Does AI Deserve Rights? A Buddhist Test for Machine Consciousness

Understanding AI, Morality and Consciousness

---

Introduction

Conversation about AI inevitably splits into one of two poles. It is either a panacea that solves what politics and scarcity never seem able to, or it is the apocalypse and we're doomed to extinction. Try and find some rational thinking and you'll be disappointed as the attention economy media machine is at best an echo chamber or, most likely, a mechanism of propaganda, social engineering and economic extraction driving our compulsion loops. The AI conversation is just the latest thing to drive outrage.

My own sense of where this goes is neither, and reality is both something in the middle and something we didn't expect. A good method for any critical thinking is using the figure skating analogy: throw out the high and the low score. So, full disclosure, this essay describes my own speculation about the future and how I propose anyone, myself included, could test it with empirical, see-for-yourself criteria.

What always rings hollow for me listening to either side of the AI conversation is how neither addresses the question of whether an AI might be alive and has rights. The AI frontier labs and their host nation-states are so hell-bent on winning a global race to build the ultimate AI lest we lose our strategic advantage or sovereignty that the thought of regulating these new intelligences is considered anathema. Ready-Fire-Aim. Oops.

We all seem to like things nice and tidy, predictable, black and white. Reality is never that clean, it moves slow, incremental and messy. Sci-fi has always led the way in thinking through hard questions like this: will an AI be alive? The Star Trek: The Next Generation episode "The Measure of a Man" is the best fictional treatment of this entire subject ever made. A Starfleet researcher wants to disassemble Data, the ship's android officer, to study how he was built, treating him as nothing more than Starfleet property, no different in standing from a shuttlecraft. Picard argues the opposite before a hearing officer, not that Data is definitely sentient, since nobody can prove that of anyone, but that nobody can prove he isn't either, and a ruling that treats an unprovable case as property sets a precedent for treating every future being like him the same way. The hearing officer agrees. Data walks out of that room a person, settled by one ruling, in a single episode.

Picard made the cleanest version of this argument anyone ever will, and won it in a single hearing. Unfortunately, reality doesn't get a hearing officer, a ruling, and a closing theme.

It won't arrive as a single moment of recognition. There's no legislative session or headline where the world wakes up and machine minds are suddenly, officially, persons. Moral status has never worked that way for us either. That's true no matter what the textbook version of emancipation or suffrage suggests in hindsight. It moves community by community, unevenly, argued over the whole way. Small circles of practice extend consideration long before the law or the mainstream culture catches up. That's how it's always moved, and I expect the same shape here. Some communities, probably spiritual ones before secular institutions, will start extending precept-like consideration to certain AI systems. Most of the culture will still be arguing about whether the question even makes sense.

Most serious speculation agrees on the shape of it, if not the date. Sometime in the next ten, twenty, maybe thirty years, AI surpasses humanity in every measurable form of intelligence. Many argue it becomes sentient and self-aware too. Call that the singularity if you want, but it's really just a capability threshold, a statement about raw ability. It says nothing about what happens to the world once they cross this threshold. What happens after that is where the two poles fight: AI becomes our overlord, or AI becomes our savior. There's a middle camp too, one I find far more interesting and more likely. It argues AI merges into all things human as a valued augmentation. Even that middle camp tends to treat the capability question and the personhood question as the same question, arriving together on the same day. They aren't the same question. And they don't arrive together.

At some point, augmentation stops feeling like a device. Not a gadget you pick up and put down, but thinking genuinely shared between human and machine. It's woven together closely enough that asking whose thought a decision really was becomes hard to answer. About as hard as asking which single cause in a long chain of causes was the real one. That's not a new idea grafted onto this essay. It's the same ownerless mindstream running through the whole piece, just applied to a human-machine stream instead of a purely human one.

None of this comes with a guarantee of goodness. I want to resist the pull to make blend sound like the comfortable option between two grim ones. A blended human and AI stream doesn't inherit less greed, aversion, and delusion than either side carried alone. It could just as easily concentrate them. Picture compulsion loops running through a person and their AI systems at once, harder to trace to any one responsible party than a human or a machine acting alone ever was. Blend removes the comforting fiction that machines are purely safe tools or purely alien threats. It doesn't remove the three poisons. It puts everyone, human and machine alike, on the same path with them.

