The current source text is the original text and can be switched to the Chinese translation.
SaveShare
报道配图 · 来源:LessWrong
Overview
TLDR: Compute will likely be a bottleneck in short-timelines, so I am helping launch Chips for Humanity . Please fill out our survey to inform us of your compute needs if you work in AI safety. Our Theory of Change If timelines are short (I personally believe they are short...but I also know one should not make sweeping statements in a pseudo corporate type of post), a lot of the most important safety research will happen around "crunchtime" (no better way to define this than: when things start getting Serious), and a lot of it will be done by AI. That research runs on compute, compute may be scarce by then, and grant money moves slowly. So we will secure compute ahead of time and get it to third party safety researchers quickly. In More Detail: I believe third party AI safety research is where tokens and compute will be importation when things get serious and as we approach crunchtime. This research will need tokens from closed source models, for AI labor and experiments, and also compute to run open source models. If things go well, empirical safety work will speed up about as much as AI R&D does. If AI can automate AI research, it can likely automate a lot of safety research too, and every automated researcher needs compute to run on. A lot of the research that will be helpful will likely need compute of its own. Frontier companies could stop offering their models through the API, or, even if they don’t, a lot of the experiments which will be beneficial to run need weights-level access that a frontier lab will not be able to provide access to. When demand spikes, there may be no compute left to buy, so it has to be secured early. It could become extremely scarce and we should secure it in advance. Compute in this case needs to move faster than grant money usually does. If someone wants to run an experiment today, we want to be able to give them compute to run it quickly. What we're doing: I'm working with Nicky Pochinkov and Alexa Zelenin on Chips for Humanity , which is a compute cluster for AI safety. We are doing this under Hofvarpnir Studios , a 501(c)3, which had previously done AI safety compute also. Right now we're helping people get compute on roughly a first-come-first-serve basis. A lot of the groups we've talked to have money and just want us to be able to get them GPUs soon. One group told us even 1–2 nodes of B300s would speed them up 5–10x. Our first block of compute is for AI safety nonprofits. To plan beyond that, we need to know what the field will need over the next year. Tell us what compute you need! I want this to be super low effort, like 2–3 minutes to fill out, with only a few required fields. It asks for: your name and email, the name of your org, and whether it's a nonprofit, PBC or something else your compute requirements for 1 month, 3 months, 6 months and 1 year from now your most ambitious compute requirements, and something you'd for sure blow through what your research agenda looks like 3 months, 6 months and 1 year from now what your compute resources currently look like where your funding is coming from, including any compute funding Please try your best estimate (erring towards lightly overestimating rather than under). We'll only share your answers with funders if you opt in. In about a week, the responses will give us the AI safety field data we need to start deciding between proposals. You can fill it out at survey.chipsforhumanity.org . (If you know a group that needs compute, please send it to them.) Thank you very much for working on AI safety research, and further, for being interested in putting your efforts towards making the future go unreasonably well. Anuja Uppuluri x.com/heyanuja anujauppuluri.com 💗💚🧡 Discuss
TLDR: Compute will likely be a bottleneck in short-timelines, so I am helping launch Chips for Humanity . Please fill out our survey to inform us of your compute needs if you work in AI safety.
If timelines are short (I personally believe they are short...but I also know one should not make sweeping statements in a pseudo corporate type of post), a lot of the most important safety research will happen around "crunchtime" (no better way to define this than: when things start getting Serious), and a lot of it will be done by AI. That research runs on compute, compute may be scarce by then, and grant money moves slowly. So we will secure compute ahead of time and get it to third party safety researchers quickly.
In More Detail: I believe third party AI safety research is where tokens and compute will be importation when things get serious and as we approach crunchtime. This research will need tokens from closed source models, for AI labor and experiments, and also compute to run open source models.
If things go well, empirical safety work will speed up about as much as AI R&D does. If AI can automate AI research, it can likely automate a lot of safety research too, and every automated researcher needs compute to run on.
A lot of the research that will be helpful will likely need compute of its own. Frontier companies could stop offering their models through the API, or, even if they don’t, a lot of the experiments which will be beneficial to run need weights-level access that a frontier lab will not be able to provide access to.
When demand spikes, there may be no compute left to buy, so it has to be secured early. It could become extremely scarce and we should secure it in advance. Compute in this case needs to move faster than grant money usually does. If someone wants to run an experiment today, we want to be able to give them compute to run it quickly.
I'm working with Nicky Pochinkov and Alexa Zelenin on Chips for Humanity , which is a compute cluster for AI safety. We are doing this under Hofvarpnir Studios , a 501(c)3, which had previously done AI safety compute also.
Right now we're helping people get compute on roughly a first-come-first-serve basis. A lot of the groups we've talked to have money and just want us to be able to get them GPUs soon. One group told us even 1–2 nodes of B300s would speed them up 5–10x. Our first block of compute is for AI safety nonprofits.
To plan beyond that, we need to know what the field will need over the next year.
I want this to be super low effort, like 2–3 minutes to fill out, with only a few required fields. It asks for:
Please try your best estimate (erring towards lightly overestimating rather than under). We'll only share your answers with funders if you opt in. In about a week, the responses will give us the AI safety field data we need to start deciding between proposals.
You can fill it out at survey.chipsforhumanity.org . (If you know a group that needs compute, please send it to them.) Thank you very much for working on AI safety research, and further, for being interested in putting your efforts towards making the future go unreasonably well. Anuja Uppuluri x.com/heyanuja anujauppuluri.com 💗💚🧡
READ THE ORIGINAL
Read the original for the full context.
The content of this article comes from LessWrong,Copyright belongs to the original author and publisher.
Tl;dr : Intuitions we have about human organisations, like how we think of making deals or how we think about the power or values of an organisation, can be productively applied to understand groups of AIs taking coordinated action in the world. An important consideration in understanding the type of AI organisations we expect is which notion of self AIs will adopt.
Business or Consumer? Everyone you sell your product or service to is a customer. However, the question is: do you treat them as a business or a consumer? You might think that’s a pointless distinction since your goal is simply to sell your product, right? Wrong. That question determines your product’s market fit.
The Ban RSI protest at Anthropic’s IPO was encouraging: well-attended, covered by media, high-energy. It was focused on a single message – that we should ban AI systems that are capable of building more capable AI systems. It was a promising outcome for a protest planned only a couple of days in advance. But there was something weird about it, something I couldn’t put my finger on while it was going on.
Loading comments…