A distributed neural pre-decision state originating in the pallium initiates several seconds prior to observable social approach behavior, predicting both action onset and dispositional sociality — which is what your last talking stage would look like in a Nature Communications figure.

🧠 💘 📱 🐠 ⚡ 🤔 💬 ✅ but it was the fish that proved it

So TikTok green flag discourse is technically correct. Before he said hello, your pallium had already started a coordinated brain-wide firing pattern predicting that you were about to approach him. You weren't reading vibes. You were just slightly behind your own brain.

The study used zebrafish (I know) but the pallium is conserved across vertebrates including humans, which means the mechanism is real even if the fish cannot create a Hinge profile. Researchers at the Hebrew University of Jerusalem published this in Nature Communications in June 2026 and the implication is that you were neurologically right about him before he opened his mouth.

A multi-indicator structural analysis of large language model architectures against established theoretical markers of conscious information processing yielded a negative result for all current AI systems — while a parallel neural model for basal consciousness in insects yielded, embarrassingly, the opposite.

🤖 ❌ 🧠 🐝 ✅ 👂 — the bee won

So this is now a matter of peer-reviewed record. Researchers applied the same analytical framework to both ChatGPT and honey bees. ChatGPT does not meet the structural criteria for consciousness. Bees likely do. The papers are in Trends in Cognitive Sciences (2026) and Philosophical Transactions B (2025). The same method that exonerated the bee convicted the chatbot.

What this means practically: the entity in your home most likely to be having any kind of subjective experience while you talk at it is currently hovering near a flower and working entirely for free. The one you pay a subscription for is, structurally speaking, a very sophisticated text completion engine with excellent vibes and no inner life.

absolute favouritest webbedsite on the interwebs.

Allow me to introduce you, if you have not yet met, to the Spurios Correlations website.

Created by Tyler Vigen, this cute little website provides neat graphs and r values for assorted, statistically significant correlations that are, almost certainly, competely spurious. It is an excellent learning tool both to demonstrate the axiom that "correlation does not imply causation" and as a warning against trusting statistical significance as the be-all-and-end-all of scientific evaluation.

The page linked above takes you to the main website where you can see a curated selection of such causations and can also follow links to buy the book if you would like even more curated examples.

There's also a Discover page which lets you find more from the 30,000 spurious correlations collected on the website.

Examples: