51–60 of 358 posts
The #1 spot was someone's Sun E15k or something like that.
Is it possible to distribute AI training like this so that people could donate their idle processing power to creating a truly open source unencumbered LLM?
As far as I can tell this isn't any different conceptually from the distributed inference setup that PyTorch/Huggingface Accelerate/etc support. Of course those were built for cooperating machines owned by the same organization with lots of high-speed I/O. In an @home/BOINC setup, you need to package things up into work units that get sent out to other people's computers, and you don't get anything back until the end. So it's higher-latency and you have to worry about people cheating the system with bogus work.
I'm not sure if this would be more efficient or cheaper than just renting out a bunch of AWS spot instances with GPUs in them.
Earlier quoted context omitted.
> The idea that all of us could together could crack RSA keys wasn't just a thought exercise, people signed up and did it. Hasn’t changed that much. Here we are in 2023 where most of this distributed power goes to cracking sha 256 hashes. Just for speculation and profit instead of leaderboards.
Bitcoin has nothing to do with "cracking SHA256 hashes", for what it's worth. The only thing remotely related is how it brute-forces inputs until it generates hashes that begin with a certain number of bits set to 0.
Is it possible to distribute AI training like this so that people could donate their idle processing power to creating a truly open source unencumbered LLM?
It's not clear that machine learning models consisting of weights are copyrightable in the first place. They are not the results of human authorship, but are instead essentially produced by a machine trying to optimize a bunch of values.
I don't specifically miss SETI@home as much as I miss the spirit of the internet back then, the "what will they think of next!?" The idea that all of us could together could crack RSA keys wasn't just a thought exercise, people signed up and did it. Now computers are so complex and task heavy and fraught with problems it seems like more hassle than it's worth. I liked when wondering "why is my machine so slow", that…
Classic SETI@home's website[1] is a blast to browse.
I don't specifically miss SETI@home as much as I miss the spirit of the internet back then, the "what will they think of next!?" The idea that all of us could together could crack RSA keys wasn't just a thought exercise, people signed up and did it. Now computers are so complex and task heavy and fraught with problems it seems like more hassle than it's worth. I liked when wondering "why is my machine so slow", that…
> The idea that all of us could together could crack RSA keys wasn't just a thought exercise, people signed up and did it. Hasn’t changed that much. Here we are in 2023 where most of this distributed power goes to cracking sha 256 hashes. Just for speculation and profit instead of leaderboards.
I don't specifically miss SETI@home as much as I miss the spirit of the internet back then, the "what will they think of next!?" The idea that all of us could together could crack RSA keys wasn't just a thought exercise, people signed up and did it. Now computers are so complex and task heavy and fraught with problems it seems like more hassle than it's worth. I liked when wondering "why is my machine so slow", that…
I don't specifically miss SETI@home as much as I miss the spirit of the internet back then, the "what will they think of next!?" The idea that all of us could together could crack RSA keys wasn't just a thought exercise, people signed up and did it. Now computers are so complex and task heavy and fraught with problems it seems like more hassle than it's worth. I liked when wondering "why is my machine so slow", that…
The recent neural network "revolution" is giving me those vibes, even though it'll ironically kill the old internet with the upcoming spam bots.