ThursdAI Recap: GLM 5.3 and Qwen 27B Drop, OpenAI Pauses RL for Safety
altryne · x · 2026-08-28
The latest ThursdAI podcast covers the week in AI: Z.ai released GLM 5.3 (API-only for now) and the tiny Qwen 3.89 27B; OpenAI announced pausing RL efforts after the scary AI swarms hacking incident, dedicating up to 20% of compute to reviewing agent thinking processes; Stripe reportedly buying OpenRouter for $8B. The episode also discusses 4 Flash models (Omni 1.1 Flash, Qwen next, GLM 5.3 Flash), a video generated in under 10 seconds with fal's new H3 Max, plus interviews with Andy Masley on DC debates and Kwindla on phone LLMs.
More from Companies & People
- Enterprise AI leaders should build model-agnostic stacks: own the evals, own the models — realmadhuguru · 2026-08-28
- Counterpoint: if everyone stays in SF, progress shrinks to easy, software-only problems — MaxUnfried · 2026-08-28
- Inaugural Unprompted AI Security Conference Set for Sydney, Sept 2026 — dyn___ · 2026-08-28
- The Atlantic defends publishing AI-written op-ed, exploring future of human prose — TuhinChakr · 2026-08-28
- Google DeepMind is losing its grip on elite AI talent, data shows — pstAsiatech · 2026-08-28
- Fitbit founders return with new family safety band — technextpreneur · 2026-08-28