Phrase of the moment: "AI fatigue"

Right when the big US AI companies set out to monetize chatbots with ads and other schemes, the public is getting fed up with AI.

My annual subscription to Perplexity came up for renewal yesterday, and I cancelled it. To me, AI chatbots are feeling commodified, and also problematic. (I’ve recently been using Kagi Search with the “?” modifier, which gives you an instant, concise AI-created summary of the search results. That combination of great search plus concise summary feels vastly superior to me than Perplexity’s long-winded, often erroneous results. (Disclosure: My son, Kevin, works for Kagi.)

I’m not alone, apparently. The Economist this week reported on US government data saying that the percentage of Americans using AI for work at big companies is hovering at a flaccid 11%, which is down one point from the previous month.

The reason may be found in other data: People don’t trust AI. That’s why, according to Pew, so few Americans turn to AI for news.

A different survey found that 39% of Americans verify AI-generated information using search engines because they don’t trust LLMs.

The bottom line is that the general hype around AI chatbots does not match the general estimation in the public about them.