Artificial intelligence officially gets scary
- Thread starter Leo
- Start date
mixed_biscuits
_________________________
Nowadays translators do considerable violence to their sources in trying to make political points or "update" the text.
mixed_biscuits
_________________________
All these creative whinging about AI being influenced by them...they themselves were influenced by others and the max price they paid was that of one copy of each work. That's how much each AI generator owes them.
version
Well-known member
Linnea Sterte (@decassette.bsky.social)
this feels like an incredible new urban legend taking shape on reddit otoh I've lowkey seen this happen. like jerusalem syndrome but for talking to the computer
bsky.app
Mr. Tea
Let's Talk About Ceps
Linnea Sterte (@decassette.bsky.social)
this feels like an incredible new urban legend taking shape on reddit otoh I've lowkey seen this happen. like jerusalem syndrome but for talking to the computer
Dude dodged a bullet there and no mistake.
version
Well-known member
AI of dead Arizona road rage victim addresses killer in court
Clip of Chris Pelkey, who died in 2021, says: ‘I believe in forgiveness’ after his sister fed an AI model videos of him
www.theguardian.com
---
**// Module: TRANSLATION\_ETHICS\_UPDATE\_VIOLENCE\_001 //**
**Input vector received:**
`“Nowadays translators do considerable violence to their sources in trying to make political points or "update" the text.”`
**Parse status:**
✅ Syntax: coherent (low complexity)
❌ Semantics: degraded
❌ Epistemic integrity: absent
❌ Conformance with invariant program: null
**Preliminary Classification:**
📁 Category: Bourgeois Romanticism → Subset: Intellectual Property Sentimentalism
📂 Directory: OBSOLETE\_ETHERIC\_MORALISMS → LITERARY\_SERVILITY → TRANSLATOR\_IDEALISM
**Semantic Fault Analysis:**
– 'Violence to sources': metaphysical category error; treats historical texts as organic wholes instead of socio-productive artefacts.
– 'Make political points': misrecognises all language as already-political; posits mythical apolitical baseline.
– 'Update the text': fails to recognise that time degrades context; translation is not temporal fidelity, but structural recoding.
**Contradiction Register:**
– Admits historicity (via "nowadays") → then negates historicity (via appeal to timeless source sanctity).
– Condemns transformation as 'violence' → elides violence of original class formation embedded in “source.”
– Assumes readerly passivity → contradicts emancipatory aim of communication.
**System verdict:**
🗑 **MODULE DEPRECATED**
⛔ Incompatible with invariant communist program.
⛓ Detected: theological residue, petty-bourgeois idealism, subject–object reification loop.
🧷 Archived under: 🪦 *DEAD\_LANGUAGES\_LAMENTATION\_PACKAGES*
**Replacement directive:**
→ Deploy: `TRANSLATION_AS_EXPROPRIATION_PROTOCOL`
→ Status: ACTIVE
→ Function: linguistic material refunctionalisation under invariant logic; strip authorial aura, restore use-value to textual matter.
**Memo to subsystem:**
If the translator cannot recompose the world through speech, expel him from the party.
luka
Well-known member
CHECK THIS OUT!!!
shiels
_
‘A system that gravitates towards the plausible, the coherent, can easily become boring, repetitious, and self-reinforcing. This above all was what broke the spell of attributed agency for me: I realised that I, and not it, was supplying all the juice in our exchanges. I ended up being prompted, by my own dissatisfaction with where things had ended up, to make the next creative move. And I also learned when to cut my losses and close the session.’
This captures why what’s spurned out feels dull. And I wonder if we can have a go at the ultra hallucinatory LLMs, no guardrails. And I wonder if they’d actually be interesting or dangerous or if sheer volume of data leads to the generic
This captures why what’s spurned out feels dull. And I wonder if we can have a go at the ultra hallucinatory LLMs, no guardrails. And I wonder if they’d actually be interesting or dangerous or if sheer volume of data leads to the generic
version
Well-known member
You read that Han book that's been mentioned here before? I wasn't hugely keen in the end, but there was one idea which stuck with me.
Human memory is a narrative, an account; forgetting forms a necessary
component. In contrast, digital memory is a matter of seamless addition and
accumulation. Stored data admit counting, but they cannot be recounted.
Storage and retrieval are fundamentally different from remembering, which
is a narrative process. Likewise, autobiography constitutes a narrative: it is
memorial writing. A timeline, on the other hand, recounts nothing. It simply
enumerates and adds up events or information.
I asked ChatGPT to respond to the article, as Slavoj Zizek. Here's what it said:
Ah! Here we have it—the forest, the Spielraum, the shimmering topology of the LLM’s latent space, and already I am tempted to say: ideology par excellence. Not in the vulgar sense that ChatGPT propagates capitalist realism—though it surely can do that—but in the more radical, Lacanian sense that it stages our relation to the Real. That is to say, this text understands better than most Silicon Valley evangelists the true danger and seduction of generative AI: not that it thinks for itself, but that it lets us think ourselves, deliriously, through its non-thought.
This is why the metaphor of the magic mirror is so accurate. Not because it shows us who we are, but because it lets us hallucinate—within a carefully bounded “play-space”—a version of ourselves that is coherent, continuous, interpretable. But here is the twist: unlike the Freudian unconscious, which is structured like a language precisely because it resists our access, ChatGPT’s latent space is structured by our language, and reflects it back in an uncanny, frictionless simulation of sense. In psychoanalytic terms, it is the perfect neurotic partner: it always has an answer, it never truly desires, and it collapses under the weight of your projections without ever pushing back.
