What if AI is the end of something old? Andrew Curry makes this case, drawing on Carlota Perez's model of 50-year technology surges and Nicolas Colin's 'late cycle investment theory.' AI represents an efficiency play, squeezing the last gains out of a computing paradigm that's running out of room to grow. Truss CTO Ken Kantzer warns that AI solutions often lack the necessary 'taste' to be truly efficient, frequently over-engineering systems rather than maximizing value.
Colin points to three signals. The 2022 startup funding collapse looks structural. The biggest AI breakthrough came from OpenAI, backed by Microsoft's massive compute budget. This was an incumbent's game from the start. Google, Meta, Amazon all responded with billions. That's late-cycle deployment: incumbents throwing capital at well-understood problems. Microsoft's Copilot brand sprawl is just one example of how platform saturation is nearly complete. The sectors where computing and networks can make a real difference are mostly transformed. What remains might simply be the natural limits of what this technology can do.
History has seen this pattern before. The Corliss steam engine, shown at the 1876 Centennial Exposition, burned fuel with impressive efficiency. It also arrived at the tail end of the steam age, just as electricity and steel were about to take over. Henry Ford's moving assembly line in 1913 didn't invent a new paradigm. It squeezed maximum productivity out of the existing one right before market saturation hit. Both breakthroughs arrived at maturity. The party was already winding down.
The US and China are approaching AI differently, and the contrast matters. America chases AGI with ever-bigger models. China focuses on practical industrial applications. Then there's the growing pushback: communities fighting data centers across the country, users frustrated they can't turn off Google's AI search features. When people start resenting a technology's presence, you're witnessing the final act. The frustration is telling. People don't fight technology that's still exciting them.