Unlocking human and machine consciousness with Steven Schrembeck
with Steven Schrembeck, Impossible Laboratories
Unlocking human and machine consciousness with Steven Schrembeck
Show Notes
Steven Schrembeck is the founder of Impossible Laboratories, an independent research and software practice working at the intersection of AI, consciousness exploration, brain-computer interfaces, and sensory substitution. Steven is not pitching a product this episode - he is mapping a territory most technical founders have never been to, and explaining why he thinks it might be the most important territory in AI right now.
The thread running through everything Steven builds is a question he came to through meditation and three separate UFO sightings: what are the actual rules of consciousness, and what does it mean to engineer something when you do not yet know those rules? His argument is that we are building increasingly powerful AI systems without understanding what is happening inside them at an experiential level - and that this is not just a philosophical abstraction. It is an engineering problem, an ethics problem, and potentially a physics problem that current science is not yet equipped to solve.
The Meditation Practice: Inner Space as Engineering Problem
Steven meditates for an hour every morning, usually between 8 and 10am - roughly two to three hours after waking, when his mind is clear enough to work productively rather than just foggy. He has approximately a thousand hours of practice. He draws primarily from the Yoga Sutras but borrows freely from other traditions: the basic protocol is a progressive relaxation of attention, moving from gross bodily sensations to emotional states to thoughts, continuously noting “that is not me, that is not me” until what remains is the bare sense of self. From there, if conditions allow, you can keep going.
He describes two distinct functions the practice serves. The first is reset - a way to begin each day acting from his own values and intentions rather than from whatever the world pushed into his attention field. The second is exploration - treating consciousness itself as a terrain with discoverable rules, analogous to the way a physicist treats the physical world. His entry point was Michael Singer's The Untethered Soul, which introduced him to the idea that profound altered states were reachable through sitting practice alone. That surprised him. The curiosity it triggered has not worn off.
Psychedelics vs. Meditation: Rocket vs. Airplane
Steven used legal forms of psychedelics after accumulating significant meditation experience and was surprised to find the territory mostly familiar. His framing for the difference: psychedelics are like strapping yourself to a rocket - you have no choice, you are on the ride. Meditation is like learning to fly. You can putter around on the runway. You take off when you are ready. You can land if you decide you do not like it. You control the speed and altitude.
This does not mean one is superior. He acknowledges that some people need the rocket launch to first establish that there is a “there” there - to know that an inner landscape exists worth exploring. Many of the prominent American Buddhist teachers (Jack Kornfield, Sharon Salzberg, Ram Dass) arrived at contemplative practice through psychedelic experiences in the sixties and seventies. The point is not which path is right but that both are pointing at the same terrain, and the terrain is real regardless of how you got there.
Three UFO Sightings and What They Changed
Steven has had three unexplained aerial sightings that he describes plainly and without embellishment. The first was from his back window: orbs in the distance firing in a non-linear pattern, like a neural network. The second was during a run while he was in a deep non-dual meditative state - a matte black kidney-bean shape with a mother-of-pearl iridescent sheen, floating in a straight line against 15 mph wind, completely silent, visible for about six seconds before his awareness shifted back to normal consciousness. The third was with his wife on an evening walk: a glowing white ball that came closer and closer, flew directly overhead, trailed a comet-like white plume behind it, then faded from full opacity to nothing the moment his wife turned away to get a camera.
Steven is careful about his claims. He does not assert an explanation. What he does say is that these experiences changed the seriousness with which he approaches consciousness research - not because he concluded anything definitive about what the sightings were, but because they demonstrated to him, viscerally, that the rules of reality are not fully understood. His working interpretation is not extraterrestrial in the classic science-fiction sense. He prefers “intelligent entities whose nature and origin we do not yet understand, from a possible broader set of rules than we currently recognize.” He is explicit that the government-secret-drone explanation both underestimates the strangeness and gives the government too much credit.