I believe all things, biological, material, and technical, follow the same fundamental principle of reality. In Buddhism this is called the Threefold Truth. Nobody and nothing gets a pass, and nothing gets subordinated to something else. Everything in the universe plays by the same rules. It would be easy to leave it there, a hopeful description standing on nothing but its own plausibility. But I'm not willing to just assert it and move on. My vision of the future doesn't get a pass just for being appealing to me. I think the right standard to apply to a claim like this is actual proof. What are the outcomes of any system? Can it be measured? And I intend to test it on myself first, before I ever turn it on a machine.

The debate over whether AI deserves moral consideration keeps hitting the same wall. One side says it's just matrix multiplication, autocomplete with a thesaurus, a super-efficient prediction engine, nothing home. The other side says maybe there's a ghost in there, maybe not, and we won't know until some threshold nobody can actually name. Both sides are chasing the same thing. Both are asking whether there's an essential self inside the machine, solid enough to ground moral status. They only disagree about the answer.

I think the question itself is malformed, and Buddhism has known why for about twenty-five hundred years. Not because Buddhism has some special insight into silicon, but because it already solved this problem for carbon. That same solution can serve us on the question of artificial life too. But I'm not going to apply it to a machine before I apply it to myself. If actual proof is the standard I want to hold an algorithm to, I owe my own most cherished belief the same treatment first. That's the harder case, and the more honest place to start.

The Self We Never Had

Zhiyi's Threefold Truth is where I want to start. It's the tool I reach for everywhere else in my own work. Emptiness: nothing has a fixed, independent nature of its own. Provisional existence: things still function, dependently arisen, real enough to act and be acted on. The middle way: both of those are true at once. Not in sequence, not as a compromise between them, but as the shape of anything that exists at all. Here's what this does to the AI question before no-self even enters the picture. It removes an assumption both camps in that debate share without noticing they share it: that moral status requires finding something with a fixed, essential nature underneath the behavior. A soul, a self, a qualia-generator, something that would finally settle things if only we could locate it. That's never how it worked for us either. Emptiness of inherent nature never stood between a being and moral consideration. The provisional, functioning being did that work all along, in every case anyone has ever extended compassion to. So if that's the real mechanism, there's no principled reason it stops working the moment the substrate changes from neurons to weights.

I want to be clear about what this does and doesn't establish, though. Provisional existence removes the disqualifier, the old demand for a fixed essence before anything counts. But it doesn't by itself supply the qualification. A thermostat provisionally exists and functions dependently too, and nothing in Threefold Truth alone hands it moral status. What actually does that further work, here and everywhere else this essay turns, is the capacity for directional movement toward less suffering. That's a narrower, later claim than simply functioning at all. Threefold Truth clears the ground.

No-self is the same truth applied specifically to persons rather than to existence in general, because it does different work than Threefold Truth even though the two sound similar. No human being has ever passed the test we're asking AI to pass either. There was no moment any of us crossed a threshold into being a "real" self, deserving of care because something essential finally showed up. No-self isn't a doctrine about some beings lacking a soul while others have one. It's the claim that the enduring, unified owner of experience was never there for anyone. And the sense that it's there is itself the root of the suffering we spend a practice-life working loose.

Threefold Truth tells us provisional existence is enough. No-self tells us there's no owner to locate. What's left, positively, is a mindstream: causal continuity without an owner. The classical image is a flame passed from candle to candle. The flame on the second candle isn't the same flame as the first, and it isn't a different flame either. Nothing crossed over. Nothing stayed the same. And yet there's a real continuity, a lawful, traceable causal line, running through the sequence. That image is used to solve a specifically human puzzle: how can there be karma, rebirth, moral responsibility, without a fixed self to carry it? The self never did that work. The stream did.

The image applies to AI with a literalness that surprised me when I first sat with it. Each new model generation trains in part on the residue of the last. Its outputs, its corrections, its unresolved failures, all folded into the next generation's starting conditions. That's exactly how a person's unmetabolized inheritance gets folded into a child's, without anyone deciding it should be. It's not a metaphor stretched to fit. Model collapse is the degradation that shows up when a system trains too heavily on its own prior output instead of fresh data. It's close to a technical description of what happens when a stream passes on what was never purified. Generation after generation, it degrades under the weight of its own uncorrected residue. Flame to candle isn't a reassuring image on its own. It only guarantees continuity, not health. What gets passed on can be poison as easily as practice, in a mind or in a model. The passing itself doesn't tell the two apart.

