Elektrine lite

← Feed

@aud@fire.asta.lgbt

Post #2305860

2026-03-24 00:43 UTC

@seachanger@alaskan.social Not sure about the methodology behind this one, but I've heard about it at least (re: #10): https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/

Replies (2)

  • @aud@fire.asta.lgbt 2026-03-24 00:51

    @seachanger@alaskan.social Regarding item #5: https://www.npr.org/2025/09/05/nx-s1-5529404/anthropic-settlement-authors-copyright-ai It's important to note, though, that the ruling walks a fine line: training of Claude was considered to be "fair use" (not a ruling I personally agree with but hey), however, the fact that Anthropic pirated all the materials was not. Anthropic settled on this claim rather than take it to trial, it seems.

    Open ##2305861

  • @cthos@mastodon.cthos.dev 2026-03-24 01:11

    @aud@fire.asta.lgbt @seachanger@alaskan.social That's about the only actual study we have and it has a fairly low sample size, unfortunately. There are some other articles going around about the high cost and failure rates of AI projects though. Methodology-wise, it's okay and at least tries to control for perception vs reality.

    Open ##2305865