SAFETY

–

01 November 2025 15h57

Cool, I'm fine with "anti-ASI condition" or "anti-extinction coalition".

1 November 2025

Terminology for AI Risk Advocates: Doomer, Pessimist, Anti-Extinction

By

–

31 October 2025 21h08

"AI pessimist" or "anti-extinction" could both be quibbled with, but anyone who recognizes the word "doomer" could parse either of these ordinary terms given slight context.

Rejecting ‘Doomer’ Slur: AI Safety Discourse and Propaganda

By

–

31 October 2025 18h01

The slur "doomer" was an incredible propaganda success for the AI death cult. Please do not help them kill your neighbors' children by repeating it.

→ View original post on X — @alexjc

Language Models Struggle With High School Math Fundamentals

By

@alexjc

–

31 October 2025 12h25

Language models perform poorly on high-school math? You don't want to hear this, but the problems started in grade-school. The moment we (collectively) found acceptable that mid-tier models could score only 75%-85% on a GSM test set of 1.32k straightforward problems…

AI Safety: Extrapolated Predictions Fall Short of Control Requirements

By

–

31 October 2025 3h21

I already extrapolated it two decades into the future one decade earlier, and my prediction came up far short of what's needed to control something smarter than humanity.

→ View original post on X — @askalphaxiv

Stanford Method Detects Copied Fine-Tuned AI Models

By

@askalphaxiv

–

31 October 2025 1h16

Someone stole your model & u can’t prove it? This Stanford paper just showed that you can find out if a model is a copy or finetuned based on your model with just its generated text So if someone yoinks DeepSeek-v3.2 and finetunes it, it’ll leave statistical traces!