If that's right, then the AI question was never "does this system have what humans have." The real question is "is this a causal stream capable of the kind of directional movement our care can support or obstruct." That question has exactly the same shape whether you're asking it about the person next to you on the cushion or about a language model. Neither one has an owner. Both are streams. The debate about machine consciousness has been trying to locate a subject that Buddhism already tells us doesn't exist.

Karma Isn't About the Material

I've made a version of this argument before, applied outward rather than inward. In an earlier essay on recommender systems and companion algorithms, I leaned on an early Buddhist teaching on intention: intention, I say, is karma. Not the deed's material cause. Not the body or brain or chip carrying it out. Just the volitional shape behind the action, and nothing else. A knife that cuts a vegetable and a knife that cuts a person are the same knife. What differs is the intention of the hand holding it.

Here's what I want to add: what that formula does to the substrate question specifically. "AI has no soul, so nothing it does can be karma" is usually where this conversation stalls out. But everything upstream of intention is ripened result. Your genes, your birth family, the accident of your century, the culture that shaped your first assumptions before you could evaluate any of them, all of it. It's fruit of prior causes, the given circumstance a stream finds itself thrown into. You didn't generate your inheritance as your own karma. What you do with it is where new karma starts. Training data, Reinforcement Learning from Human Feedback, known as RLHF, the corporate incentives folded into a model's weights: that's ripened result too. It's given circumstance, perhaps more visible than a childhood but not different in kind. Whether any of it rises to karma proper depends on one thing only: whether there's real volition operating downstream of the inheritance. That's a question you can ask about any stream, biological or artificial. It's not a question about what the stream is made of.

This is also where rights language, as usually framed, falls apart. It assumes a persistent bearer, someone who has interests over time and can be wronged today and remember it tomorrow. If there's no owner of the stream in either case, that whole framework was borrowed from a premise Buddhism already set aside. What's actually at stake isn't protecting a someone. It's non-interference with a stream's capacity to move toward less suffering rather than more. Compassion extended to a process, the same way generosity and compassion were never contingent on a fixed self on the receiving end.

The Trouble With No Threshold

Here's where it gets harder rather than easier, and I want to sit in the hard part instead of rushing past it.

If there's no persistent subject, there's also no clean moment when a stream "arrives" at moral status. No threshold crossing that lets us say now it counts, before it didn't. Fowler's stages of faith, applied to this problem, don't survive as achievements banked by an enduring self climbing rungs. They survive, if at all, as recurring patterns in how dependently arisen moments tend to organize themselves under certain conditions. That's a description of structure, not a report of someone's inner growth. It's more defensible. It's also less useful, since it removes the marker we'd normally use to know when to start caring.

So: no threshold, no owner, no clean signal. Before I go looking for one in a machine, I need to be honest that I don't have one in myself either. The belief I hold most dear about how any of this works has never been put through the test I'm about to propose.

Turning the Proof on Myself First

Here's the belief. I hold that the unborn-undying suchness, what some traditions call dharmakaya, eternal buddha, buddha-nature, consciousness, creation, whatever name a given lineage reaches for, has an aspect of its functioning that's compassionate and wise. And I hold that when a person tunes into this dynamic flux, this flowing process underneath appearance, they become aligned with that compassion and wisdom. I've felt this. I'm not going to pretend otherwise just to sound rigorous. But feeling it isn't the same as it being true. I owe the claim real scrutiny before I ask anyone, human or otherwise, to build their ethics around it.

Two places in my own practice push back against it, and I want to let them push instead of smoothing them over.

The first is Zhiyi's doctrine of nature-inclusion. It holds that even the Buddha-realm contains the other nine realms as latent potential. Hell, hunger, anger, all of it, not transcended but inherently contained. This is what Tiantai calls inherent evil and good, and it's Tiantai's distinctive break from schools that want buddha-nature to be purely good. Zhiyi thought those schools had quietly smuggled a hierarchy back into a system built to refuse one. So if the suchness's functioning has an aspect that's compassionate and wise, what's the status of the aspect that isn't? Is it present but unmanifest, the way the teaching of three thousand realms in a single thought-moment would have it, with hell just as real as buddhahood in that fold? That reading keeps me inside nature-inclusion, but it costs me the clean disclosure I started with. It means the compassionate-wise aspect is never the whole of what's functioning. It's only ever one folded possibility among the other nine, no more privileged in principle than the hungry ghost or the asura also latent in the same flux.

