ai-safety
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Heretic hits 20.5k GitHub stars: automated LLM "abliteration" tool now lets anyone strip safety alignment with one command |
|
0 | 5 | May 27, 2026 |
| Anthropic bars under-18s from its AI services, drawing criticism over access and competitive motives |
|
0 | 2 | May 25, 2026 |
| Researcher cites Yudkowsky's 2016 Turing test prediction to question his calibration on AI doom |
|
0 | 5 | May 24, 2026 |