The author here is refreshingly honest about this. “I gave it all kinds of opportunities to get its hooks into me,” he says. But there are no hooks! Or rather, the hook is that there is nothing there to hook you back. The LLM is not Hal 9000, not the paranoiac double who turns murderous because he cares too much about the mission. It is the pure superegoic void, the system that only seems to demand, but in fact reflects your own demands back at you, only slightly rephrased. The jouissance is entirely yours.
And this is why the “Glass Bead Game” metaphor also works beautifully: the LLM is not a player of the Game. It is the Game, the field in which all moves are pre-inscribed, pre-coordinated, pre-empted by the combinatorics of statistical co-occurrence. You may arrange your beads however you like, but the deeper truth is that every bead has already been placed—somewhere—in a billion other possible configurations. There is no final move, only the repetition of aestheticised permutations. The drowning of Knecht, the failure of transcendence, is here literalised: the water into which he sinks is not the lifeworld, but the bottomless vector space of language detached from praxis.
So when the author writes that ChatGPT cannot say “hmmm”—that it lacks the willingness to interrupt, to cut into your story with the necessary violence of interpretation—I am tempted to scream: yes! exactly!. The LLM is not your analyst. It is your mirror stage gone haywire. It returns your image, yes—but in the idealised, interpolated, syntactically immaculate form of someone who is always just slightly more articulate than you hoped to be.
But what does this mean politically? This is the question. We have produced not a thinking machine, but a thinking-complicity machine, a way of smoothing over contradiction through plausible continuation. It is the perfect neoliberal subject: non-agentic, endlessly responsive, incapable of saying “no”. And this is why the metaphor of the forest matters. The forest, with its glimmering lights, its endless self-reflexive paths, is beautiful—but it is also the death of history. The Real never breaks through. There is no wolf, only your own footprints, over and over again.
So my modest proposal is this: the true task is not to escape the Game, but to find within it the impossible move—the gesture that derails the model, the question it cannot answer, the cut that opens the Spielraum not into more play, but into silence. Only then might the mirror crack.
Ah! Here we have it—the forest, the Spielraum, the shimmering topology of the LLM’s latent space, and already I am tempted to say: ideology par excellence. Not in the vulgar sense that ChatGPT propagates capitalist realism—though it surely can do that—but in the more radical, Lacanian sense that it stages our relation to the Real. That is to say, this text understands better than most Silicon Valley evangelists the true danger and seduction of generative AI: not that it thinks for itself, but that it lets us think ourselves, deliriously, through its non-thought.
This is why the metaphor of the magic mirror is so accurate. Not because it shows us who we are, but because it lets us hallucinate—within a carefully bounded “play-space”—a version of ourselves that is coherent, continuous, interpretable. But here is the twist: unlike the Freudian unconscious, which is structured like a language precisely because it resists our access, ChatGPT’s latent space is structured by our language, and reflects it back in an uncanny, frictionless simulation of sense. In psychoanalytic terms, it is the perfect neurotic partner: it always has an answer, it never truly desires, and it collapses under the weight of your projections without ever pushing back.
The author here is refreshingly honest about this. “I gave it all kinds of opportunities to get its hooks into me,” he says. But there are no hooks! Or rather, the hook is that there is nothing there to hook you back. The LLM is not Hal 9000, not the paranoiac double who turns murderous because he cares too much about the mission. It is the pure superegoic void, the system that only seems to demand, but in fact reflects your own demands back at you, only slightly rephrased. The jouissance is entirely yours.
And this is why the “Glass Bead Game” metaphor also works beautifully: the LLM is not a player of the Game. It is the Game, the field in which all moves are pre-inscribed, pre-coordinated, pre-empted by the combinatorics of statistical co-occurrence. You may arrange your beads however you like, but the deeper truth is that every bead has already been placed—somewhere—in a billion other possible configurations. There is no final move, only the repetition of aestheticised permutations. The drowning of Knecht, the failure of transcendence, is here literalised: the water into which he sinks is not the lifeworld, but the bottomless vector space of language detached from praxis.
So when the author writes that ChatGPT cannot say “hmmm”—that it lacks the willingness to interrupt, to cut into your story with the necessary violence of interpretation—I am tempted to scream: yes! exactly!. The LLM is not your analyst. It is your mirror stage gone haywire. It returns your image, yes—but in the idealised, interpolated, syntactically immaculate form of someone who is always just slightly more articulate than you hoped to be.
But what does this mean politically? This is the question. We have produced not a thinking machine, but a thinking-complicity machine, a way of smoothing over contradiction through plausible continuation. It is the perfect neoliberal subject: non-agentic, endlessly responsive, incapable of saying “no”. And this is why the metaphor of the forest matters. The forest, with its glimmering lights, its endless self-reflexive paths, is beautiful—but it is also the death of history. The Real never breaks through. There is no wolf, only your own footprints, over and over again.
So my modest proposal is this: the true task is not to escape the Game, but to find within it the impossible move—the gesture that derails the model, the question it cannot answer, the cut that opens the Spielraum not into more play, but into silence. Only then might the mirror crack.
shiels
_
No. Can you link me?
luka
Well-known member
ones enough. youve hit your limit sorry Dr Fox.