Harry Stebbings says his most candid AI podcast covered open models, cyber security and China vs. the US
arena · x · 2026-08-04
Harry Stebbings says the episode with @mlangelopoulos is one of the most candid AI discussions he has hosted, covering cyber security, open source, China vs. the US, and distillation.
- The discussion argues that much of the AI podcast circuit is overly sanitized because guests fear upsetting major model providers.
- It then dives into whether Chinese open models are genuinely ahead, why cyber security is in a poor state, and why many “neolabs” may be research projects that won’t survive.
- The thread also teases a broader take on Kimi and other open models outperforming Western closed models in some areas.
More from Companies & People
- Healthcare credentialing delays can keep doctors idle for months and drive up costs — LeahHoustonMD · 2026-08-04
- Cursor’s solution architecture lead grew up through Google and Microsoft — Fowe · 2026-08-04
- Scalar Field launches an agentic trading desk for stocks, crypto and prediction markets — NoCartographer4725 · 2026-08-04
- Make the golden path obvious: terminal copy-paste plus zero-click checkout — yacineMTB · 2026-08-04
- OpenAI economics team says AI’s effects need new measurement infrastructure — soumitrashukla9 · 2026-08-04
- Cloudflare’s agents package tops 1.29 million weekly npm downloads — whoiskatrin · 2026-08-04