Kelly argues that historians often make the best futurists, as understanding the past provides a framework for predicting technological trajectories. He compares our current grasp of AI to early scientists' confusion about electricity, suggesting intelligence is a compound of many cognitive elements, not a single thing. Kelly advises finding a niche where you can be 'the only' rather than 'the best,' a strategy he used to build Wired's roster of star writers. He also shares his personal use of AI for creative joy, like generating a private saga with historical figures, and views AI as a mirror for self-reflection. Reflecting on past predictions, Kelly admits he was wrong about VR's slow adoption and underestimated eBay's potential, highlighting the importance of humility in forecasting.