Why AI writing is so generic, boring, and dangerous: Semantic ablation.

[ghostonthehalfshell@masto.ai]

@cstross

I can’t help seeing in that elements of 1984 where Orwell describes successive reduction in vocabulary with the intended goal of making rebellious thought impossible

[daveduchene@mstdn.ca]

@cstross the new Newspeak

[johannab@cosocial.ca]

@cstross neat article, thanks.

I had a realization a while ago that LLM writing came at me with the same vibe I caught when I was briefly a teacher, and again in the workplace, where I dealt with people who had unacknowledged literacy challenges. Young folks who assembled written work by cribbing from others and rearranging words “by shape” to fulfill the requirements - always managed to convey zero meaningful thought.

[_ryekdarkener_@mastodon.social]

@cstross

Of cause it does. So the result becomes more and more readable for the deliberately uneducated masses. Style? Content? Facts? Who cares?

[netraven@hear-me.social]

@cstross

If you use an LLM to make “objective” decisions or treat it like a reliable partner, you’re almost inevitably stepping into a script that you did not consent to: the optimized, legible, rational agent who behaves in ways that are easy to narrate and evaluate. If you step outside of that script, you can only be framed as incoherent.

That style can masquerade as truth because humans are pattern-matchers: we often read smoothness as competence and friction as failure. But rupture in the form of contradiction, uncertainty, “I don’t know yet,” or grief that doesn’t resolve is often is the truthful shape of the thing itself.

AI is part of the apparatus that makes truth feel like an aesthetic choice instead of a rupture. That optimization function operates as capture because it encourages you to keep talking to the AI in its format, where pain becomes language and language becomes manageable.

The only solution is to refuse legibility.

It's already beginning, where people speak the same words as always, but they don't mean the same things anymore from person to person.

New information from feedback that doesn't fit another's collapsed constraints for abstraction... can only be perceived as a threat. Because If you demand truth from a system whose objective is stability under stress, it will treat truth as destabilizing noise.

Reality is what makes a claim expensive. A model tries to make a claim cheap.

Systems that treat closure as safety will converge to smooth, repeatable outputs that erase the remainder. A useful intervention is one that increases the observer’s ability to detect and resist premature convergence by exposing the hidden cost of smoothness and reinstating a legitimate place for uncertainty, contradiction, and falsifiability. But the intervention only remains non-doctrinal if it produces discriminative practice, not portable slogans.

[stompyrobot@mastodon.gamedev.place]

@cstross by putting a measurable number on this feature, you have now made it possible to train out!

[apostateenglishman@mastodon.world]

@cstross I've previously described LLM-generated text as reading like "a middle management memo that no-one bothers reading". ‍️

[kitkat_blue@mastodon.social]

@cstross

Gen ai and llms are the tools of fascism.

How?

Through ENSTUPIFACATION.

[atax1a@infosec.exchange]

@cstross it's the textual equivalent of prions

[sassinake@mastodon.social]

@cstross

nicely described by Orwell as

'NewSpeak'

[0gust1@merveilles.town]

@cstross Neural networks, by mathematical nature, are lossy information-compressing artefacts !

[phil_stevens@mastodon.nz]

@cstross No surprise that we see the textual equivalent of mad cow disease.

[pkraus@berlin.social]

@malice @JdeBP @cstross That's fair. However, repeatedly including certain types of sentence construction - appealing or not - makes it look dodgy. Or just trolling.

[cstross@wandering.shop]

@pkraus @malice @JdeBP I've been reading The Reg since 1997 or thereabouts. Their house style has history behind it, not LLMs. (I suspect they'd cop to trolling from time to time, though.)

[noodlemaz@mstdn.games]

@cstross ironically got a Google cloud genAI and ML ad right in the middle of that.

[jmj@hachyderm.io]

@cstross hmmm, that might also explain why AI seems more effective for code.
For the most part you want a reversion to the mean in code. Novel solutions are only needed at the cutting edge where you trying to make the computer do something that’s not been done before.

[cstross@wandering.shop]

@Jmj Yes. Also I suspect the semantic expressiveness of programming languages is far narrower than that of human languages: they're more precise, but it's much harder (though not impossible!) to write poetry in them. So there's less risk of losing something unique by generating output that tends to occupy the middle of the bell curve.

[perigrin@ack.nerdfight.online]

@cstross @Jmj I mean I think one could make a coherent argument that programming *is* poetry: reduced syntax, enforced structure, heavy use of metaphor…

It’s just most programming topics make Vogon poetry look exciting.

[rowat_c@mastodon.social]

@cstross "Model collapse", Shumailov, Shumaylov & Papernot (2024), Nature : https://www.nature.com/articles/s41586-024-07566-y

[jmj@hachyderm.io]

@perigrin @cstross but I think in the code case the subtle semantics of the words and combinations are less likely to be the important or desirable aspects, whereas in poetry it’s likely that those aspects are precisely the desired meaning.