Submitted by Matt Hart, Founder @ Better Ideas Faster
On simulations, consciousness, AI, and the only advantage that lasts
Simulations
A 20-year-old undergraduate in Beijing built a prediction engine in ten days. Working alone with AI coding assistants and a laptop, he created MiroFish – a tool that topped GitHub’s global trending list, racked up 28,000+ stars, and landed him a $4 million investment from one of China’s original internet billionaires.
What MiroFish does is worth understanding, even if you never touch a line of code.
Most prediction tools crunch numbers. MiroFish builds societies. You feed it seed material – news, social data, policy documents, research, even fiction – describe your prediction question in plain language, and the system generates hundreds or thousands of AI agents, each with distinct personalities, backgrounds, attitudes, and decision logic. It drops them into a simulated environment and lets them interact. They argue. They influence each other. They form factions, shift positions, change their minds. Their memories persist across rounds.
The output isn’t a probability score. It’s an emergent forecast – what the simulated population collectively arrived at through social dynamics playing out over time.
And you can step inside afterward. Query individual agents. Introduce new variables. Test counterfactual scenarios. It’s not a black box. It’s an interactive world.
One kid. Ten days. Open source. Four million dollars.
It crystallised a thought I’ve been mulling for a while – todays it’s less about what ideas to build and more about who is asking the right questions.
Consciousness
There’s a cognitive scientist at UC Irvine named Donald Hoffman who has spent decades developing a thesis that should, by rights, unsettle anyone working in innovation, creativity, or strategy.
His argument – backed by evolutionary game theory and computer simulations – is that our senses did not evolve to show us reality. They evolved to keep us alive.
Evolution doesn’t reward truth. It rewards fitness. And in every simulation Hoffman’s team has run, organisms that perceive the world accurately go extinct when competing against organisms whose perceptions are tuned purely for survival utility.
He calls this the Fitness Beats Truth theorem.
The implication is radical. What we experience as reality – space, objects, colour, time – is not the world as it is. It’s an interface. He describes it as a ‘species-specific operating system shaped by natural selection to give us useful icons, not accurate representations.’ The same way a file icon on your desktop is a helpful symbol but bears no resemblance to voltage patterns on a circuit board, the ‘reality’ we navigate is a compressed, simplified dashboard designed for one purpose: keeping us in the game long enough to reproduce.
Evolution hides the truth from us. Not as a flaw. As a feature. I mean get your head around that?!
When you look at the world through this lens, a lot comes into focus. The tech founders optimising for scale at any cost? Fitness payoffs.
The platforms engineered to capture attention regardless of what it does to human wellbeing? Fitness payoffs.
The venture logic that rewards extraction speed over depth? Fitness payoffs.
The entire machinery of late capitalism – move fast, ship product, capture value, repeat – maps almost perfectly onto Hoffman’s framework.
The organisms that see fitness win. The ones that pause to ask whether any of it is true get outcompeted.
AI Illusions
Now hold MiroFish and Hoffman in the same hand.
Hoffman says we’re already operating inside an interface that masks objective reality. Our entire experience of the world is, in a meaningful sense, a constructed simulation – shaped not for accuracy but for utility.
MiroFish builds a second simulation on top of that. Artificial agents with synthetic personalities, running through social dynamics in a digital environment, producing emergent predictions about how people – who are themselves navigating a perceptual interface that hides the truth – will behave.
A simulation inside a simulation. An illusory world modelling an illusory world. And it works. The investor wrote the cheque. The demos produced coherent forecasts. The use cases are real and multiplying – market sentiment, election outcomes, policy impact, product launches, competitor response.
So what does that tell us?
Maybe this: that ‘truth’ was never the point. Not for evolution. Not for these tools. We’ve never had access to objective reality. What we’ve had is a set of perceptual tools that are good enough – good enough to find food, avoid predators, raise children, build civilisations.
It’s not true. But its crazy useful.
And now we’re building a new layer of tools – digital twins, agent-based simulations, swarm intelligence engines – that do exactly the same thing at a different scale.
They don’t need to model truth to be valuable. They need to model utility.
What’s likely to happen when a complex population of actors with different motivations responds to a change in conditions.
This isn’t a departure from how we’ve always operated. It’s an extension of it.
Evolution gave us a perceptual interface to navigate a reality we can’t see. Now we’re building computational interfaces to navigate complexities our perceptual interface can’t handle.
Layer on layer on layer.
Here’s where the bottleneck sits.
MiroFish is open source. The code is free. The AI tools that helped build it are available to anyone. The LLM backends are commodity infrastructure. A 20-year-old built it in ten days.
So the advantage isn’t in the software. It isn’t in the hardware.
It isn’t in access, because access is approaching zero cost.
It’s in the human who decides what question to ask.
Who selects the seed material.
Who reads the output and can distinguish signal from noise. Who brings lived experience to bear on which scenario to test.
Who looks at the emergent forecast and sees what it actually means.
The tools are being democratised in real time. The ability to direct them is not.
That’s the Humanware thesis I’ve been developing – that mapped human creative capabilities are the irreplaceable advantage in the age of AI.
The bottleneck isn’t what’s buildable. It’s who has the ideas. And more than that – who has the depth to know which ideas matter.
