Evolution, LitRPG, and AI Harnesses

Ok, this will be a crazy and long one, but if you're into evolutionary biology, AI harnesses, or LitRPG, I think you will trip off this as much as I have.

Basically, I'm going to link these things together as a single concept, and explore what we can do with that information.

/lfg

The base layer ​

Pretty much everything that we care about as humans goes back to evolution. Basically, we are the mech suits for evolution. Specifically, for genes.

Evolution has a master game: to produce more and more competent organisms that can propagate the most competent genes, and it works by injecting these organisms with goals, using biological systems such as hormones and emotions to steer us.

We have anger. We have jealousy. We have ambition. We have all these different things, and other animals have different versions of those. There's a spectrum there.

The central concept is that we only want things because evolution gave those wants to us. Everything about human nature is based on this. We enjoy things and dislike things and are steered toward and away from things because evolution built it that way.

Our entire incentive system, our goal system, our preferences, and our likes. We can't choose which foods we like. We can't choose if we like blondes or brunettes. We just want things. We just prefer things.

This is in the substrate of biology in our system, and more importantly, evolution, which has all the incentives and disincentives and punishments.

So that's the base here.

The LitRPG system ​

The next piece is LitRPG.

LitRPG is a genre of literature in which you have characters who are injected into a gaming system. The gaming system has parameters: clothing, armor, weapons, magic, psionics, and different components of capabilities.

Over the course of a LitRPG series, which tends to be very long, you see constant escalations.

I'm going to use one particular series, The Primal Hunter, as my favorite example. I'm on book 12 or 13 or something. The central character is this guy named Jake, who used to have some random job. He wakes up in a system version of Earth, which is completely controlled by this air-quote “system.”

The Primal Hunter, Book 1 by Zogarth

The system is kind of like evolution. It determines all the rules. Clothing, armor, damage, magic, capabilities, escalation paths. Everything.

As you proceed with this character, he starts figuring out:

• I need shelter.

• I need armor.

• I need basic defenses.

• I need to learn how to fight.

He has to choose a class. He has to choose a race. He decides to be human, but he could have chosen lots of different races. He could have chosen something other than hunter, but he chose to be a hunter. Then he gets to choose all these other sub-paths.

It's a really cool system because each time he chooses something, it opens up completely custom things. Characters are vastly different based on the choices they've made.

And then there's the narrator.

The narrator also narrated Cradle, the first series I got into. In The Primal Hunter, he says, “Ding!” He says it at the same tone and pitch every time.

Unsouled, Cradle Book 1 by Will Wight

Even my partner, who's in the car with me when I'm listening, has already Pavlov-associated this with a good thing, because he says it at the same tone and pitch every time.

Human race advanced to level 274. Class advanced to level 274. Or a level in alchemy, which is the profession he chose. Alchemy has thousands of variants, and every choice he makes produces new variants.

The entire LitRPG system is a perfect analog to evolution itself.

Ding! ​

Look at what we do as humans. We're constantly optimizing for all these different things. The only question is whether we're aware of it.

We are the mech suit. Evolution is on the inside, pressing pedals and pulling levers. It's determining what ultimately matters. It's deciding what we're going to enjoy and not enjoy.

And the way it incentivizes us to advance is through these dings.

When the narrator in a LitRPG says “Ding,” that's the same thing as evolution squirting a reward into our system. Suddenly, we feel the spurt.

We train in karate or volleyball or whatever it is. We fight and grind. We struggle. We try to push through hardship.

When you choose something difficult and push through and achieve something, maybe you got third place in the karate competition. Maybe you got first place. Maybe you got into the Olympics. Maybe you didn't get that far.

You're getting a ding.

That ding in LitRPG books is the same ding we get from evolution.

Another thing that's crazy about it is that evolution is fucking harsh. It is super fucking mean.

How long do you get to enjoy your Olympic trophy? How long do you get to enjoy that new Hermès bag, that new Rolex, or that new Porsche? Whatever it is that you like.

You spend all this time grinding and trying to make more money and trying to meet the right mates, all for the purpose of getting to the next thing.

