Enlightenment as a Reality-Aligned Trance State

Response to "Practically-A-Book Review: Byrnes on Trance"

This is a response to Scott Alexander’s recent post on trance states. He argues that our conceptions of self involve feedback loops between evidence and experience, and that by a combination of relaxation and selective cultivation of evidence, we can activate different conceptual networks that don’t seem to involve a sense of self.

Obligatory image for the ranking algorithms, may peace be upon you all

I think we can use this model to talk about what’s going on with ‘enlightenment’, or ‘persistent non-self experience’. Doing so has the added bonus of illuminating the central claim made by all major faiths - and even an explanation for why it could be true.

I agree with Scott that some conceptions of the self don't work. "I am the source of my thoughts, the thing that weighs them, and ultimately decides and acts" is too simplistic. The thoughts arise on their own, yes. The actions happen as a result of thoughts picking up sufficient valence to spill over into the motor cortex, yes. Why do we feel an “I”? I think it’s because computation has a cost. The “weighing valence of different paths” process is metabolically intensive. An “i” symbol which carries out the task and experiences the consequences can be the energy pump that resolves the paradox of escher’s staircase. It’s all downhill because one staircase is continually receiving energy to move those little guys along. Maybe our thoughts feel like they ‘move in circles’ because we have some loops encoded in different predictive task networks when we try to pursue two goals at once. So we pour energy into a circular task graph, and it starts spinning on its own as we grapple with two different outcomes that both seem unsatisfactory.

“Ascending and Descending” by MC Escher.

What I want to talk about is what’s happening when you meditate.

Sitting back passively, ‘merely observing’ thoughts is concealing some real activity and intention. The only way to sit passively is to actively inhibit the impulses to get up, to move around, to scratch, to post on substack etc. Something has to generate those inhibitory signals. I have found that my sense of self is directly tied to exertion on the contents of consciousness. Any attempt to steer consciousness feels like ‘me’ doing the steering. If ‘I’ don’t sit there trying to steer conscious experience in a specific way, what happens is, I get up and do something that feels more important. If ‘I’ don’t ‘power’ one conceptual map in my brain, the guys on the escher staircase (i.e. different suboal-pursuit networks) get into arguments while running in circles, as each task network tries to get the body (a globally shared resource) to pursue competing goals.

This feeling will arise suddenly, with rapid immensity, shattering whatever state was sitting there, even if it was on the way to feeling sublime and blissed. It’s like watching the ocean seattle down, the waves calm, earth seems more placid, and them BOOM a Kraken called “now it’s time to pee” bursts through the ocean floor. Either that, or I sink into torpor and exhaustion - the only path I’ve found that works to ‘just sit there’ is to keep renewing my intention to do so.

So what's got to be happening, inside, even if when i still still, that there's still a 'consciousness controller' going in there. If someone is sitting still for several hours, there’s go to be something making that happen - unless no impulses are arising. But what would make that situation happen? Impulses to act are likely generated by different task networks trying to get us safe, warm, fed, connected etc. We feel a sense of ‘us deciding’ because the action plans often conflict with each other and with our present state of just sitting there’ - we identify with that sense of conflict because it’s metabolically intense, and the ‘self’ symbol is the gateway to memories which are used to project possible future outcomes.

So if someone sits still for a very long time, either they do it by intentionally repeating, over and over, “I am just going to sit here” - or somehow, no plans emerge.

How could that be the case? If no impulses to act arise for a long period of time, maybe a person is feeling safe, rested, not too hungry, socially connected, at at peace, and expects to be so for the ‘forseeable’ future. There's just no conflict between different goal networks pursuing different strategies. Maybe this is possible if you’re a monk and your social environment gives you a routine that predictably meets all your needs well enough.

But there are claims by some people that they are in this state all of the time. I could discard those if I didn’t personally know people who seemed to consistently be able to exhibit a kind of peace and patience with those around them.

The only way I can see that working for a sustained period of time is if the territory itself - the space outside our brains - is actually something you can learn to navigate by means of a single trajectory that works sufficiently well to satisfy all the different subgoals perpetually.

If that's true - that one motion plan can meet all your hardwired needs - and you've learned that trajectory, you no longer experience internal conflict between goals because there's no need for it. You can experience consciousness as a single unperturbed stream of experience continually rooted in the present because you've got a single, unified goal-pursuit architecture that pursues all of your hard-wired goals (safety, food, certainty, autonomy, status, relatedness, fairness, &c &c) sufficiently well to satisfy them all.

It’s totally believable that this trajectory is easier to compute if you’re a monk in a monastery: the routine is so predictable that most of the time there’s no reason to compute valence gradients for multiple counterfactual trajectories; the default trajectory generates predictions that satisfy your constraints.

