Nathan Benaich
@nathanbenaich
solo member of investment staff @airstreet @airstreetpress @stateofaireport @raais
🤓 i'm running an ai usage survey for the @stateofaireport the results will be open sourced with the State of AI Report in october '25 it takes 10 mins and i want to hear from everyone, regardless of their level of expertise and usage head to stateof.ai/survey thanks!
buying candy and cigs, european summer
🚨 Novak Djokovic spotted Shopping in a Supermarket..
It took 36h for Tomorrowland to rebuild a decent main stage after the freak fire. Meanwhile, the Brooklyn Mirage is still not open. Why are things so slow in America?
agree
It's trendy for politicians to show off extremely short timelines. In the UK, this is especially curious given (a) the Secretary of State for technology is one of them and (b) with a few notable exceptions, the UK is behaving as if AGI is centuries away Link below 🇬🇧
oh boy
The WSJ is reporting that xAI is working with Valor Equity Partners to raise an additional $12 billion in capital to buy the chips for Collosus 2. The report also claims that the Grok IP was used as collateral for the $5 billion raise in June.
from move 37 to drug design: a talk by @maxjaderberg of @isomorphiclabs at @raais 2025 a radical mission: solving disease. it's not a metaphor - it’s a systems-level wager that the same kinds of models that cracked go can crack biology... hope you enjoy it!
interesting
Today (w/ @UniofOxford @Stanford @MIT @LSEnews) we’re sharing the results of the largest AI persuasion experiments to date: 76k participants, 19 LLMs, 707 political issues. We examine “levers” of AI persuasion: model scale, post-training, prompting, personalization, & more 🧵
i cannot wait to have an imo gold grade model for my trip to the supermarket planner
Our IMO gold model is not just an "experimental reasoning" model. It is way more general purpose than anyone would have expected. This general deep think model is going to be shipped so stay tuned! 🔥
IMO: here are the rules for announcing the results. GDM: sounds good. OAI: we can just do things and not stick to the announcement rules.
An advanced version of Gemini with Deep Think has officially achieved gold medal-level performance at the International Mathematical Olympiad. 🥇 It solved 5️⃣ out of 6️⃣ exceptionally difficult problems, involving algebra, combinatorics, geometry and number theory. Here’s how 🧵
lol wut on earth did i just watch
It’s official: @Grammarly has acquired @Superhuman! We’ve already built the AI superhighway that brings AI everywhere users work. Now, we’re building the agents that will collaborate across all communication tools—and bringing the Superhuman team together with Grammarly and…
i continue to argue that gov should be the buyer of first resort for new technologies in lieu of handing out cheques to investors or directly investing in the market to fix “funding gaps”, esp. those that are critical to our future (eg ai, computing, energy) policymakers should…
there is no time.
The next large-scale war on European soil won’t offer the luxury of time. If we find ourselves wiring up commercial drones again, it won’t be out of necessity. It will be a failure of preparation and imagination. We’ve seen the factories. We’ve had the warnings. There will…
wait are we meant to be surprised?
Something I noticed at ICML: for the most part, VCs are strikingly disinterested in the technical content of the conference, and the discipline at large. You could be describing what obsoletes the transformer, and they couldn’t care less.