Re:Show HN: Is AI hijacking your intent? A formal control al...
by stuartjohnson12 on Sunday January 11, 2026 @09:50AM
(#46576189)
Since OpenAI patched the LLM spiritual awakening attractor state, physics and computer science is what sycophantic AI is pushing people towards now. My theory is that those things tend to be especially optimised for deceit because they involve modelling and many people can become confused between the difference between a model as the expression of a concept and a model as in the colloquial idea of "the way the universe works".
[ Parent
| on HN ]
Re:Show HN: Is AI hijacking your intent? A formal control al...
by cap11235 on Sunday January 11, 2026 @09:53AM
(#46576217)
I'd love to see a new cult form around UML. Unified Modeling Language already sounds LLMy.
Re:Show HN: Is AI hijacking your intent? A formal control al...
by daikikadowaki on Sunday January 11, 2026 @10:08AM
(#46576351)
'The Church of UML' does have a certain ring to it. But that’s exactly the trap I’m trying to avoid.
The reason JTP focuses on the 'Ghost'—the traces of what the model rejected or what was lost in translation—is to prevent exactly that kind of cult-like devotion to the output. A cult forms when you forget the model is just a map.
I’m not interested in worshipping the map; I’m interested in ensuring that when the machine draws it, we can still see the ink on our own hands. If we can't see the delegation, we can't see the deceit. That’s the 'sovereignty' part of Cognitive Sovereignty.