America already lost one AI race | TWiAI Ep 23

<p><strong>This Week In Startups is made possible by:</strong></p><p><strong>PAYPAL OPEN</strong></p><p><br></p><p><strong>Today’s show:</strong></p><p><strong>This week, our panel of AI experts — Anand Kannappan (Patronus AI), Ori Goshen (AI21 Labs), and Alex Finn (Henry Intelligent Machines) — join special guest host Lon Harris to debate whether Washington will move to formally limit powerful foreign AI models like Moonshot’s Kimi K3 and</strong><a href="http://z.ai"><strong> Z.ai</strong></a><strong>’s GLM-5.2.</strong></p><p><strong>What does “winning the AI race” actually mean on a practical level? Why is cybersecurity, not benchmarks, likely the actual battleground. And why did Hugging Face have to use GLM-5.2 to investigate their recent breach rather than a frontier US model like Fable 5?</strong></p><p><strong>PLUS why agents still fail so much on long horizon tasks, why proper model routing has such a HUGE impact on enterprise AI bills, and we take a look at Alex’s hacked digital typewriter he uses to code without distractions.</strong></p><p><br></p><p><strong>Guests:</strong></p><p><br></p><p><strong>Anand Kannappan on X:</strong><a href="https://x.com/anandnk24"><strong> https://x.com/anandnk24</strong></a></p><p><strong>Patronus AI:</strong><a href="https://www.patronus.ai/"><strong> https://www.patronus.ai/</strong></a></p><p><strong>Ori Goshen on X:</strong><a href="https://x.com/origoshen"><strong> https://x.com/origoshen</strong></a></p><p><strong>AI21:</strong><a href="https://www.ai21.com/"><strong> https://www.ai21.com/</strong></a></p><p><strong>Alex Finn on X:</strong><a href="https://x.com/AlexFinn"><strong> https://x.com/AlexFinn</strong></a></p><p><strong>Alex’s Vibe Coding Academy:</strong><a href="https://www.skool.com/vibe-coding-academy/about"><strong> https://www.skool.com/vibe-coding-academy/about</strong></a></p><p><br></p><p><strong>Relevant Links:</strong></p><p><br></p><p><strong>Axios: “The Secret Trump Administration Battle to Fight Chinese AI”:</strong><a href="https://www.axios.com/2026/07/20/ai-us-china-open-source-kimi"><strong> https://www.axios.com/2026/07/20/ai-us-china-open-source-kimi</strong></a></p><p><strong>Chamath post on US AI options:</strong><a href="https://x.com/chamath/status/2079457219892871458?s=20"><strong> https://x.com/chamath/status/2079457219892871458?s=20</strong></a></p><p><strong>Brian Armstrong post on spend vs. token usage:</strong><a href="https://x.com/brian_armstrong/status/2070670644577280109"><strong> https://x.com/brian_armstrong/status/2070670644577280109</strong></a></p><p><strong>Hugging Face: Security incident disclosure:</strong><a href="https://huggingface.co/blog/security-incident-july-2026"><strong> https://huggingface.co/blog/security-incident-july-2026</strong></a></p><p><strong>VentureBeat Hugging Face coverage:</strong><a href="https://venturebeat.com/security/safety-guardrails-blocked-hugging-faces-defenders-not-the-attacker-when-an-ai-agent-breached-its-systems"><strong> https://venturebeat.com/security/safety-guardrails-blocked-hugging-faces-defenders-not-the-attacker-when-an-ai-agent-breached-its-systems</strong></a></p><p><strong>David Sacks post on cyber guardrails:</strong><a href="https://x.com/DavidSacks/status/2078984980588531855"><strong> https://x.com/DavidSacks/status/2078984980588531855</strong></a></p><p><strong>Databricks Unity AI Gateway:</strong><a href="https://www.databricks.com/product/artificial-intelligence/unity-ai-gateway"><strong> https://www.databricks.com/product/artificial-intelligence/unity-ai-gateway</strong></a></p><p><strong>Ramp’s Router:</strong><a href="https://ramp.com/router/"><strong> https://ramp.com/router/</strong></a></p><p><br></p><p><strong>Timestamps:</strong></p><p><br></p><p><strong>0:00 Will the US limit foreign open source models?</strong></p><p><strong>5:40 Are the American frontier AI labs in danger of bankruptcy?</strong></p><p><strong>7:15 The risks of over-regulating AI</strong></p><p><s