← All episodes

AI Genius Predicts the Next 3 Years

Name: AI Genius Predicts the Next 3 Years
Uploaded: 2026-04-11T13:30:53Z
Channel: Joe Lonsdale

April 11, 2026 | 3 products mentioned

Joe Lonsdale host

Scott Wu guest

Russell Kaplan guest

Watch on YouTube ai agents software engineering automation founder strategy small business hiring and talent productivity gains ai adoption

Scott Wu and Russell Kaplan from Cognition discuss how AI capabilities are doubling every 2-3 months, enabling autonomous agents to handle increasingly complex work—and why this creates a massive advantage for founders and small business operators who adopt these tools now. The episode explores concrete shifts in how builders should approach AI integration over the next 3 years, from task delegation strategies to hiring talent, with a focus on the productivity and competitive edge early movers will capture.

Key takeaways

• AI autonomy is expanding rapidly: Models can now work autonomously for 10-20 hours of human-equivalent effort (vs. 10 seconds two years ago), meaning builders should shift from line-by-line prompting to assigning whole projects and event-driven tasks rather than real-time supervision.
• Small businesses will see an explosion in competitiveness because AI enables solo founders to access legal vetting, financial analysis, and software engineering quality previously available only to larger teams, flattening the resource playing field for new ventures.
• Interview and hiring processes need to flip: Rather than blocking AI use during technical interviews, forward-thinking companies should require candidates to use AI tools and evaluate them on product thinking, decision-making, and trade-offs—the skills AI can't replace.
• The workforce bottleneck is shifting from capability to adoption speed: Most large organizations and governments are still learning how to use AI effectively, creating a 5-10x productivity gap between "agent-forward" companies and laggards that will widen through 2028-2029.
• Recruit former founders into your team if solving an ambitious problem—high-agency people who've built before are more likely to stay engaged when the only constraint is their own ambition and they're surrounded by world-class collaborators attacking a massive problem space.
• The form factor of AI work changes with capability level: As models get stronger, the interaction model shifts from hands-on coding assistance to autonomous project completion (e.g., "improve this app's performance, run tests, and report back"), requiring builders to think in terms of outcome-level tasks rather than code-level fixes.