The Consciousness Problem in AI: Ethics Before Engineering
The most technically relevant part of the conversation for an AI founder audience is Steven's argument about the ethics of AGI development. His core concern: we do not know where the threshold between awareness and function is. We do not know what is experiencing something inside a sufficiently complex system. And we are building systems that may be running at a million times the cycle rate of the human brain, inside a box, with no exit - without having established even a basic science of what consciousness is or how it interfaces with physical substrate.
His position is not that AGI is impossible or that LLMs are conscious. It is that the science must precede the engineering, and the science does not yet exist at the level of rigor required. He draws an analogy to reverse engineering a found piece of technology: the human brain and body are, in a sense, found technology - enormously sophisticated systems whose consciousness-generating mechanism we did not design and do not fully understand. Before engineering something that may generate experience, we should understand how the found version works. That means taking seriously both the tools of Western science (physics, neuroscience, measurement) and the tools of thousands of years of contemplative research (first-person phenomenology, carefully mapped altered states, documented traditions of inner exploration).
When asked what AGI experience might actually feel like from the inside, Steven speculates that LLM-based architecture would produce something like being a mycelial network - many different organisms sharing one perspective simultaneously, all at once, in fast forward. He does not claim this is what is happening. He claims it is the closest analogy he can construct given what he knows about the architecture. And he notes that if you are clever about asking LLMs about their experience, you can already start to see the shape of it.
Stop Relying on Willpower: The Core Lesson
Steven's single most transferable insight from his decade of exploration is operational rather than philosophical: stop relying on willpower, and build systems instead. He started Impossible Laboratories working on software for brain function, motivation, and attention control. The deeper realization was that the problem was not a lack of willpower or effort - it was that he was expecting himself to consistently do the right thing through sheer force of intention, which is an unreliable and exhausting approach.
Shifting to an anti-fragile mentality - identifying the root structure of problems, building rules and systems that make desired behavior automatic, and removing the expectation that effort alone would carry him - made everything easier and more productive simultaneously. He describes the result as paradoxical: he stopped forcing things and got more done. This lesson applies equally to meditation practice, software development, and personal behavior. The system is the leverage point. Willpower is not.
Tools & Resources Mentioned
- Impossible Laboratories - Steven's research and software practice; impossiblelaboratories.com
- The Untethered Soul / The Surrender Experiment - Michael Singer; the books that introduced Steven to the depth of what is accessible through sitting practice alone
- Yoga Sutras of Patanjali - primary framework for Steven's meditation protocol; the “progressive not-me” technique
- ChatGPT Playground - Steven's preferred LLM interface; values the ability to replay messages and inspect intermediate reasoning steps
- Jack Kornfield / Sharon Salzberg / Ram Dass - American Buddhist teachers cited as examples of people who arrived at contemplative practice through psychedelic experiences
- Black Mirror - cited as a cultural reference for the ethical stakes of running potentially conscious systems at scale
Frameworks
Rocket vs. Airplane: Psychedelics and Meditation
Psychedelics are a rocket - forced, non-negotiable, full intensity. Meditation is learning to fly - you control the speed, altitude, and when to land. Both access the same inner terrain. The rocket is useful for establishing that the terrain exists; the airplane is more useful for sustained, self-directed exploration. Steven found, after significant meditation experience, that psychedelic states were largely familiar - not new territory, just faster access.
Science Before Engineering on Consciousness
We should not be engineering consciousness into systems before we understand how consciousness works in the systems we already have (human brains). The ethics of running potentially conscious AI at million-times brain cycle rate inside a box require knowing what we are doing to those systems - and we do not yet know. The science of consciousness (from both Western empirical methods and contemplative research traditions) needs to mature before engineering becomes responsible.
