Imagine for a moment you’re on a flight, 35,000 feet in the air. Out of nowhere there’s a huge jolt of turbulence. Your stomach sinks. You start sweating. All of a sudden you’re religious.
You close your eyes and visualize the plane landing safely, I’ll never take life for granted again, just put this thing on the ground. Unfortunately, your internal beliefs have absolutely no effect on the plane.
If this seems obvious to you, why do you do it?
We do it because we are hardwired to engage with reflexive systems, systems where our beliefs actually can change the outcome of a situation.
A plane is not one of those systems. Planes are completely insensitive to your beliefs.
But this is not always the case.
Planes are relatively simple, typically made up of about ~6 million parts. Most of these parts are irrelevant and none of them are aware of you, or sensitive to your beliefs.
Essentially, planes are not complex enough.
People, on the other hand, are much more complex (by at least two orders of magnitude). The average person is made up of 30 trillion cells networked into goal-seeking cybernetic loops at multiple levels of organization.
People are so complex that your beliefs can affect them (e.g. believing you will be liked makes you more likely to be liked).
Walk in to a room full of people and try it. If your self-confidence is felt deeply enough, it will leak out of your body. You express this belief unknowingly, through small gestures and micro-facial expressions, your poise and the way you carry yourself, what you choose to say, how you say it, pheromones, who you choose to interact with, and more.
“It is absolutely imperative that you keep your thoughts in check and your awareness where it needs to be. Your consciousness physically affects material reality. Literally. No exaggeration. The double slit experiment taught us particles behave differently when observed. We know as a matter of fact that random number generators can be swayed and tilted to one side within a binary, by intent alone. Extrapolate the above principles across your every day life. Energy flows from the immaterial to the material in a very real way. Your reality is a direct reflection of your thoughts. By extension, changing your thoughts physically changes the reality you experience.” - @upskillyourlife
Reflexivity
Reflexivity has a long history of mention. It’s been referenced in movies, stories, thought experiments, art, and the sciences, throughout history (see: Borge’s The Circular Ruins, M.C. Escher’s Drawing Hands, Russell’s Paradox, Turtles all the way down, The Alchemy of Finance, I am a Strange Loop, and more).
“Some things have to be believed to be seen.” —Ralph Hodgson
Across these examples, a common theme is a collapse of the usual subject–object divide. The observer and the observed, the cause and the effect, the map and the territory get entangled or even merged into a confusing loop. Becoming aware of this often forces us into a higher perspective.
“To make progress we must step outside the loop or recognize the loop as such.”
Personally, I can’t help but get excited when I see the word reflexivity. It feels like magic. You can just think something and change the world?
But, we don’t seem to understand it, or know how to control it.
Most references to reflexivity gesture at it with an obscure humility, expecting us to just accept that it’s confusing and move on.
Nobody talks about reflexivity as a practice, as an application. How do we harness this obviously powerful (and increasingly important) force?
The coining of the term hyperstition by Nick Land made progress towards this, but it remains obscure and science fictiony.
This essay is an attempt towards a more advanced understanding of applied reflexivity.
Applying Reflexivity
As a general rule of thumb:
The more complicated a system is, the more likely it’s reflexive. When something is made up of enough parts, some parts become aware, or sensitive to, the internal states of the entities that observe it.
This is important: reflexivity isn’t a binary switch, it’s a dial. Complexity increase the dial, bit by bit, until belief (or the system’s own dynamics) becomes a primary causal factor of its dynamics.
Examples of this in the real world are obvious to anyone who’s paying attention:
That date didn't fail solely because of objective incompatibility. It failed because one person (or both) went in thinking it probably would.
That startup didn’t succeed because the idea was objectively brilliant, but because the founders’ belief infectiously convinced investors and potential hires, whose backing amplified belief, and so on.
That stock’s price didn’t rise because the company’s fundamentals improved, but because enough people collectively imagined it would.
Here’s how Dan Reardon frames it, in the context of memestocks:
The entire GameStop scenario is a case study in reflexivity. Reflexivity is the idea that our perception of circumstances influences reality, which then further impacts our perception of reality, in a self-reinforcing loop. Specifically, in a financial market, prices are a reflection of traders’ expectations. Those prices then influence traders’ expectations, and so on.
And here’s a more chaos-theoretic framing:
There's a way to describe systems as other L1 Chaotic or L2 Chaotic. So an L1 Chaotic system is like the weather: the weather doesn't change if you predict it, at least for now. Whereas the financial markets are an L2 system. As soon as you make predictions about the market, you change the market. And this is true about economists: if you make predictions about the economy, if a central bank issues their predictions about the economy, they immediately change the economy just by creating predictions about them
Reflexive dynamics are likely at play anywhere humans are present. Fake it till you make it, social proof, the pygmalion effect, the list goes on.