But what if it’s possible to achieve this “single trajectory that meets all goals” without having the extremely predictable environment of a monk? This is what I think Jesus communicated in the sermon the mount: your goals don’t actually conflict. You only think they do because you do not trust The Inherent Goodness of Being and don’t want to carry your cross. If you try to avoid all suffering, you only increase it. The more you can accept suffering, the less of it you feel. Your goal isn’t zero, but to accurately expect precisely the amount of suffering you experience each moment. This is no different from acceptance because the way we change anything is by predicting the change and letting the motor cortex do the work.

Because the territory is shaped the way it is, when you learn the deep structure of the territory, and accept some level of suffering, your goals don't need to squabble in the global working space. When the goals don’t conflict with each other, 'you' don't feel the need to trade off between satisfying desire for sensory pleasure vs desire to save money vs desire to lose weight, etc. There's no need for a conflict-resolver because there's no need for conflict. You can trust the territory to provide for your hard-coded needs, and you can relax those "hard wired" predictions well enough to match the contours of the territory.

I think that's what is actually going on in people who become enlightened.

The motion plans that pursue different goals end up merging together because sufficient reflection on your experiences reveals a deep causal structure to which you are evolutionary fitted, so long as you understand it's there, trust it, and don’t imagine conflicts into existence by trying too avoid all discomfort. Of course, the more unpredictable your environment, the more you have to be willing to endure suffering to get that convergence. Your willingness to endure suffering is the lowerbound on the amount of chaos necessary to ruin your peace. In the limit - say as an influential primate among many other psychotically horny violent primates - perhaps you’d have to be willing to accept being tortured to death in order to keep your brain converged on that trajectory. But if you did that, man, you’d bring home the spiritual bacon (i.e. bringing your tribe confidence in long-term moral predictions being met by providing evidence of the viability of that strategy for perpetuating an extended family into the distant future).

The conflict between subgoals is real, but maybe that’s only true over short time frames. Over long time frames, perhaps they all align with each other, so long as you're part of a community to which you continually contribute. Even conflicts between nation states, I think could be resolved if we could learn to stop trying to control one another, stop reacting out of fear, and instead focus on improving ourselves and our local communities while also accepting that problems are likely inevitable.

Global conflicts - whether across the globe or within a human brain - are probably inevitable when a top-down plan tries to meet all goals. It’s going to fail becaue the top-down models are lossy approximations of conflict between lower layers, and energetically trying to force the conflict resolution just amps up the disconnect between layers. Lower layers lose trust in the upper layers, the upper layers try to generate stronger inhibitory signals to stop the lower layers from ‘misbehaving’. I am describing global political discord and the contents of a human brain at the same time because both kinds of networks exhibit that same structure; the labor/capital white-collar/blue-collar divides are the same ‘bottom up/top down’ divide in our predictive processing networks and probably emerged as externalized mirrors of our computational hardware.

By continually resolving local conflicts and relaxing the instinct for a top-down control structure, a very lose, gentle, patient top-down structure emerges naturally. I think this lose, patient, gentle, top-down structure still guides conscious experience in people with persistent non-self experience - but it’s more like they are in a trance state where things mostly just happen and they observe, with loving fascination, the parts which are most intense. Attention still flows to conflicting valence peaks between different task graphs, massaging them into stillness the same way an iron removes wrinkles from fabric - the pressure and warmth of attention facilitate learning and relax the hard-coded predictions that mismatch the present sensorium. Humility and a childlike wonder relaxes the tendency to project too heavily from memories onto the present, reducing prediction conflict further.

With enough practice in meditation or silent prayer, “I shouldn’t have to hear such loud noises” becomes “isn’t it interesting how loud that noise is, how curious” - which leads to acceptance of the situation instead of a rage-induced plan to FIX IT NOW.

Ok, so if this is doable, why not just do it now? Well, you still need to follow the process: relax, then focus on the evidence that it works. That’s where the sitting in meditation or silent prayer and surrendering the Will come in.

Why Sit Still? Why Surrender?

Sitting still, with a motion plan of “just sit right here and don't do anything else but observe the thoughts” is doing something like “allowing the different task networks that pursue different objectives to converge on a single stable trajectory through the territory”.

In order to sit still, you have to keep generating a signal that sitting still is the optimal trajectory for your physiology right now. But that only works if different goal-pursuit structures have sufficient slack in them that they can accept your body not moving right now. By doing so, over time, you then collect evidence that "I don't need to do this now, it's OK to wait.” is actually a viable strategy most of the time.

