Tesmaia 🏳️‍⚧️
banner
tesmaia.nl
Tesmaia 🏳️‍⚧️
@tesmaia.nl
She/Her ❤️💚
Software dev, hobby gamedev, I like bouldering

Making a swimmy metroidvania
Website: https://tesmaia.nl
I think the conflation comes both ways. The makers of “the con” want it broad so they can claim it’s more than it really is, while people instructed to do “the new big thing” want it broad so they don’t actually need to change anything (just rebrand whatever we’re already doing as “the thing”)
December 24, 2025 at 8:11 PM
Zero tolerance also makes it way too easy to damage a company’s reputation. If not even corrective measures down the line are good enough, there is even less incentive to be transparent about AI use, while people can just lie about a company that did actually achieve an AI-free result.
December 24, 2025 at 2:39 PM
This is why I don’t agree with zero tolerance. It will contaminate any business of a medium scale even if there is a policy to not use AI, there will be someone that uses it and hides the fact. Either as employee, as contractor, or as part of things further down the supply chain.
December 24, 2025 at 2:34 PM
Courts need to deny any attempt of tech-faked necromancy at any point in proceedings. Wtf.
December 24, 2025 at 6:58 AM
Huh?
December 21, 2025 at 10:58 PM
In particular, you can launder copyright infringement without consequences. Devs would have loved to be able to steal assets from other games if that was accepted and not legally dangerous, just that alone will boost productivity.
December 21, 2025 at 10:50 PM
Wanneer kan de eeuwige andersomdag eindigen?
December 21, 2025 at 7:23 PM
Why would you use a text forecast to classify things? Classification systems are not LLMs.
You can put the output of a classification system through a bullshit generator for lolz, but that’s not useful.
December 20, 2025 at 3:00 PM
Systemen en prikkels verslaan altijd de intentie, dus steek je hand niet in een extreemrechtse cactus anders blijft het voor altijd aan je plakken.
December 19, 2025 at 4:27 PM
Zucht en ughh is inderdaad het sentiment dat hierbij hoort.
December 19, 2025 at 4:24 PM
Confederal is more accurate, but there are some federal institutions.
December 19, 2025 at 3:24 PM
Juist door deze motie niet te steunen heeft d66 straks terecht de schuld van alles wat er mis gaat door Wiersma’s beleid.
Oh je wil het toch weer aanpassen, dat is wispelturig beleid, foei. Maar je moet het wel oplossen he. Hoe zit het met je 100k woningen garantie btw?
December 18, 2025 at 10:03 PM
Zelfs de overheid heeft schijt aan de overheid.
December 18, 2025 at 9:24 PM
That’s what they need the storytellers for. The bots can’t do the work but they need people to believe they can.
December 18, 2025 at 7:31 PM
What did I just read??
December 18, 2025 at 6:58 PM
Any placeholder that’s not obviously ugly should be assumed to be ready for release. They’re fine with it ending in the final product.
December 18, 2025 at 5:02 PM
It’s also not a summary, but a prediction of what a summary could look like.

LLMs extrapolate text, the output is a prediction.
December 17, 2025 at 5:05 PM
Next they’ll do it the other way around and give you predictions of a paper based on the abstract. And that’d be cited more than the real thing because the prediction won’t be behind the paywall…
December 16, 2025 at 10:07 PM
Instead of looking outside the window, when you’re right next to a window, to see if it’s raining, you choose the weather forecast of an hour ago because it’s The Future.
December 16, 2025 at 10:03 PM
Volgens mij gebruik je gewoon de ov chipkaart. Maar het liefst zou ik het voor iedereen gratis zien.
December 15, 2025 at 10:49 PM
Ok, so now I'm interested in the carpenter bee lore.
December 15, 2025 at 9:05 PM
Persoonlijk? Die schrijfstijl doet me wel heel erg denken dat het AI slop is.
December 15, 2025 at 4:05 PM
Al die info gaat vast rechtstreeks de Palantír database in. En met DNA geef je ook biometrische gegevens van je familie door, nooit dat vergeten.
December 14, 2025 at 10:24 PM