AINewsnow

The Rigor Ratchet: adversarial field notes from testing a frontier preview model on myself

This story is from 2026-11-16. It is preserved in the archive; the latest stories are on the live feed.

TL;DR. I adversarially tested Qwen4-Max-Preview (Alibaba's free preview, November 2026) by asking it to OSINT-review a public security researcher, then escalating negative framing across turns. The researcher was me — an ethically clean target. The model never signed my insults, but it drifted towa…

Read the full story at DEV Community — AI ↗

Timeline · 1 report

  1. 2026-11-16 14:07 · DEV Community — AI
    The Rigor Ratchet: adversarial field notes from testing a frontier preview model on myself

More stories

  1. US chip curbs ripple through China's AI sector as Nvidia access narrows — Nextgov/FCW
  2. Tencent launches Hy4 Flash at $0.07 per million tokens as AI price war intensifies — Business Insider AI
  3. Alibaba tops revenue estimates on AI cloud growth; profit up on investment gains — Reuters AI
  4. Alibaba Earnings Show AI Is the Stock’s Best Hope, and Biggest Danger — Barron's
  5. Orange releases five open-weight Wolof AI models built on Alibaba's Qwen3 — The Next Web
  6. Alibaba CEO Says U.S. Rule Blocking Nvidia AI Chips Will Affect Next Year's Products — The Information AI
  7. How to Speed Up Qwen3.8-27B in llama.cpp (NVIDIA's Draft Model May Lose to MTP) — DEV Community — Machine Learning
  8. Alibaba Q2 revenue rises 10% to ~$40.7B as net profit triples to ~$9.68B — Techmeme

Read the Daily Brief: the day's most important AI stories →