When it comes to the application of reflexivity, I’ve found the investing/entrepreneurial framing to be the most helpful.
This is where it gets interesting:
The *best* investors and founders are experts at seeking out the opportunities where reflexivity matters the most, where their belief does the most work. Just as investors seek undervalued stocks, people who understand reflexivity seek ideas and opportunities that are undervalued in terms of belief; where simply holding a strong belief can trigger a self-reinforcing cascade.
In other words, ideas exists along a spectrum of reflexivity, from ‘as insensitive as a rock,’ to ‘extremely susceptible to cascading reflexivity.’
You might call these areas ‘starving for belief’, or in need of ‘belief liquidity’. They're neglected, overlooked, or actively unpopular, but there’s something there.
If you identify something genuinely worthwhile that nobody else believes in yet, your conviction can do enormous amounts of work for you. It’s probably the closest thing to free lunch in modern society.
Ask yourself: "Where can my belief achieve the greatest reflexive leverage? Where can I get the highest ROI on belief?”
When Sam Altman says, "you can just do things," he’s describing precisely this dynamic. But what he really means is if you get good at spotting those under-appreciated pockets of potential (areas desperate for belief liquidity) then yes, you really can “just do things” and watch massively disproportionate results unfold.
If we assume belief is a currency, then reflexivity is compound interest. Early belief causes others to believe, impacting objective outcomes, causing others to notice, then believe, etc, etc.
If the original idea is powerful enough, it may cross the belief contagion threshold.
“Every idea has a reflexivity threshold—below it, belief remains niche; above it, explosive viral propagation occurs.”
A Framework for Reflexivity
Rather than viewing reflexivity as a nice-to-have, it should be at the center of how we view and assess opportunities, beyond the narrow framing provided by George Soros.
I’d like to propose a framework for reflexivity. I believe there are roughly four orders of reflexive leverage; discrete stages in the chain reaction of reflexive maturation that an idea can benefit from. The higher up the meta, the more reflexively ‘mature’:
1st order (Classical reflexivity): Your own belief in something being true, increasing its trueness directly (e.g. delusional optimism).
2nd order (Social reflexivity): Other people’s belief in something being true, increasing its trueness, e.g. Keynesian Beauty Contest (guessing what others are thinking). In some sense this is the number of people involved in 1st order reflexivity. As Keynes described, by the “third degree” of reasoning “we devote our intelligences to anticipating what average opinion expects the average opinion to be”
3rd order (Reflexive momentum): The speed in which a multi-person reflexive dynamic is spreading. In this case, the outcome is already being effected by others’ belief, but is propagating, increasing its impact (e.g. capitalizing on meme-stock craze). This is essentially a measure of the rate of change of 2nd order reflexivity.
4th order (Meta reflexivity): At this stage, the reflexivity of a specific dynamic is noticed and is made explicit. The awareness of its reflexivity spreads, causing it to be more reflexive, reinforcing all of the above. This essay itself—by explicitly teaching people about reflexivity—is operating meta-reflexively.
Reflexivity and Markets
Most reflexive ideas, despite being reflexive, are already crowded with belief. Every public equity in the stock market, for example, has lots of believers. There are not many publicly traded stocks that are starving for your belief, relatively speaking. You believing in GameStop won’t create an explosive cascade of reflexivity, at least not you alone.
Obviously GameStop could benefit from having another long term investor, but your belief is low leverage. You’re not getting enough bang for your belief-buck anymore.
To create an autocatalytic cascade of belief on a publicly traded stock, one must coordinate thousands, or tens of thousands of people. Too crowded to be sensitive to one person’s story-telling abilities, something that only rears its head occasionally.
The opportunities to get a *huge* multiplier on your belief are rare, and most of them are already taken. When it comes to such a powerful source of leverage, it makes sense that competition is fierce. Reflexivity has a crowding problem.
Even most startup ideas are experiencing this problem. Founding a startup used to be fairly contrarian. But today, most startup ideas have a low ROI on belief. Startups as concept, may be over-crowded at this stage. The world already believes that startups are a worthwhile undertaking, given certain risk/reward tradeoffs. Hundreds of accelerators, hundreds of seed funds, dozens of immediate copy cats, nation-state sized tech companies as direct competitors. Belief is no longer scarce, in startups as a category. One must look even further out. I can’t tell you what those areas starving for belief are, but I can help point you in the right direction.
The key is not just reflexivity, but identifying the potential for a reflexive domino effect.
There’s an opportunity to start seeing the world through this lens, looking for the areas where a cascade is most likely. I’ve developed a set of questions we can ask ourselves, at each stage.
