Elektrine lite

← Feed

@glyph@mastodon.social

2026-09-26 18:06 UTC

@eevee@slime.global @zzzeek@fosstodon.org @hynek@mastodon.social I've also explained, both in that article, in general posts, and to you personally, why people *massively over-perceive the benefits and under-perceive the costs*. it's too early to have any data on this but neurodivergence is theoretically a huge risk factor for AI psychosis

Replies (3)

  • @eevee@slime.global 2026-09-26 18:12

    @glyph@mastodon.social @zzzeek@fosstodon.org @hynek@mastodon.social and even your post hinges on the charitable assumption that the end user is actually verifying the output in some way. at least for code, there's *some* way to do this: you can run it. but the sales pitch is "phew, the computer did it for me" and that's the sort of thing that haunts me. it seems unlikely that ari drennen is checking that the automated tax returns or summaries of their email are *accurate*, because that would involve manually doing exactly the same work

    Open ##4860433

  • @glyph@mastodon.social 2026-09-26 18:09

    @eevee@slime.global @zzzeek@fosstodon.org @hynek@mastodon.social is a *perceived*, self-reported improvement to productivity worth the risk of an emotional dysregulation spiral? the null hypothesis at this point is that these people are deluding themselves. I've seen a thousand people claiming that it's helping with so many things, and yet zero times someone observed, say, their partner or their child adopting AI and managing their disability better as a result

    Open ##5274381

  • @zzzeek@fosstodon.org 2026-09-26 18:33

    @glyph@mastodon.social @eevee@slime.global @hynek@mastodon.social yes it's this "no you're just 'over perceiving' your benefits" thing that I think is bullshit frankly, at least for me

    Open ##5274382