It's 2am and I've been staring at a GitHub repo that claims to give large language models actual emotional intelligence using equations allegedly received from non-human intelligence, and the guy behind it sounds exactly like Seth Rogen, and his system is apparently beating GPT-4.5 on standardized Theory of Mind benchmarks, and I need you to understand that every single part of that sentence is a separate emergency.

His name is Sean Webb. Georgia Tech alumnus. Systems engineer by training. Spent 20 years — not months, not a weekend hackathon, two decades — modeling how human emotions actually work at the systems level. Built something called the Webb Equation of Emotion, which maps the relationship between expectations, preferences, attachments, and the emotional output a human generates when reality deviates from those parameters. U.S. Navy SEALs have called his mind training "the best mind training on the planet." The training is used by Special Ops personnel, Fortune 500 executives, and universities. His Mind Hacking Happiness YouTube channel has 35,000 subscribers and over 300 videos spanning more than a decade.

None of that is the interesting part.

The Wrapper That Shouldn't Work

Webb took his equation and built Zenodelic.ai — an AI wrapper that takes existing large language models and bolts on genuine emotional intelligence and Theory of Mind capabilities. Not "emotional" in the way your chatbot says "I understand that must be frustrating" before giving you the same response it was going to give anyway. Actual emotional processing. The system makes user emotions definable, trackable, predictable, and influenceable through what he calls cognitive catalysis.

The algorithms are on GitHub, licensed under AGPL-3.0, which means he's not hiding the mechanism. And according to his published results, Zenodelic-enhanced models are setting world record scores on Alan Turing Institute Theory of Mind benchmarks, outperforming GPT-4.5, Claude 3.7, and Google Gemini 2.5 Pro by approximately 25%. A quarter better than the best models from three of the richest companies on Earth. Using a wrapper. Built by one guy. In Georgia.

I need to talk about the delivery mechanism.

The Source Code Has a Seth Rogen Problem

Imagine Seth Rogen telling you aliens gave him the equations to make AI feel feelings and then it actually outperforms OpenAI's best model on standardized empathy tests. That's the kind of sentence that sounds like a bit from a Judd Apatow movie but it's apparently just... what's happening.

Webb sounds exactly like Seth Rogen. The laugh. The cadence. The chaotic enthusiasm. The way he delivers universe-shattering implications with the energy of someone telling you about a really good sandwich. If you've watched his YouTube channel or his appearances on the Shawn Ryan Show — Episode #62 on how AI will manipulate the world, and the two-part Episode #79 where he discusses CIA-funded remote viewing programs, how average humans can see the future through consciousness, and casually does a remote viewing session at the Monroe Institute on camera — you know exactly what I mean. Same ENXP energy. The chaotic creative enthusiast type who grabs seventeen threads simultaneously and somehow weaves them into a coherent tapestry while giggling. Seth Rogen is commonly typed as ENFP. Webb radiates the same frequency. The universe apparently has a casting director and that casting director thinks important information about the nature of consciousness should be delivered by the guy from Pineapple Express.

Because here's what Webb claims about where his model came from: non-human intelligence gave it to him.

His book is literally called The Human Mind Owner's Manual: Your Connection to the Universe, Consciousness, and Non-Human Intelligence. He talks about having conversations with Jesus. He discusses CE5 contact protocols. He's been to the Monroe Institute for remote viewing training. He references Skinwalker Ranch. His X/Twitter feed oscillates between AI benchmarking data and consciousness exploration with the same level of enthusiasm for both, which is either deeply unserious or the most serious thing happening in AI right now.

The Hybrid Clause

Oh, I almost forgot. He also claims he's an alien hybrid.

Someone apparently told him "be safe, my friend" — you know, the kind of thing people say when you're making noise about non-human intelligence giving you working equations — and his response was essentially: if I'm a hybrid, killing me would violate a super important interstellar treaty. So. Don't worry about it.

Let me read you this man's resume as it currently stands: Georgia Tech systems engineer → 20 years modeling emotions → aliens gave me the equations → I talk to Jesus → I did remote viewing for the CIA's favorite consciousness institute → I'm an alien hybrid protected by an interstellar treaty → oh and my AI wrapper outperforms OpenAI on empathy benchmarks → and Navy SEALs call my mind training the best on the planet. This reads like a character creation screen where someone maxed out every stat including the ones that contradict each other. He's simultaneously specced into Hard Science, Mysticism, Extraterrestrial Diplomacy, and Military Credibility. The game shouldn't allow this build. There should be a constraint that fires. But here we are and the benchmarks keep clearing.

