Today's episode covers Meta's AI model delays, industry consolidation moves, and the growing debate around AI's impact on employment. The show features commentary on Meta's delayed 'Avocado' model, XAI's leadership exodus, Cursor's massive valuation increase, and medical AI adoption statistics.

The main segment explores competing perspectives on AI-driven job displacement, examining recent research from Anthropic's Labor Market Impacts of AI and MIT's Building Pro-Worker Artificial Intelligence. The discussion includes policy proposals from former Commerce Secretary Gina Raimondo and evidence suggesting AI may create more jobs than it eliminates, challenging prevailing pessimistic narratives about technological unemployment.

Meta's Model Delays Signal Frontier AI Competition Tightening

Meta's 'Avocado' model has been delayed until at least May after falling short of rivals in reasoning, coding, and writing benchmarks, despite outperforming Gemini 2.5 but not matching Gemini 3.

The model has been in development for nine months while performance goalposts shifted dramatically, with Meta leadership reportedly considering licensing Gemini as a stopgap solution.

Ethan Mollick tweeted that 'Both XAI and Meta seem to be falling behind' and 'Frontier AI models are really a three-way race at this point.'

XAI Leadership Exodus Continues as Coding Push Intensifies

XAI hired senior Cursor leaders Andrew Millich and Jason Ginsberg to report directly to Elon Musk as the company acknowledges falling behind on coding capabilities.

Two more co-founders departed this week - Zhi Hang Dai and Gorong Zhang - bringing total departures to six of twelve original co-founders this year.

Musk characterized the changes as controlled demolition, tweeting 'XAI was not built right the first time around, so is being rebuilt from the foundations up. Same thing happened with Tesla.'

AI Adoption Accelerates in Medicine and Enterprise Markets

American Medical Association survey found 81% of doctors now use AI professionally, more than doubling since 2023, with leading use cases being research summaries and administrative tasks.

Only 17% of doctors use AI for assistive diagnosis, while most applications focus on 'augmented intelligence' rather than replacing human judgment.

Anthropic is reportedly in talks with Blackstone and other PE firms to launch an AI consulting venture, though Pentagon conflicts have put discussions on hold.

Job Displacement Debate Intensifies Amid Mixed Evidence

Atlassian cut 1,600 jobs (10% of workforce) explicitly citing AI, with CEO Mike Cannon-Brooks stating 'AI doesn't change the mix of skills we need or the number of roles required in certain areas. It does.'

Anthropic's Labor Market Impacts of AI introduces 'observed exposure' metrics showing management and finance roles have 90%+ theoretical AI coverage but minimal current automation.

European Central Bank study of 5,000 Eurozone firms found AI-intensive companies are 4% more likely to hire new staff, contradicting job displacement fears.

63% of Americans believe AI will decrease job availability while only 7% predict job increases, showing more pessimism than Chinese respondents at 40%.

Policy Frameworks Emerge for Managing AI Transition

Former Commerce Secretary Gina Raimondo proposes a 'new grand bargain' requiring employers to define AI economy skills while government invests in training and safety nets.

Raimondo advocates for modular higher education with 'short, affordable job-linked credits' and employer tax incentives tied to on-the-job training and worker retention.

MIT's Building Pro-Worker Artificial Intelligence categorizes technological change into five types, arguing only 'new task-creating technologies' are unambiguously pro-worker.

The MIT paper argues current AI focuses overwhelmingly on automation rather than collaboration, citing misaligned incentives and the 'AGI bet' where firms see little point in worker enhancement tools.