1st order:
“How much work will my belief do on its own?”
Is this domain so crowded that another believer won’t matter?
Is the system still sensitive to a small injection of authentic belief?
Could just my conviction alter perception, engagement, or resource allocation?
2nd order:
“How likely is it that my belief sparks belief in others?”
Are people looking for someone to believe so they have permission to believe too?
Are memetic channels (X, Substack, etc) undersaturated?
Could my conviction catalyze a sort of latent consensus?
3rd order:
How quickly can this existing belief dynamic propagate?
Is this the early stage of a cascade or is it already mature and reflexively saturated?
Can I amplify the loop, or is it losing energy?
4th order:
“Is awareness of the reflexivity itself part of what’s driving it?”
Will pointing out the loop amplify the loop?
Is the narrative now about the narrative?
Is there an opportunity to be the person who names the dynamic and thereby accelerates it?
Reflexivity and Time
Even the Lindy Effect is a special case of reflexivity, across time.
The Lindy effect proposes the longer something has survived or existed, the longer its remaining life expectancy. Longevity implies a resistance to change, obsolescence, or competition, and greater odds of continued existence into the future. Where the Lindy effect applies, mortality rate decreases with time. Mathematically, the Lindy effect corresponds to lifetimes following a Pareto probability distribution.
Basically, if something has existed for long enough, the biggest “cause” of its continued existence is a reflexive dynamic between the history of its existence and its future existence.
Time is a ‘great filter’ for 2-4th order reflexivity dynamics.
Reflexivity and Societal Sense Making
As society becomes more networked and socialized via communication technology, it becomes more complex. According to our generalization, that means it becomes more reflexive (more parts means more reflexive potential). This changes the nature of reality and the causal mechanisms of society unfolds.
On a daily basis, we interface with people, society, social media, markets, which are highly reflexive.
In a reflexive society, claimX is true because people believe it to be true and social proof of claimX’s increasing trueness have been communicated across social networks increasing it’s perception of trueness, but not necessarily because it is actually true.
—
In general, we are increasingly exposed to concepts that we have very little understanding of, where we must borrow our conviction on them from others.
All of a sudden, the average person on social media is forced to have an opinion on respiratory viruses, horoscopes, stocks, AI, crypto, microplastics, political candidates, the economy, inflation, celebrities, tech platforms, geopolitics, and much more.
When you don’t have a real belief about something because you don’t really understand it and are spread so thin, you’re forced to rely on other people’s beliefs. The problem is, most others don’t really understand it either, therefore, a large % of their belief comes from other’s beliefs as well.
You get where this is going. Eventually, the concepts that are complex enough to be reflexive become more reflexive than they’ve ever been, because the main source of information about that thing is other people’s beliefs (borrowed conviction).
The more reflexive a society, the more unstable it is, because surface-level beliefs can just as quickly flip the other way. Conversely, societies in which belief is mostly dependent on reality, say, value or revenue, or other non-social dynamics, are more resilient.
If a reflexive society doesn’t eventually translate it’s belief-based cognitive memeplex into an evidence-based cognitive reality, the bubble will pop.
Actions vs. Reactions
When it comes to modeling the world, reflexivity makes things much more difficult.
Understanding culture, politics, and other social dynamics has always been difficult, but it was simpler before. The simple, action-consequence model of understanding about the world is less predictive than ever. Why? Because it’s not about the direct consequences of your actions, it’s about people’s reactions to those actions. Those reactions are reflexive, mostly determined by other people’s reactions.
“Your real action is your target’s reaction” - Saul D. Alinsky
In the book “Rules for Radicals,” the prescient Saul Alinsky, explains the case for switching to a reaction-centric model of the world. You’re better off thinking about the ‘reactions to the reactions of an event’ than modeling the event and it’s ‘logical implications’ in isolation.
This is a defining feature of a simulacra, where perception, and reactions, matter more than reality and direct consequences.
You should start understanding 2-4th order reflexivity more intuitively, if you want things to make sense going forward.
Reflexivity and Humans
It turns out, happiness may be a quality that emerges from a reflexive loop. In fact a lot of emotions may only exist in a reflexive memetic dynamic, originating from an internal dynamic and growing stronger based on self-reinforcing external evidence.
‘Gaining control of your emotions’ then, involves indexing more heavily on internal reflexive dynamics and away from external ones.
Turning to human behavior, a series of productivity studies in the 1920s highlighted an unexpected feedback loop. Researchers found that any change they made to working conditions, brighter lights or dimmer lights, more or less breaks, seemed to increase worker productivity for a time. The mystery was resolved by sociologist Henry Landsberger, who suggested that it was the novelty of being observed and made to feel important that boosted the workers’ output, not the specific change in conditions
It was named the Hawthorne effect. Essentially, people’s knowledge that they were being watched altered their behavior. The workers, consciously or unconsciously, worked harder because they felt they were being observed.