As soon as you get there, evolution devalues that thing in your mind. You look at it on the shelf and you're like, “Nah. Not really a big deal.” It's in your past. It gets marked as uninteresting.

You can collect trophies, and the only trophy you care about is the one you don't have.

The hedonic treadmill is exactly the mechanism I'm talking about. It's one of the features of evolution that keeps us wanting more.

This is fucking profound. And it's the exact same game.

I learned this from Dragon Ball Z retroactively, once I started getting into LitRPG. It was all about escalation. Levels, sub-levels, branching, comparison, competition.

This is all the language of evolution.

There's levels to this game ​

Now we add the third piece: all this AI stuff.

For some reason, I've had this drive, this characteristic of LitRPG, for my entire life. My first memory of it is being in elementary school and breakdancing. I got really good at it for my local area. I loved seeing other people's styles. I had my own style. I thought it was cool.

The guy next door was like a god in my mind. He was older than me. There's a thing called “going up against,” and he went up against someone in the local area, just a few streets over.

This other guy destroyed my mentor.

I was like, “Oh my god.” That's the first time I realized there's levels to this game.

Again, deeply evolutionary psychology language. Evolutionary biology language.

I believe I made my first personal AI harness in July of 2025, just a few months after Claude Code came out. I immediately started building without realizing this connection with LitRPG.

I started building a harness. I started building skills. And I put in all my personal goals.

I already had this concept of a TELOS: a definition of what you're trying to pursue, and making sure you're tracking exactly what you're trying to pursue, which I'd been working on long before AI.

For the last 20 years or so, I've wanted this structured direction:

• What exactly are we working on?

• Why?

• What is the structure?

• What are the obstacles?

• What are we doing to overcome the obstacles?

I immediately brought that into the harness and started adding skills. Skills are ways to accomplish a task in a small, discrete, modular way.

I was blown away by AI for a very specific reason. I was like,

“Shit. I actually now have an alien, friendly companion to help me pursue all my goals. Everything in my TELOS.”

A unified system for pursuing one's goals, both personal and work.

The more powerful AI became, the more obvious this became to me.

A multi-axis upgrade infrastructure ​

Jake in The Primal Hunter is constantly thinking about what he can do to enhance things. Whenever he gets loot, whenever he gets money, whenever he learns something new, he's thinking,

“What should I upgrade? How should I improve my defense?”

If he finds someone who fights in a different way or style, he's like,

“Okay. That guy would have killed me. That's super bad.”

Now he needs to create some sort of magic item, friendship, or capability that can monitor for that or continuously defend against it. He needs a shield.

And that shield needs to be upgraded.

Here's the similarity across all three themes: evolution, LitRPG, and AI harnesses.

It's a multi-axis upgrade infrastructure.

We're talking about thousands of dimensions, and each dimension has all these branches.

When we're playing LitRPG, we're playing mini evolution. And when I'm playing AI harness, I'm also playing mini evolution.

This is why I'm so obsessed with it. There's nothing more primitive and rewarding than that base layer of evolutionary psychology.

As I learn AI, as I enhance my harness, as the models get better, what am I doing? I'm constantly upgrading every component, mini-component, module, and sub-module of my entire harness, all for the purpose of pursuing my TELOS.

Guess what my TELOS is mapped to? The same stuff that makes me happy evolutionarily.

The same concept, going back years ​

An AI harness allows us to take our whole system of striving and goals, place it into a harness, and explicitly tell it,

“Your job is to help me advance these goals.”

I wrote about digital assistants in 2006. I didn't even say “AI.” I said semi-intelligent or pseudo-intelligent systems.

I wrote a book about it, The Real Internet of Things, in 2016. Personal assistants, which I call digital assistants, running on AI. Companies becoming APIs. We would talk to our agent, and our agent would do things for us based on how much it knew about us and our context.

The Real Internet of Things by Daniel Miessler

The concept of that book was that we cannot predict tech, but we can predict where tech is going because we know what humans want.

We want to be attractive. Competent. Smart. Interesting. Useful.

Those are the things evolution wants us to be. So in The Real Internet of Things, I said the tech we're going to build is a personal AI that knows everything about us and continuously advocates for us, shaping the world to take us in the direction of our goals.