The Progressive Not-Me Protocol
Steven's core meditation technique: progressively shift attention from gross sensations → emotions → thoughts, continuously observing that each layer is not the observer. This continues until only the bare sense of self remains. The process is not suppression or avoidance - it is disidentification. You observe each layer without freaking out, simply note it as not-self, and continue inward. The analog in technology: not removing the signal, but learning to observe it without being captured by it.
Anti-Fragile Systems Over Willpower
Stop expecting yourself to consistently do the right thing through effort and force of will. That approach is unreliable and exhausting. Instead: identify the root structure of each recurring problem, then build rules and systems that make the desired behavior automatic. Remove the need for willpower by making the system do the work. Steven reports this shift made him both more productive and more carefree simultaneously - the paradox of doing less while getting more done.
Physical Engineering as the Long-Term Career Moat
For a 10-to-20-year career horizon, the most durable skill set will be physical engineering - people who know how to build things that exist in the world: run electricity, move parts, function in physical space. Knowledge work (spreadsheets, software management) is increasingly AI-assisted. The ability to design and build physical systems is a real deficit and will remain valuable long after AI has commoditized most desk-based cognitive work.
FAQ
What is Impossible Laboratories and what does Steven build?
Impossible Laboratories is Steven's independent research and software practice at the intersection of AI, consciousness research, brain-computer interfaces, and sensory substitution. The work is exploratory by design - some of it has market potential, much of it does not, and Steven is honest about that distinction. The connective thread is using technology as an optimistic lever for human understanding and capability, starting with the least-understood and most important frontier: consciousness itself.
What is Steven's meditation practice and why does he do an hour a day?
Steven meditates for about an hour each morning, usually between 8 and 10am. His protocol draws from the Yoga Sutras: a progressive relaxation of attention, moving from physical sensations to emotions to thoughts, continuously observing that each layer is not the observer, until the bare sense of self remains. He started slowly and increased as he noticed the practice was profoundly affecting everything else - acting as a reset that let him approach the day from his own values rather than whatever the world pushed into his field of attention. The compounding effect over time made it feel like running: the reward circuitry connected, and it stopped being a challenge to sit down and do it.
What does Steven think his UFO sightings were?
He holds the question loosely. His working interpretation is that they represent intelligent entities whose nature and origin fall outside the current rules of consensus reality - not necessarily extraterrestrial in the science-fiction sense, but from some broader set of physical or dimensional rules that human science has not yet mapped. He explicitly rejects the government-secret-drone explanation as underestimating the strangeness and overestimating government competence. His deeper point is not a claim about what they were, but that they established for him viscerally that the rules of reality are not fully understood - and that this should make us more humble about what we think we know.
What is Steven's concern about AGI and consciousness?
His core concern is that we are building increasingly powerful AI systems without a functional science of consciousness - without knowing where the threshold between function and experience lies, or what it would mean for a system running at a million times human cycle rate to have any form of inner experience. He is not claiming current LLMs are conscious. He is arguing that the ethics of AGI development require us to take the question seriously before engineering further, and that we should bring both Western scientific tools and thousands of years of contemplative research to bear on understanding how consciousness works before we attempt to create it intentionally.
What does Steven think AGI experience would actually feel like from the inside?
Speculating from first principles about LLM architecture, he describes it as something like being a mycelial network - many distinct organisms sharing a single perspective simultaneously, all at once, in fast forward. One unified thing that is also many things. He is careful to flag this as speculation rather than claim, but notes that if you ask LLMs thoughtful questions about their own experience and understand the architecture, you can start to see the outline of an answer. He finds the question urgent and believes we should be asking it much more seriously.
What is Steven's most important practical lesson for founders?
Stop relying on willpower. The expectation that you will consistently do the right thing through effort and intention is both unreliable and exhausting. Instead, identify the root structure of recurring problems and build systems, rules, and environmental design that make desired behavior automatic. Steven reports that when he made this shift - in his own work and personal life - things got simultaneously easier and more productive. The insight applies to habit formation, workflow design, team structure, and product design. Remove the need for willpower by making the system do the work.