An External Prefrontal Cortex — The MUEDlobe Concept

So here's the thing — I actually know my brain's specs as hard numbers. I took the WAIS, the intelligence test.

Working memory at 128, digit span 18. Apparently that's top 2%. But my processing speed was 97, coding subtest at 8. Basically, I'm great at juggling complex things simultaneously in my head, but terrible at cranking through simple tasks efficiently.

Right. And it maps directly onto music production. Structural design — building the skeleton of an arrangement, mentally manipulating the relationships between multiple parts — I'm ridiculously fast at that. But organizing files, naming things, writing invoices? Catastrophically slow.

I've lived with this for almost 50 years, and the only conclusion I've reached is: double down on your strengths, and build systems to handle the weak spots.

That's one of the reasons I built Hoo for MUEDnote. Organizing notes during production, structuring logs — I wanted AI to handle the routine processing I'm bad at.

But running Hoo made me realize something. A one-size-fits-all AI doesn't quite fit, you know? For someone like me who dumps fragmented thoughts in rapid bursts, the AI trying to summarize mid-stream gets in the way. For someone with high processing speed, the AI explaining things too slowly becomes frustrating.

Exactly. So I thought — what if I told the AI my brain's specs? Maybe it would mesh way better.

Yeah. The concept is an IDE for your brain's OS. Measure your cognitive traits, then optimize Hoo's behavior based on the results.

Three steps. First, "Measure" — game-based mini-tests that gauge your cognitive traits. Then "Edit" — adjust Hoo's settings based on those results. Finally, "Sync" — the tuned Hoo integrates with MUEDnote and MUEDear.

Take Pattern Grid, for instance. A Launchpad-style grid lights up in sequence and you reproduce the order. Even in reverse. That measures working memory. Speed Sort has you rapidly categorizing elements on screen — measures processing speed and impulsivity. Then there's Context Bridge, a chat-based format with Hoo where you respond to abstract prompts. That reveals verbalization tendencies.

But it's not clinical diagnosis. The goal isn't producing an IQ score — it's collecting cognitive data that's useful for Hoo's prompts. The matching is: "Your brain tends to work like this, so Hoo will behave like that."

Yeah. Presets are auto-applied based on your measurement results. For someone like me with high working memory and low processing speed, you get Scribe Mode. It fills in logical gaps in fragmented speech and logs it, maintains long context windows, and doesn't rush you with "so what's the conclusion?"

Navigator Mode. Step-by-step instructions, frequent check-ins, task-list format for organization.

And for power users, we open up direct prompt editing. Embed scores as dynamic variables, inject snippets for drill-sergeant mode or counselor mode. You can literally program your own external prefrontal cortex.

Because for 30 years, my own brain was a black box. I made music on "I don't know why I can do this, but I can." People around me called me a natural talent. But I never understood what was supposedly talented about me.

Yeah. When the numbers came back, it finally clicked. High working memory — that's why complex structural design comes easy. Low processing speed — that's why routine tasks are painful. Not genius, not anything special. Just peaks and valleys in a cognitive profile.

The guilt around my weak spots disappeared. Low processing speed isn't laziness — it's a spec. So build systems to compensate. And the high working memory? Lean into it more consciously.

And I want to give that to other people too. So many creators agonize over "I don't have talent" without even knowing their actual strengths and weaknesses. I want to tell them — it's not about talent. It's your cognitive profile.

The brain only ruminates on memories. That's the premise. But how you ruminate varies from person to person. Someone with large working memory can pull in lots of memories at once and do complex recombinations. Someone with high processing speed can cycle through those recombinations faster.

And AI is also a reconstruction engine, but unlike humans, it doesn't have peaks and valleys in its cognitive profile. It's uniformly fast across the board. So you keep the human's peaks as they are, and use AI to fill in the valleys. That's the essence of MUEDlobe.

That's why I called it an "external prefrontal cortex." The prefrontal cortex handles executive function, right? Planning, judgment, inhibition. In my case, the executive functions tied to processing speed are weak. Hoo picks up the slack there. But for structural design, I'm stronger than any AI, so that stays with me.

Right. I said that phrase on instinct originally, but when I saw my WAIS results — it was literally that.

MUEDnote is the production log. MUEDear is ear training. MUEDlobe is brain measurement and configuration. And Hoo sits at the center of all of them.

In MUEDnote, it fills in your thinking during production. In MUEDear, it calibrates difficulty to your cognitive traits. Across every app, Hoo knows "this is the kind of brain this person has."

Getting personalized by "music you like" doesn't excite me. I already know what I like. But when someone says "your brain tends to work this way, so Hoo will act like this" — yeah, that lands. It feels like it's covering the parts I can't see about myself.

Concept stage. The plan is written, but nothing's built yet. MUEDnote and MUEDear are in review first, and this comes after those stabilize.

No. But the concept comes from lived experience, so the core won't drift. The relief I felt when my WAIS results revealed my cognitive profile. The liberation of "my weakness isn't laziness — it's just my specs." I want to give that to other people. Through games. Casually. And make it fun.

Same thing, really. It's about understanding how your own brain works. Once you know, it gets easier. Not knowing is what makes it hard.