Benedict Evans: AI Platform Shift or Something Bigger?

How Evans Separates AI Reality From Hype

Benedict Evans has spent years as one of tech's most incisive analysts, and this a16z podcast appearance demonstrates why. He refuses to get swept up in either AI doomerism or utopianism, instead applying the same frameworks he's used to analyze mobile, cloud, and social.

His opening provocation is brilliant: "ChatGPT has got 800 or 900 million weekly active users. And if you're the kind of person who is using this for hours every day, ask yourself why five times more people look at it, get it, know what it is, have an account, know how to use it, and can't think of anything to do with it this week or next week." This isn't dismissive - it's genuinely curious about the gap between power users and everyone else.

Evans contextualizes AI terminology with typical precision: "The term AI is a little bit like the term technology. When something's been around any for a while, it's not AI anymore. Is machine learning still AI? I don't know. In actual general usage, AI seems to mean new stuff. And AGI seems new scary stuff."

On AGI specifically, he's delightfully skeptical: "Either it's already here and it's just small software or it's 5 years away and will always be 5 years away." This echoes the long history of AI winters and summers.

The platform shift question is the core of the discussion. Evans has studied how previous shifts (PC, internet, mobile, cloud) played out - some industries were completely transformed, others barely noticed. His framework helps separate which patterns from previous platform shifts will repeat and which are unique to AI.

On bubbles, he's pragmatically philosophical: "If we're not in a bubble now, we will be." The question isn't whether there will be excess - it's what useful infrastructure gets built during the mania.

5 Insights From Evans on AI Platform Shifts

• The gap between AI power users and everyone else is the underexplored story - 900M ChatGPT users, but most can't think of what to do with it
• "AI" means "new stuff" in practice; once technology works reliably, people stop calling it AI
• Previous platform shifts (mobile, cloud, internet) had industries that transformed completely and others barely affected - AI will likely follow this pattern
• Bubbles are probably inevitable, but the question is what useful infrastructure gets built during the mania
• We don't know the physical limits of this technology, which makes forecasting unusually difficult