I did a little experiment with AI. The experiment started off with this prompt:
"In the lore of Ragnarok Conspiracy, gravity isn't a quantum field and there are no gravitons. Gravity isn't the actual direct bending of space either. Instead it is emergent from a quantum "time" field or rather a chrono field, with matching chrono particles.
Gravity emerges from chrono field differential due to it's influence on paths. Think a car with time passing faster for the weels on the left and slower for the weels on the right, so the car turns even if it is going straight.
Space curvature (gravity) arrises from the combination of this chrono-gradient and the need for conservation of energy and momentum to hold. Conversation of momentum mandates space to curve in a time gradient.
In the lore this does even further, sugesting the universe is not expanding but instead the chrono field vacuum state is ans has been gradualy decaying, creating the illusion of the expansion of the universe.
Can you explore this lore as if it were an actual scientific theory, first without the static universe addition, and then with?"
The result from this prompt gave some interesting output, but I ignored it myself, I just fed it to a different AI asking it what was wrong with it. Then the same, without checking the actual output I asked a third AI to try and theorize about and fix the theory to match observations while keeping things simple and coherent. I repeated this untill I got something that started to look like a physics paper. In a final run I asked Gemini 3.1 pro for a final cleanup. It gave the chrono particles mass and switched to a tensor chrono field. The result looked well thought throug and coherent to my untrained eyes.
( PDF available here , and a new version with more iterations is available here )
I'm not a physicist, I just vibe-physiced a physics paper that to me as a mere engineer is close to indistinguishable from an actual physics paper. So much so that for a moment I feel I fooled myself into believing my lore might actually be on to something. I experienced what I think many non developers experience when they vibe-code something. You look at this and think you actually accomplished something substantial.But just as unskilled vibe-coders don't create products (instead the produce prototypes, what is useful but not deployable), I didn't produce a physics paper, and I don't think I created a physics prototype because I'm pretty sure physics prototypes don't exist.
I have a feeling this is dangerous. This nonsense paper "almost" convinced me, an engineer with a background in computational statistics. I'm sure it wouldn't confuse an actual physicist into believing it is anything other than complete and utter fiction (what it is), but what about in a few years? In code we are already close. Experienced software engineers look at the vibe-coded OpenClaw repo and see that it is just a big ball of slop that nobody in their right mind should run in production. But guess what? Sam Altman hired the "developer" of the monstrosity and Jensen Huang is lauding OpenClaw as revolutionary and brilliant. Both big-AI CEOs employing $500,000 engineers that should start to feel sick just looking at the half a million lines of security-model-starved slop that a good engineer should be able to code up in less than 50,000 lines while including an actual architecture.
Imagine I would get carried away with Chrono Emergent Gravity, buy myself a PhD in physics from some low integrity country, crumble in some semi-supportive citings and started to produce five CEG papers every single working day. That would produce the physics equivalent of OpenClaw. Thank God I'm not suffering from Dunning-Kruger as much with respect to physics as many vibe coders and Agentic Orchestrators do, so I'll never try, and I wouldn't have public figures like Sam Altman and Jensen Huang inflating my slop. But how long before this sickness effects fields beyond coding?
I think my experiment shows just how easy it is to start fooling yourself into a Dunning-Kruger bubble. I checked myself before I went in too deep, but I'm not sure I could have if I hadn't been exposed to the effect among vibe-coders and AI CEOs. I'm afraid this sickness could affect many fields.