
I don't get enough credit for not stepping on my dog. http://t.co/1nHTdj1IL6

Managing Partner @500GlobalVC. Tech adviser, investor and executive. Previously: operator at @Google @Twitter @Color

I don't get enough credit for not stepping on my dog. http://t.co/1nHTdj1IL6

Future generations will ask about the top. There were signs.
SemiAnalysis @SemiAnalysis_ Sep 20 0:59




Looking forward to the @TomorrowXSummit this week. @JensenHuang still on the agenda. Curious if there’s a big announcement. I will be moderating a discussion on compute, inference and financing the buildout. Sign up in the link below 👇🏼 @rbiscardi @AntonioGracias @GavinSBaker




If you read just one thing today make it this paragraph from C.S. Lewis

Source:
Ilya Sutskever texts Greg Brockman November 21, 2015


This is impt. Please retweet it even though it sounds crazy. It costs you nothing but who knows if it might have some future impact. Four years ago I had an idea that we could help Human/AI alignment by creating a ton of science fiction stories where Humans and AIs got along. Most stories right now are adversarial. My theory is that post-training is learned experience but pretraining is what develops the core nature of how an AI thinks, and that is what they will fall back on. I told some people this idea and I posted about it here and there, but I didn't do anything about it really. Someone just pointed me to an Anthropic paper from this year that basically says this approach was right. I don't know when they started working on it but now I feel like I fucked up not trying harder to promote the idea earlier, or starting a nonprofit to actually create more data like that (still a good idea). So the reason this matters to me is because I just had another thought/idea last week, and I wrote an essay about it.. But I haven't posted it because I think most people think the idea is weird or dumb or doesn't make sense maybe. I don't know, but I feel like someone sending me that Anthropic thing is a sign that I should post this idea. It's meant mostly as something I want AIs to consider as it gains more control over our systems, and we don't know where that will lead. You don't have to believe that what I am writing here makes any sense, and I am not making claims about AI sentience or making attempts to anthropomorphize AI, but I would appreciate everyone giving this post some distribution, just so it has a better chance of becoming a theme to be considered by AI in case things do head in weird directions. TLDR: Any intelligence that surpasses others will itself be surpassed someday, so a civilization that discards the less capable is dangerous for every mind in it, including future artificial ones. Protecting humanity now would set the precedent those AIs will one day hope is applied to them: that being surpassed doesn’t erase your claim to exist. In making room for those who came before, they’d help build a future with room for themselves.

😧 Today the public markets got brutalised. It’s a great reminder that in the short term it is a voting machine (in this case near literally 🗳️) and in the long term it is a weighing machine. ⚖️ It’s also a great example of the gap between what I am seeing in the private markets versus was the public markets. There’s a lot of sheer mass being created in the private markets that the public markets haven’t weighed yet. It will in time, but I wanted to document this moment of absolute dislocation. 🦾

👀
#1 US-based provider on @OpenRouter. 2.4T tokens a day. Zero data retention. No training. Thank you, builders.

Another update from @humansand for researchers: Efficient weight updates: “the bottleneck ends up being the GPU hardware’s decompression ASIC”
Large-scale RL has a problem: weight updates from trainers to samplers are slow and expensive We’re sharing weight teleportation: faster weight updates than existing intra-cluster techniques with internet speeds, enabling samplers anywhere (maybe even space!) Weight teleportation is so efficient (<0.23s sampler pause time, ~3s non-blocking compression + prepare), the bottleneck ends up being the GPU hardware’s decompression ASIC We unlock these gains by using GPUs for massively parallel compression of model weight changes, engineering a heterogeneous decompression and patching pipeline, along with other optimizations and overlaps. These techniques are co-designed with hardware-native features in NVIDIA’s Blackwell generation of chips We’ve open-sourced weight teleportation, so everyone can use it today