About this episode

Ari Morcos and Rob Toews discuss the current state of AI research, the impact of capital on innovation, and make bold predictions for 2026.

Ari Morcos and Rob Toews return for their spiciest conversation yet. Fresh from NeurIPS, they debate whether models are truly plateauing or if we're just myopically focused on LLMs while breakthroughs happen in other modalities. They reveal why infinite capital at labs may actually constrain innovation, explain the narrow "Goldilocks zone" where RL actually works, and argue why U.S. chip restrictions may have backfired catastrophically—accelerating China's path to self-sufficiency by a decade. The conversation covers OpenAI's code red moment and structural vulnerabilities, the mystique surrounding SSI and Ilya's "two words," and why the real bottleneck in AI research is compute, not ideas. The episode closes with bold 2026 predictions: Rob forecasts Sam Altman won't be OpenAI's CEO by year-end, while Ari gives 50%+ odds a Chinese open-source model will be the world's best at least once next year.

People in this episode

Host: Jacob Effron

Guests: Ari Morcos, Rob Toews

Topics covered

• AI research
• LLMs
• innovation constraints
• chip restrictions
• self-sufficiency
• predictions

Keywords

• AI
• research bottleneck
• compute
• innovation
• chip restrictions
• predictions
• NeurIPS

Mentioned in this episode

Organizations: OpenAI, SSI

Places: U.S., China

More episodes of Unsupervised Learning with Jacob Effron

• Ep 85: Has AI Infra Stabilized, FM Vibe Shift, & What's Next for Coding Agents · April 23, 2026 · 55 min
• Ep 84: OpenAI’s Chief Scientist on Continual Learning Hype, RL Beyond Code, & Future Alignment Directions · April 9, 2026 · 59 min
• Ep 83: Owning the System of Record, AI-Native Org Charts, & Why ITSM is The Most Vulnerable Legacy Category · April 2, 2026 · 54 min
• Ep 82: Behind Legora's $550M Raise, Model Competition, Doubling Revenue Every Quarter, & US Expansion · March 11, 2026 · 54 min
• Ep 81: Ex-OpenAI Researcher On Why He Left, His Honest AGI Timeline, & The Limits of Scaling RL · January 29, 2026 · 1h 3m
• Ep 80: CEO of Surge AI Edwin Chen on Why Frontier Labs Are Diverging, RL Environments & Developing Model Taste · December 15, 2025 · 48 min

Explore listener stats, chart rankings, contacts and more on the Unsupervised Learning with Jacob Effron podcast page.