Post #3739641
2026-06-21 10:31 UTC
@pozorvlak@mathstodon.xyz @Migueldeicaza@mastodon.social @matt@toot.cafe at some point we will go 3 months without a front-page story about a hundred-billion dollar company making an excruciatingly embarrassing mistake in production with their “good enough now” models and I will *start* to be curious about whether things have actually changed on the efficacy front. I have wasted an enormously frustrating amount of time on attempting to re-validate the error potential already.
Replies (1)
-
@glyph@mastodon.social 2026-06-21 10:31
@pozorvlak@mathstodon.xyz @Migueldeicaza@mastodon.social @matt@toot.cafe However, to reiterate and clarify what I have said above, I am still not interested in *using* these things even if they do one day become extremely effective because the *mechanism* is plagiarism, and thus, both unethical and unsustainable. It is a blessing that they are still, in aggregate, extremely ineffective, and we still might have a chance to fight their adoption on that basis even for those who have no moral compass.