John is a former AI-focused journalist who led TechCrunch's AI coverage, writing technical pieces on how companies commercialized machine learning for enterprise problems, giving him pattern-recognition across hundreds of AI teams and go-to-market approaches. His team includes machine-learning engineers and data scientists who actively brainstorm and prototype AI strategies with portfolio companies, providing a more technically embedded support model than traditional venture capital approaches.

• Intrigued by the evolution of artificial intelligence beyond basic model power, with a specific emphasis on system memory and autonomous agents. Believes real enterprise value comes from these advanced applications rather than raw computing strength, noting that systems thinking is crucial to avoid brittle setups. As an example, highlights the contrast between how robotics and software agents handle external memory and praises teams building dedicated memory infrastructure for machine learning.
• A strong focus on supply chain modernization and physical infrastructure. Views the supply chain sector as one of the few business areas where massive innovation can directly improve everyday life and combat economic pessimism. Points to macro trends like rising commercial transport fuel costs and the massive adoption of electronic logging devices in commercial trucking as prime catalysts for new fleet management startups.
• Dedicated to upgrading outdated business infrastructure, noting that enterprise tools are severely lagging behind consumer technology. Sees massive opportunities to bring modern conveniences like seamless search, better product recommendations, and upgraded payment systems into the enterprise space. Specifically mentions backing teams that are overhauling clunky business payments to bring them up to modern consumer standards.
• Offers practical advice on startup execution and go to market strategies. Cautions founders against expecting massive customer conversion from media coverage and encourages focusing on macroeconomic trends for new ideas. For instance, frequently warns founders that public relations hits are a long game and that actual click through rates from major press outlets are surprisingly low.
• Fascinated by the intersection of human and machine collaboration for solving complex workflows. Believes that while machine learning alone cannot handle every task perfectly, combining human intelligence with artificial systems creates highly effective synergies. Envisions a future where machines themselves are part of the operational loop, breaking imprecise generative tasks down into smaller, more reliable composite steps.