Financial Times says stop talking about AI — is the discourse weaker than the tech?

Summary

The Financial Times published a piece arguing that public discourse on AI has become intellectually weak relative to the technology's actual significance.

The piece opens by comparing income growth across millennia. Living standards stagnated from 1000 BC through the late 1700s, then shot upward with industrialization. The takeaway: be skeptical when someone likens AI to the Industrial Revolution. At best, AI might match the telephone or incandescent lightbulb—transformative, but not civilization-altering.

But the argument then reverses. Before recorded music, hearing your favorite piece meant waiting for an orchestra to pass through town. Before air travel, crossing an ocean was a weeks-long ordeal. AI could deliver similar quality-of-life improvements. The author acknowledges the weakness in this position: it relies too heavily on historical precedent, lacks technical grounding, and falls into the trap of all AI skepticism by offering no actionable guidance for investors or policymakers.

The core criticism is structural. The discourse is trapped between two camps with inherent conflicts of interest. The most bullish voices often work in AI or hold significant equity stakes, creating strong incentive to talk the technology up. Skeptics resort to weak anecdotes or tautologies. Neither side illuminates what the technology actually does or will do.

One useful observation: the AI debate reveals temperament more than truth. Those predicting seismic disaster tend to be highly strung already. Optimists are naturally chipper and credulous. Skeptics are complacent.

On the practical question of how much worse your life would be without AI tools, the assessment lands in single digits. AI improves daily life by perhaps 3 to 5 percent. That roughly aligns with the trillion dollars added to market cap in markets worth hundreds of trillions. The technology appears correctly priced, not overhyped.

One wrinkle remains: AI may show minimal GDP impact while simultaneously destroying jobs. If that occurs, skeptics and alarmists would both be right. The technology would lack sweeping economic benefit but generate genuine social friction. The discourse offers no framework for navigating that outcome.

Generative AI tools still feel like incremental SaaS improvements or toys. Video game-playing bots once looked similarly toy-like. The question is not whether the toy phase ends, but how quickly. Despite a trillion dollars in capital, we are still somewhat in it.