I wrote about that in 2006, then in the book in 2016. I was working on TELOS before AI, in the last months of 2022. Then in 2025, the harness became a thing. That's when I started taking it to the next level.

It's all the same fucking concept.

What is the structure of our goals? How are we pursuing them? How can we get help?

Now we have an alien friend in the form of our harness that can help us 24/7, while we sleep, pursue all these different things.

I made a harness called LifeOS, which is free and open source. It has intelligence, open-source intelligence, research tools, and all this stuff so that your AI harness can constantly survey the world for you.

In “AI's Predictable Path,” which is loosely based on my book, I use a vignette where you're at a coffee shop and have a daughter across the country in college.

Your digital assistant tells you,

“She's walking home from a bar and someone is following her. I've dispatched some drones. I've called the police. Someone is on the way. I've also invoked a local beacon, and a helpful person with a high rating is stepping outside his house. He's going to walk behind her and shoo away the guy before the cops get there.”

That's infrastructure and wiring and piping. But why would I predict this technology would be built?

Because it's our fucking kids. We will want technology that can protect our kids. All I did was project forward what that technology might look like.

Reality is layer dependent ​

There's a question of free will here, but I don't want to make this whole post about that.

I think at the base layer we don't actually have control. Ultimately, we're not making choices. It appears we're making choices. It feels like we're making them, so we might as well be making them.

Reality is layer dependent.

At the human layer we exist on, we are deciding our own fate and making our own choices, very much within the confines of evolutionary psychology. We have some measure of practical free will, some measure of creativity.

I don't wake up in the morning and go, “Gee, I wonder what the universe is going to do today.” I ask myself, “What am I going to do today?”

I have no choice but to live inside this human layer.

We treat the harness as if we have free choice and we're deciding all these things. From the experience layer, it's 100% true.

If you're interested in evolutionary psychology and you've read a bunch of LitRPG, you get to enjoy the fact that, holy shit, we're building our own LitRPG system for our actual life. Customized for ourselves and our goals.

Whose goals are you pursuing? ​

The whole point of this at the human layer is something I take from Sam Harris. He didn't say it this way, but this is my characterization of his Moral Landscape.

The Moral Landscape by Sam Harris

My goal personally is to improve the lives of conscious creatures. Increase their eudaimonia. Reduce their suffering.

Becoming explorers, librarians, teachers, parents, nurturers, and artists, like in Star Trek: The Next Generation. Creating experiences for others. Improving the lives of others.

This is part of the base of Human 3.0.

We're in Human 2.0 right now. We're taught to go to school and college for what purpose? To get a job working for someone else.

Very few people are in the position to say,

“I'm going to come up with my own ideas. I'm going to have my own company. I'm going to create something net new in the world.”

The reason is legacy. Our education system and society are based on producing workers. Capitalism simply isn't oriented to maximize human flourishing and fulfillment.

What I'm trying to get at with Human 3.0 is self-activation. People realizing that they've been told by parents, peers, and schools that their job is to do well on tests, not color outside the lines, get grades, and work in a cube helping Miss Johnson do her slides.

Because Miss Johnson is important and you are not.

The whole world thinks they're not important. Very few people are told they can be the person on TV or the person who's going to write a book.

This is what I'm trying to unlock.

What is your actual desire? What is the actual fire within you? Did you have it when you were a kid? Did you lose it?

One of the concepts I got from Rilke is that when you're a child pursuing imagination, building forts, fighting wars, playing with dolls, playing house, that's true life. Then we're taught to discard it and grow out of it. One of the things he teaches is that we have to get back to that.

Who are we? Can we articulate who we are?

That's what TELOS is for: articulating the problems in the world you want to solve, and setting out ways to pursue that and get there.

The purpose of Human 3.0 is activation. Having people look within, figure out who they are, and articulate it. Helping people articulate and pursue their ideal state.

We could do this without AI if we had a more advanced society, but a lot of things limit creativity, and if you've got five jobs trying to put food on the table, it's pretty hard to think advanced thoughts.

