🔥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

Today, we propose the Conditional AI Safety Treaty in TIME as a solution to AI's existential risks. AI poses a risk of human extinction, but this problem is not unsolvable. The Conditional AI Safety Treaty is a global response to avoid losing control over AI. How does it work?

Big new review! 🟦Open Problems in Mechanistic Interpretability🟦 We bring together perspectives from ~30 top researchers to outline the current frontiers of mech interp. It highlights the open problems that we think the field should prioritize! 🧵

Science moves so fast! Wild to see "The Geometry of Categorical and Hierarchical Concepts in LLMs" get an ICLR2025 oral while our blogpost undermining their findings wins runner-up best blogpost. No consensus yet! What do you think?

1/10: In our new paper, we develop scaling laws for scalable oversight: oversight and deception ability predictably scale as a function of LLM intelligence! We quantify scaling in four specific oversight settings and then develop optimal strategies for oversight bootstrapping.

What research do we need to get to aligned AGI? Our AI Security Institute alignment research agenda covers what we think the critical research problems are, and has a list of tractable problems! Broadly, we divide things into two sets of questions: [1/5]

This Singapore conference was an amazing AI safety comeback after the Paris flop: great consensus between a who's who from the US, China, top companies, AISI's, etc on what safety research needs to get done: aisafetypriorities.org

Holy shit this piece. 'Multiple AI platforms now offer tools to leave AI-generated feedback on students’ essays. Which raises the possibility that AIs are now evaluating AI-generated papers, reducing the entire academic exercise to a conversation between two robots — or maybe

🚀 Excited to announce the launch of the AISAR Scholarship, a new initiative to promote AI Safety research in Argentina! 🇦🇷 Together with Agustín Martinez Suñé, we've created this program to support both Argentine established researchers and emerging talent, encouraging

I want to remind everybody again that there is a vaccine for Lyme disease. It‘s just that… nobody cared. It‘s right on the article for Lyme disease. One of the most insane paragraphs ever on Wikipedia:

OpenAI catches a lot of shit because it promised the public so much and is now falling short. But don't forget that it remains miles ahead of: • Meta • xAI • Microsoft • DeepSeek which all get off light because they've only ever promised little or nothing!

A 2g antiviral gum neutralized multiple strains of influenza and herpes in under 15 minutes. The active compound, FRIL (Foetuin Binding Ricin B like Lectin), binds to and disables viruses. Formulated with LabLab bean gum, it remains shelf stable for over two years at room

Huge news. The U.S. Senate has voted overwhelmingly (99-1) to reject the moratorium on state-level AI legislation, recognizing the growing need for common-sense safeguards to protect American jobs, families and lives.

BIG AI/biotech Breakthrough In a groundbreaking study published in Nature Communications, University of Pennsylvania researchers used a AI system called APEX to scan through 40 million+ venom encrypted peptides -proteins evolved over millions of years for attack and defense. In