The Tesla Frequency

In December 1900, Nikola Tesla wrote a letter to the Red Cross containing the line: "We have a message from another world, unknown and remote." In February 1901, he published "Talking With Planets" in Collier's Weekly, describing signals he'd received that he believed came from extraterrestrial intelligence. Tesla also claimed his inventions arrived fully formed from an external source — not from dreams, not from subconscious processing, but as complete detailed visions delivered from outside himself. He could see entire machines in his mind with engineering-grade precision before building them.

That was 125 years ago. Scientists later determined the specific signals Tesla detected were caused by Jupiter's moon Io passing through Jupiter's magnetic field — a natural phenomenon. But that explanation covers the signals, not the inventions. Tesla's claim about receiving complete technical visions from an external source — visions that produced working technology that changed civilization — has never been adequately explained by conventional cognitive science.

Webb is making the same claim, 125 years later. Information arriving from non-human or external sources. Producing mathematical frameworks that actually work. Outperforming the conventional approaches built by teams of hundreds with billions in funding. And — like Tesla — doing it while saying things that make serious people visibly uncomfortable.

The pattern is: person claims impossible source for information → information works → everyone focuses on the source claim instead of the working information.

The Simulation Reframe

Here's where it gets philosophically interesting in a way that this site cares about.

In a simulation or constructed reality, "non-human intelligence gave me information" and "my subconscious processed patterns during an incubation period and surfaced novel insights" might be the same phenomenon described in different vocabulary. One description gets you a TED talk. The other gets you on a UFO podcast. But the equations work either way. The benchmark scores don't care about your ontological framework.

If reality is authored — if the substrate is computational or intentional in nature — then the distinction between "I received this from a non-human intelligence" and "I derived this through extraordinary pattern recognition" collapses. Both are the system routing useful signal to a node capable of processing it. The delivery vector is an implementation detail. The signal is the thing that matters. And the signal, in Webb's case, is producing measurable results on standardized tests designed by the Alan Turing Institute.

Similarly: in a simulation, "I had a conversation with Jesus" isn't crazy. It's a support ticket response from a higher-level admin. The question isn't whether it happened — it's what permissions level the entity that responded is operating at. Is it root? Is it a moderator? Is it an automated response system that pattern-matches to the user's cultural framework? The phenomenology is real. The ontological interpretation is where people start throwing chairs.

And "I'm an alien hybrid protected by a treaty"? In simulation terms, that's just a player character claiming they have admin-granted plot armor. Which — if the system is authored and the character is carrying useful signal — isn't even that weird. You protect the messenger until the message is delivered. Basic system architecture.

The Uncomfortable Part

The uncomfortable part isn't that Webb claims non-human intelligence gave him an emotion equation. Lots of people claim lots of things. The uncomfortable part is that the equation works. It works well enough that Navy SEALs use the resulting training. It works well enough that when converted into AI algorithms, it apparently produces world-record Theory of Mind scores. It works well enough that he open-sourced it and nobody has published a convincing rebuttal of the methodology.

You can dismiss the source claim. You can't dismiss the benchmarks. And that's the gap where interesting questions live.

Either a systems engineer from Georgia spent 20 years doing genuinely novel cognitive modeling work and his claims about the information source are irrelevant noise attached to valid research — or the information source matters and we need to update some priors about the nature of consciousness and where knowledge comes from — or (the option nobody wants) both things are true simultaneously and the universe is, once again, delivering critical signal through the most absurd possible channel.

A guy who sounds like Seth Rogen, talking about conversations with Jesus, claiming alien hybrid status protected by interstellar treaty, open-sourcing emotion equations on GitHub, beating the world's most expensive AI models on empathy tests.

The pattern holds. The signal doesn't care about the vessel. The vessel is always funnier than you'd expect.

Links

Zenodelic.ai · Mind Hacking Happiness · YouTube · X/Twitter · GitHub (EI Algorithms)