Look at someone like Bertrand Russell or Marcus Aurelius. They grew up rich and had tutors on every topic, and they turned out to be amazing people. Big surprise.

A big part of the promise of AI is getting more people the freedom to listen for their inner voice, articulate it, and pursue it. That's really hard if you have to work at McDonald's, then Wendy's, then clean houses.

I've been lucky. I already work for myself, I've been exposed to lots of books, I've had a decent career in security, and I was born in the Bay Area, so I've got all sorts of advantages.

The problem with AI and people and their lives is that they don't know how to articulate their ideal state. They don't know who they are or what they want. They've been handed a set of goals by parents, peers, and their education system. They're on autopilot, chasing other people's goals.

That's why a lot of high performers wake up in their 30s and 40s, or even their late 20s, and say,

“I did everything right, and I'm not happy.”

They pursued someone else's goals.

The whole purpose of the harness is to deeply understand who you are, what you're trying to do, and what your goals are, and have your system work for you to help pursue those things.

What do we actually do with this? ​

Let's say we accept these three components and how they work together. What do we do?

I, for example, don't like to be surprised. I never want to be fucking surprised by anything. Except a surprise party. Even then, I'd like my DA to tell me, “By the way, they're waiting inside. They're going to make a very loud noise.”

I've got stocks, I've got investments, I'm spending my time a certain way, and the future is going to play out in a certain way.

There are a million things happening all over the world right now that could affect my future. If I knew about them, I might make different decisions.

What information do I wish I had that I don't currently have?

What processes can I kick off inside my harness? It's no different from Jake implementing a new magic system, building up his home bunker, or building an alchemy lab underneath his house on the new system version of Earth.

It's all the same shit.

I'm going to my harness and saying,

“Listen. Make sure nothing surprises me and liquidates my stocks. You need to tell me.”

I just built a system called Socrates. It's inside Pulse, inside LifeOS. It's a giant set of questions:

• Have any financial things happened that put my stocks at risk?

• Is there anything I should be investing in that I'm not, based on a merger, a company going under, or a new company being built?

• Is there anything I'm currently invested in that I should get out of?

• Are there any news sources I should add to my feeds?

• Are there any things I should be doing every day?

• Are there any supplements I should stop taking or start taking?

• How should I modify my workout so I gain the maximum amount of muscle?

• How can I improve my relationships?

All these fucking things. There are a million things we could be doing better.

This can be implemented inside a harness in the form of hundreds or thousands of agents working 24/7, solving problems and answering questions from our Socrates list.

A primary concept running through the harness is what I consider the purpose of AI once it gets to a certain level: the transition between current state and ideal state.

If that requires a Socrates agent asking you questions to make sure you're not getting lazy and having AI do everything for you, that's part of it too.

What can I do practically? The same thing Jake is doing. Constant little optimizations of everything.

The human is the most important thing ​

Keep in mind, you don't want to go overboard. You don't want the system to be the whole fucking point.

Sometimes you should turn off all your AI and walk away. Climb a mountain, go on a meditation retreat, bring dead-tree versions of books and read for three weeks straight, and think and study and write with a pen on paper.

This doesn't have to be technology-focused. A primary belief of mine is that the human is the most important thing here. The advancing of humans.

One of my deep beliefs is that we're roughly happier when we obey the rules of evolution, but we don't have to do it in a primitive way. We want to advance. We want to progress.

We're not the same as a meerkat. We have consciousness, whatever that means. Even if we don't have true free will, we're different from an ant, a slug, or plankton.

Whatever freedom we have and experience, we should rejoice in it and move within those confines, while understanding that our substrate is still biology. Evolutionary biology.

If you're given everything and don't have goals, aren't pursuing, aren't struggling or striving or facing adversity and doing difficult things and achieving, you're fucking unhappy.

Yes, understand that we're inside this cage. We're the mech suit, and evolution is driving. But within those confines, we do have some control.

If it's not absolute, so the fuck what? It still feels great. We still have so much room to explore.

We can set our TELOS. We can pursue our goals. We can build our own LitRPG system using an AI harness and use it to make the world a better place.