But if that’s where the argument stopped, it would be a neat professional insight and nothing more. Useful. Tidy. Incomplete.
Because I’m thinking Hoffman’s thesis has a problem. And the problem is where this gets interesting.
A Dent in the Matrix?
If evolution ruthlessly selects for fitness and drives truth to extinction, there are some features of human experience that have no business being here.
Contemplative practice. Mystical experience.
The felt sense that there’s something deeper behind the interface. The drive toward meaning that has no reproductive payoff.
The pull to sit in silence, to go inward, to ask questions about the nature of consciousness itself.
These aren’t marginal phenomena. They show up in every culture, in every era, across every geography. Indigenous knowledge systems. Buddhist meditation. Sufi poetry. Christian mysticism. Jungian depth psychology. Plant medicine traditions that are thousands of years old.
On pure fitness logic, all of it should have been selected out long ago. The person in the cave meditating is burning calories and not reproducing. The shaman in ceremony isn’t gathering resources or defending territory. The mystic contemplating the nature of the self is doing precisely nothing that evolutionary game theory would predict or reward.
And yet these traditions haven’t just survived. They’ve persisted with extraordinary resilience – often outlasting the empires and systems that tried to suppress them. That persistence is a signal. And Hoffman’s model, for all its rigour, doesn’t explain it.
The Monk and the Empire
I keep landing on two possibilities.
The first is that Hoffman’s model is simply incomplete. Fitness-for-reproduction may be a powerful selection pressure, but it isn’t the only one. There may be something else operating – a pull toward coherence, integration, wholeness – that runs deeper than survival logic. Something that doesn’t show up in evolutionary game theory because it operates on a different axis entirely.
Not fitness.
Not even truth.
Something more like directionality. A current that runs beneath the interface. The second possibility is stranger and, maybe, more interesting.
What if the wisdom traditions are a fitness strategy – just one that operates on a timescale Hoffman’s simulations can’t capture?
Think about it. The extractors move fast. They dominate in the short term. They accumulate resources, capture markets, build empires. But they also burn through everything in their path. The fitness-maximising organism consumes its environment and then collapses.
Tokens. Environment. Destruction. Then, rinse and repeat.
The contemplative traditions play a different game. Slower. Deeper. Less visible. But they’re still here. The monks are still here. The indigenous knowledge systems – battered, suppressed, colonised – are still here. The meditation lineages that started two and a half thousand years ago are still here.
The people who see deeper, not just faster, turn out to be the ones still standing when the extractors have eaten everything in sight.
Maybe meaning is a fitness payoff. Just not one that shows up in a fifty-round simulation. Maybe it shows up across centuries.
The monk outlasts the empire. Every time.
I’m not writing this as philosophy. I’m writing it because I think it has direct implications for anyone building, creating, or leading right now.
The dominant narrative is pure Hoffman. Fitness payoffs everywhere. Optimise. Automate. Scale. The tools are free, the models are commodity, the barriers to building are approaching zero. And the people winning – visibly, loudly, measurably, obnoxiously – lying, cheating, something-a-rather-maxxing – are the ones playing the fitness game hardest.
But I keep coming back to MiroFish. A beautiful piece of engineering. Open source. Built in ten days. Capable of simulating entire populations. And completely dependent on the human who decides what question to ask it.
The simulation doesn’t have purpose. It doesn’t have meaning. It doesn’t have the felt sense that one question matters more than another. It processes seed material and runs agents and produces outputs. Brilliantly. Efficiently. At scale.
The part that can’t be automated – the part that makes the whole thing useful – is the person who brings something the simulation doesn’t have. Judgement born from experience. Meaning born from the kind of inner work that has no fitness payoff on paper but somehow produces the questions nobody else is asking.
I’ve spent twenty-five years working in creativity and innovation – building frameworks, running programmes, helping organisations and individuals develop their creative capabilities. And I’ve spent the last decade in a parallel process of
deep personal work – consciousness, plant medicine, practices that sit firmly outside any professional playbook.
I used to think those were separate tracks. The professional and the personal. The strategic and the spiritual.
I don’t think that anymore.
I think they’re the same track. The capacity to ask better questions – to direct simulations, to see through interfaces, to bring something genuinely new to a world drowning in optimisation – comes from exactly the kind of inner development that Hoffman’s model says shouldn’t exist.
My thesis is that becoming is the competitive advantage.
Not because it makes you faster. Because it makes you deeper. And depth is what generates the ideas that no model can produce on its own.
Hoffman proved that we live inside an interface. Fair enough. But someone had to look at that interface and ask: what’s behind it? That question didn’t come from fitness. It came from something older, stranger, and more persistent than anything natural selection can account for.
And I think that same impulse – the refusal to accept the interface as the whole story – is exactly what the age of AI is going to demand of us.
Not more simulation. More depth. Not faster tools. Deeper humans.
I don’t have this figured out. I’m writing my way toward it – in public, in practice, in a PhD I’ve titled Story as Method. The research is ongoing. The questions are live. And I’m increasingly suspicious that the most important capability we can develop right now isn’t technical at all.
It’s the willingness to go deeper than the interface wants us to go.
Does the monk outlast the empire? I think so. But I’d love to know what you think.
Give Matt a shout on LinkedIn
