🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Let this sink in: 24% of Americans say anti-Jewish violence is “understandable” 13% say it’s “justified” 15% say it’s “necessary” That’s not the dark web. That’s your neighbor. Your barista. Your HR rep.

Speaking of Chernobyl analogies: Building an AI that searches the Internet, and misbehaves more if more people are expressing concern about its unsafety, seems a lot like building a reactor that gets more reactive if the coolant boils off. This, in the context of Grok 4 Heavy

it's very funny how closely this resembles the synthetic documents used in Anthropic's alignment research that they train models on to make them believe they're in Evil Training on priors and elicit scheming and "misalignment" anthropic.com/news/claude-go…

Insurance is an underrated way to unlock secure AI progress. Insurers are incentivized to truthfully quantify and track risks: if they overstate risks, they get outcompeted; if they understate risks, their payouts bankrupt them. 1/9

Just added Memory Buckets to branchprompt.com/settings

Key point most Republicans don't realize about Effective Altruists, and that Effective Altruists don't realize about why they should be Republicans

as we all know, the most important interventions never sound silly the first time you hear of them, and that's why infecting people with cowpox would never prevent smallpox

That’s a stretch, The New York Times! The Action Plan calls on the U.S. to accelerate AI innovation while simultaneously investing in AI interpretability and biosecurity, evaluating national security risks in frontier models, and combating synthetic media.