CharlieKey takeaways from your favorite podcasts
← 14 November 2026TBPN

Jensen Out, TBPN Merch Drop, GPT 6 Cyber, Dylan Patel Live

With Jensen Huang out as Nvidia CEO, the hosts and guests argue the loss of a founder-mode decision-maker threatens Nvidia's pace, open-source AI and China policy, while opening room for rival chipmakers.

Watch on YouTube ↗

Charlie's summary

  1. Nvidia's founder-mode edge is at risk: committees will slow technical bets, supply deals and the 'central bank of AI' investing

    The hosts and Dylan argue Jensen personally picked one chip option from A, B, C and D and got suppliers behind it. Dylan names three exposed areas: technical decisions, supply chain (a less charismatic CEO won't get 60 to 70% of TSMC capacity at below-market prices) and investments. Deals Jensen once handshook may now go to committee, which is “going to make decisions slow.” Talent recruiting also loses its founder pitch.

    ▶ 11:14
  2. Open-source AI in America is the biggest loser if Jensen's China defense collapses

    Dylan thinks “open source in America would not exist without Jensen.” Reflection and Thinking Machines have Nvidia backing, and companies like Harvey, Sierra and inference providers rely on Chinese open models. With Jensen's credibility gone, he expects tighter China restrictions, which could stall Chinese models or stop them from being open-sourced. Application companies might then have to use closed models and become wrappers.

    ▶ 1:04:15
  3. The best CEO picks are outsiders who can run both chip supply chains and technical details: Charlie Kawwas and Johnny Srouji

    Dylan says Colette Kress is excellent but cannot lead the company technically. Internally he sees Ian Buck and Jonah Alben as the best fits, but neither covers technology, deal-making and supply chain. His favorites are Charlie Kawwas, number two at Broadcom, and Johnny Srouji, Apple's head of hardware. He argues defending a category you created differs from competing as number two, which is why Nikesh Arora is a less obvious fit.

    ▶ 1:08:18
  4. Nvidia's troubles tilt the field toward Anthropic over OpenAI and toward diversified Chinese labs

    Dylan says Chinese labs that leaned on Nvidia are now weaker than those already diversifying, such as Zhipu, which supports seven Chinese accelerators, and DeepSeek, which is close to Huawei. The same logic applies to the US labs: OpenAI is 90-plus percent Nvidia, while Anthropic is about a third or less. Weaker open source also helps closed-source labs: “The closed source is winning.”

    ▶ 1:12:53

Give your email to unlock the rest of the key points — for every episode.

Do more with Charlie

Free — sign in with your email, no password.