The experiment was reflexive: the study wasn’t merely measuring productivity, it was changing it, by the act of measuring. Modern psychology and medicine take this into account; for instance, in clinical trials, participants often improve just because they are being monitored and cared for (a blend of Hawthorne and placebo effects). The Hawthorne studies teach us that observers must be wary: to observe is often to perturb.
Reflexivity and Science
Since the advent of the scientific method, we’ve been able to find regularities across countless domains in nature. By zooming in on the really small, zooming out to the really big, or removing anthropocentric bias, we have been able to detect robust patterns that hold true across space and time, regardless of belief. Most of these physics-based dynamics do not have a reflexive component. Partially, because they are simple. They are not complex enough to be reflexive.
The fact that we have produced verifiable ‘laws’ regardless of when, where, or who did the testing, is a sort of miracle. Gravity is the same no matter who, what, where, or when it’s tested (outside of Einsteinian exceptions). There’s something inherently beautiful about this, because we can say it about such few things.
In fact, the limits of the scientific method are the areas in which these global ‘regularities’ do not exist, or are difficult to test for.
You can see where this is going: reflexivity breaks the scientific method.
In fact, this has been the primary struggle of modern science: our tools of measurement are not sophisticated enough. Using the scientific method to reason about a non-linear, reflexive systems is like trying to use a thermometer to predict the political climate of the country you’re in, it’s just not that simple.
Science has hit a wall in biology, quantum physics, social systems, psychology, cognitive science, economics, and more. Systems in which the act of measuring or observing the system effects the systems. Systems where simply thinking about the system can change it. How does one even begin to formalize these types of systems, where the observer and the observed are intertwined? How does one find the regularities that allow us to speak confidently or model such things, as separate from ourselves?
Physics hit wall whenever physicists were no longer able to reason about the universe as an isolated, non-reflexive system in which human observer didn’t matter (physicists don’t like thinking about humans or themselves). Quantum physics stumped them because the act of observing a quantum particle changed it’s state.
As soon as physicists had to think like biologists (embodied, dynamical, strange feedback loops), they stopped making progress.
Enter AI. AI finally gives us a chance, it feels like it was born for this.
References:
https://www.ft.com/content/0ca06172-bfe9-11de-aed2-00144feab49a?ref=danreardon.com
https://www.ft.com/content/0ca06172-bfe9-11de-aed2-00144feab49a?ref=danreardon.com
https://www.ft.com/content/dbc0e0c6-bfe9-11de-aed2-00144feab49a
https://thephilosophersmeme.com/2022/04/04/homesteading-the-memeosphere/
https://www.cairn.info/revue-de-philosophie-economique-2013-1-page-29.htm
https://x.com/upskillyourlife/status/1845321152308429084?s=46
This is brilliant. I've been thinking carefully about reflexivity (calling it "self-reference") for about a decade and found this article really clarifying. I hadn't at all thought about there being kind of a reflexivity market in terms of low-hanging fruit for affecting complex systems with belief. Lots of new stuff here for me, but that part in particular stands out for me.
I think science CAN tackle reflexive systems by the way. I like David Deutsch's approach to defining science, basically saying it's about coming up with explanations that could be proven false and then checking if they're wrong. It's just that, as you note, non-reflexive systems are much easier to come up with stable explanations for. So those are the domains that science has focused on so far. We usually call such things "objective". Then science fails when it hits reflexive topics because its practitioners keep wanting to use objective measures and methods on them, which fail for spectacularly obvious reasons if you're used to thinking in reflexive terms.
Basically anything having to do with subjectivity is like this. If we're coming up with theories about the nature of the self, for instance, then the act of understanding and stating the theory is probably part of the domain the theory is supposed to explain. You can't factor the scientist out of that process the way you can factor out lots of reflexive stuff in biology via an RCT. You sort of have to bite the bullet and just let your theories be kind of self-aware and check whether their self-reference makes sense and is stable (e.g. less like "This sentence is false" and more like "This sentence is true").
I've been on about "subjective science" exactly because of this point. We can do real science in reflexive domains. It just requires being willing to TRY, and to develop new methods. And that lets us tackle things like wholesomeness and wisdom — the really precious stuff that objective science has to ignore.
Thanks so much for writing this post! Very thought provoking for me.
I think the idea of reflexivity is still underrated. I wrote this piece to put it closer to the center of how I see the world, not just how I see financial markets. Unsurprisingly, the concept of reflexivity, is itself, reflexive. The more we understand it, the more prevalent it becomes.