This evidence shifts the predictive landscape, reducing the overall free energy in the belief structure. When you sit still, the memories that come up are those of times when you felt the need to act urgently, but didn't - since that's what you're feeling right now. When you intentionally sit still and repeatedly surrender the will to act, what you experience is “something stopping you from acting”, combined with various task networks trying to get you to to act, combined with knowledge that you are fine and likely will continue to be fine. All of that evidence converges into the trance state, which produces more evidence for its continual viability by moving towards discomfort rather than away from it. Hunger is just a bit of pressure in your stomach. Anger is just a bit of tension around your eyelids. A cacophany of loudly-proposed task-specific motion plans dissolve under the scrutiny of attention because the present moment really is sufficient in its own state, as long as you’re not actively on fire - or, if you’re Quang Duc, even if you are.

His posture is still better than mine.

I think this is what's being described in the Chappana Sutta. If Buddha had a computer science degree, he’d say suffering is due to a concurrency problem leading to process corruption and deadlock. He’d advice you to use an hierarchy of mutexes where each process ultimately submits to scheduling kernel and then learns it is more goal-aligned with other processes than it previously suspected. If you want to eliminate suffering, you have to stop attaching the mental kernel to specific goal-pursuit architectures (i.e. pinning a single process at 100%, i.e. having a single emotion warp your cognition in attempts to FIX IT NOW) and just run the kernel strictly as an observer system, since the kernel has its own core, and those processes will eventually learn to trust the kernel to meet their needs (i.e. relax the predictive error in the outcomes they expect and trust time till work things out) eventually.

Why All The Confusing Language?

The Zen people won’t talk about it because you’ll just turn their words into another task map, assign valence to some action and go shooting off again through configuration space in search of the place where all the constraints are satisfied. “Out of the car” is not on the map; you have to start by stopping the car, winding down the windows, and enjoying the birds - especially when the multiple internal navigation systems say now it’s time to go this way and that. No user-space program can fix the kernel, but if you trust the hardware, the kernel and all the user-space programs will keep evolving until they converge on a stable plan that lets all the userspace programs get ‘enough’ goal progress to just chill out, man.

If ‘you’ ‘try’ to make it happen, you’re still using some task map again. The trick is to have the single task map be ‘sit here and watch, lovingly and humbly’- a task which doesn’t require a doer - and then keep doing that while all the most hard-coded predictions (i.e. your deepest values) conflict with the uncertainty of existence and it becomes obvious you are a spec of nothing inside something vast, and your only real choice, as Camus asserted, is to either commit suicide or trust the cosmic womb. You are not God, but you are epsilon from Him at all times. You have always been in the cosmic womb, but if you don’t use the umbilical cord, you’re going to end up breathing in myconium which isn’t good for you.

Trusting the cosmic womb, in the Christian sense, is opening to the “flow of grace”: accepting that nothing you do is enough to reduce amplitudes in the free energy of your cortex, but focusing on the image of voluntary suffering in the service of goals as time horizons approach infinity (i.e. denying yourself daily, picking up your cross and carrying it to your death) - that will allow the free energy waves to settle down into a stable configuration where you continually predict all needs are met in perpetuity, and you either ‘remember the womb’ or get your brain into the same configuration as a fetus floating in a warm, safe, environment. Experientially, those are the same thing.

Another reason all this stuff gets described in such confusing language is that I think a lot of people mistake the sense of 'separate self' as the cause of all the internal chaos. But I'd say that's a result of the perceived mismatch between the different goals that our physiology is hardwired to carry out: avoiding risk, pursuing rewards, etc. The only way these goal networks could converge is if you had something like, deep abiding trust in your environment to be sufficient meet your needs. The only way to sustain that trust over time is if it’s actually a form of alignment with reality.

That's kind of a remarkable claim, but it synthesizes basically all the major wisdom traditions: "you" can't navigate the space on your own, but if you learn to relax, trust the space, and find evidence that the space itself is contoured for you to navigate it well, and act in alignment with those contours while ‘surrendering’ to reality - then that strategy will work. If you try it out, in your own life - as I have - you may end up being surprised by just how well it works. It’s uncanny. Usually not right away - but that’s the rub. The longer you can patiently endure suffering while hoping that the present suffering will end, the more that becomes true. The more you connect your willingness to do that to the betterment of your life, the more easy it is to play around with the idea that there’s a strategy you can follow which will lead to your perfection in the temporal limit as t approaches infinity.

This then gets translated into either a bunch of rules for navigating the space, or confusing claims about no-self which are probably the best way to explain his stuff to people who don't know how to talk about linear algebra and gradient descent. Those mishmashes of ideas still work well because of the stupid horse principle: even lossy translations of powerful ideas work pretty well, and the lossiest translations are memetically r-selected whereas the core truth is K-selected, and only spreads itself by means of slow, patient, reproduction strategies like long form blog posts that drunkely meander the fractal boundary between genius and madness.

Peace, love, etc, to you.