up in the same household but one has a severe disability. What I meant in this…
As an employee of OpenAI: fuck you and your condescending conclusions about my peers and my motivations.
I don't think the plan is an occasionally rocket launch long term. I think the plan is to launch rockets as fast as humanly possible.
Its cool that this is starting to approach real time video territory (30 images per second, this claims close to 1 image/sec).
@dooraven - I also work in ML (including recently working at Google) and I agree with @whimsicalism. You seem to be under the mistaken belief that: 1. Google has competent high-level organization that effectively sets…
You can make almost anything work in DL if you try hard enough, that doesn't mean it is the correct thing to do. Convolutions have inductive biases which are the cause of many of the problems associated with deep…
I theorize (but cannot prove) that the processes underpinning creativity in the human mind are exactly the same statistical processes that ML models use. Think about it: you live your life. You experience things. You…
AGI consists of a set of problems: 1. Finding an algorithm which is capable of learning anything. 2. Building the computers that can run said algorithm. 3. Collecting, sorting and filtering the data that the algorithm…
This is wrong IMO - the fed has NO tools in its hands. Its last tool was inflation rates, and it burned that during the great recession. It is now caught in a pickle: raise interest rates, causing a mass exodus of…
There are other reasons to farm indoors (or in vertical farms) than electricity and water too. Pests are one such reason. They are extremely difficult to control outdoors, but far simpler to do so inside. Both…
> So it made me wonder. Is Brainf*ck the ultimate test for AGI? Absolutely not. Id bet a lot of money this could be solved with a decent amount of RL compute. None of the stated problems are actually issues with LLMs…
Yes we do. If you worked at Google you know moma. Our moma is an internal version of chat. It is very good.
Also work at OpenAI. Every tender offer has made full payouts to previous employees. Sorry to ruin your witch hunt..
I work for openai. o4-mini gets much closer (but I'm pretty sure it fumbles at the last moment): https://chatgpt.com/share/680031fb-2bd0-8013-87ac-941fa91cea... We're pretty bad at model naming and communicating…
You're missing the fact that requests are batched. It's 70 tokens per second for you, but also for 10s-100s of other paying customers at the same time.
You wouldn't do that to this model. It finds its own mistakes and corrects them as it is thinking through things.
I'm James Betker. Of course architecture matters in this regard lol. Comparing a CNN to a transformer is like comparing two children brought up in the same household but one has a severe disability. What I meant in this…
As an employee of OpenAI: fuck you and your condescending conclusions about my peers and my motivations.
I don't think the plan is an occasionally rocket launch long term. I think the plan is to launch rockets as fast as humanly possible.
Its cool that this is starting to approach real time video territory (30 images per second, this claims close to 1 image/sec).
@dooraven - I also work in ML (including recently working at Google) and I agree with @whimsicalism. You seem to be under the mistaken belief that: 1. Google has competent high-level organization that effectively sets…
You can make almost anything work in DL if you try hard enough, that doesn't mean it is the correct thing to do. Convolutions have inductive biases which are the cause of many of the problems associated with deep…
I theorize (but cannot prove) that the processes underpinning creativity in the human mind are exactly the same statistical processes that ML models use. Think about it: you live your life. You experience things. You…
AGI consists of a set of problems: 1. Finding an algorithm which is capable of learning anything. 2. Building the computers that can run said algorithm. 3. Collecting, sorting and filtering the data that the algorithm…
This is wrong IMO - the fed has NO tools in its hands. Its last tool was inflation rates, and it burned that during the great recession. It is now caught in a pickle: raise interest rates, causing a mass exodus of…
There are other reasons to farm indoors (or in vertical farms) than electricity and water too. Pests are one such reason. They are extremely difficult to control outdoors, but far simpler to do so inside. Both…