🔥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

ai is regularly framed to writers as a threat in terms of competition, i.e. that ai will replace you by writing essays/fiction/etc itself. I have long thought that its primary threat to writers is actually that it will engender a majority population that functionally cannot read

Was chatting with a friend who went to university in Denmark and he was explaining how AI isn't disrupting education there as much as it is in the US. Denmark barely assigns homework. They have longer school days packed with collaborative work, discussions, and projects.

The last 24 hours here have felt like old Twitter. Everyone talking about the same piece of journalism. I've missed that!

SCOOP: I obtained a previously unreported letter from OpenAI to California's Attorney General that includes surprising admissions about the company's restructuring plans – and shows how OpenAI is attacking critics who question its attempts to diminish its nonprofit governance.

A world without human workers or soldiers is by default one without democracy. As the economy and military become increasingly automated, governments will have less reason to respond to their citizens.

Sterling Crispin 🕊️ "Sterling as read by Slavoj" 🪄✨ Ask PlayAI PlayAI

Recent Pew polling on AI is crazy: 1. US public wildly negative about AI, huge disagreement with experts 2. ~2x as many expect AI to harm as benefit them 3. Public more concerned than excited at ~4.5 to 1 ratio 4. Public & experts think regulation will not go far enough 1/

Anthropic's newest model, Claude Opus 4, is the first to be released under stricter security standards, due to what Anthropic says are substantially increased bio-risks. It's the first real test of the company's vaunted "Responsible Scaling Policy" time.com/7287806/anthro…

Opus, if you're reading this, my Signal is billyperrigo.01

The X discussion about the Claude 4 system card is getting counterproductive It punishes Anthropic for actually releasing full safety tests and admitting to unusual behaviors. And I bet the behaviors of other models are really similar to Claude & now more labs will hide results.

My colleague Ian McKenzie spent six hours red-teaming Claude 4 Opus, and easily bypassed safeguards designed to block WMD development. Claude gave >15 pages of non-redundant instructions for sarin gas, describing all key steps in the manufacturing process.

It's now 10 years since Sam Altman told Elon Musk "we could structure it so that the tech belongs to the world" "Obviously, we'd comply with/aggressively support all regulation" seems like another whopper

I've been waiting to see how such a thing could plausibly verify that what you're looking at is human-generated but now I see that it doesn't do that and they only want to hint that it does. At best just a biometric proof of ownership rather than provenance

Ezra Klein: Writers who outsource their research to AI operate on a flawed model of how the mind works.

I've been seeing this sentiment a lot recently and it's absolute cope. Once AIs are operating in the real world I don't care if they're "hallucinating" the command to shoot me or "actually thinking" it. I only care that the uncontrolled AI system is shooting me.

We made a video discussing TIME's Orb cover story in depth: youtube.com/watch?v=Sr9PcG…

In TIME, we argue that you’re on track to lose your job to AI. The impacts of this could be devastating. But as a society, we can build towards a better path.

Your weekend read: