Dailytech

Accidental Tech Podcast Asks Why So Many People Dislike AI

AI·October 9, 2026

The 692nd episode of The Accidental Tech Podcast, titled "A Thinking Hitch," spends much of its time on a question that is hard to avoid right now: why does the public seem so unhappy with AI? The hosts organize their show notes around a Pew Research survey that measures how people feel about the technology, from concern versus excitement to risk versus reward.

The themes they pull out are less about features than about people. Respondents are asked whether AI will weaken their ability to think, form meaningful relationships, and stay self-sufficient. The survey also gets at whether people can tell AI-generated content from human-made work. The sharpest point in the notes is that heavier use among young adults appears to go along with more dislike, not less. The hosts connect that to commencement coverage from the University of Central Florida and the University of Arizona, and to an interview with a recent journalism graduate.

The episode also looks at the physical side of the AI boom. One linked item reports that Americans really do not want data centers in their backyards. The hosts bring in a Stratechery piece on so-called phantom data centers, along with an Energy Information Administration chart tracking US nuclear power from 1970 to 2023. At the other end of the spectrum sit wearable "AI friend" devices from friend.com and omi.me, which show how far companies are pushing AI into personal companionship.

Not everything was about AI. The follow-up segment covered cut, copy and paste in Finder, problems with Terminal.app restoring its state, tcsh and .cshrc configuration, and a set of terminal resources, including Julia Evans' guide to the secret rules of the terminal. On the storage side, the hosts discussed keeping node_modules out of Dropbox and Time Machine, and pnpm as an alternative to npm. The notes end with Reminders snoozing, Apple's 26.5 updates, and a rumor that the 20th anniversary iPhone will get a curved display a year late.

Reporting based on an external source.