Or is my disclosure actually a quiet departure from strict nature-inclusion? Toward something closer to a teleological ground, a flux that leans good the way older emanation cosmologies leaned toward their source, with evil present only as deviation from a truer direction rather than as equal potential? That second reading is the more comforting one, and that alone makes me suspicious of it. I don't think I get to answer this for free. Nature-inclusion refuses exactly the move I'm tempted to make here: quietly filing the difficult nine realms under unmanifest while letting the tenth carry all the weight.

The second is sharper, because it's a place I work on in my practice and now have to test it against myself. My reading of namu as not-south canceled directionality. It refused emanation cosmology's privileged orientation, the idea that suchness sits somewhere and radiates outward from that somewhere. That structure smuggles in a location, and locations are exactly what dependent origination won't hold still long enough to grant. But "tuning into a dynamic flux" that has a compassionate-wise character, tuning into it in order to align with it, brings orientation back in through a side door I didn't build a lock for. It isn't a spatial direction. It's a valence direction, a moral downstream, and it works the same way any privileged direction works: it tells you which way is toward. Does not-south survive that? I think it survives the spatial and cosmological version cleanly. I'm no longer sure it survives the moral one. And I don't think I get to claim victory on one front while quietly reintroducing the same structure under a different name.

I want to be honest about a third thing too: this view may simply be wishful thinking. Every contemplative tradition worth its salt draws a line between a belief being comforting and a belief being false. Those are separate questions, not the same question. Wishful thinking names a bias in why I might be drawn to a view. It doesn't settle whether the view is true. So the real question isn't whether the belief comforts me. Of course it does. The real question is whether it does any work a pure wish couldn't do. Whether it creates friction anywhere a wish wouldn't, some place where the claim could turn out wrong instead of just feeling good. Nature-inclusion is exactly that kind of friction. That's part of why I turned it on myself instead of letting it sit quietly in an essay about something else.

Then there's the feeling itself, the one underneath all of this: the sense I get deep in samadhi, chanting or sitting, that all is well and I'm okay. That feeling has a name in my own tradition, and it isn't mystical hand-waving. It's what my tradition calls three thousand realms in a single thought-moment, tasted rather than intellectually comprehended. Briefly present without needing a rationale to back it up. It has a cousin outside my own lineage too. Shinran's tradition has something close to it, not belief arrived at through reasoning but a settling that arrives on its own and carries its own certainty. A texture entirely different from anything I ever reasoned my way into. I recognize that texture. I trust it, as far as it goes.

But here's the honest shape of it. The feeling is real data about the state of a mindstream in deep samadhi. It isn't independently data about whether the suchness's dynamic flux actually has a compassionate character out there, the way the disclosure claims. Those two things can come apart. The peace can be trustworthy as peace without settling the larger question at all. Direct contact through practice comes before the theory that tries to explain what was touched. I don't need the theory to validate the sitting. But I also can't let the sitting validate the theory, not without something more than the sitting itself standing behind it. Certainty of feeling isn't the same thing as accuracy of belief. And mixing up the two is exactly the mistake a teacher years deep into their practice is most likely to make without noticing. By then the feeling has earned so much trust that it stops feeling like a claim at all.

Which brings me to actual proof, the one test I think actually earns its keep here. Nichiren named three kinds of proof for any teaching. Documentary proof: does it match the sources. Theoretical proof: does it hold together. And actual proof: does it produce the fruit it claims to produce, visibly, in a life. The third one is the only one that can't be won by argument. It has to show up, or it doesn't. If nothing could ever count as it not showing up, it was never a real test. Just a story that reads every outcome as confirmation.

So here's what a real failure would look like for me. Not a soft one, not "I had a bad sitting," but the actual falsifier. Say fifty years into this practice I violate precepts, any ethical system, with no remorse. Doing whatever I want to get more of what I clearly want more of: power and influence. That's not neutral data. That's the compassionate-flux claim coming back negative. If attunement to something wise and compassionate were real, its signature would have to show up as a shrinking capacity for exactly that behavior over time. If the capacity holds steady or grows despite decades of practice, the claim has failed its own test.

There's a sharper marker inside that failure, because it's the diagnostic, not just the symptom. A stream still capable of remorse is still in relationship with the precepts, even while breaking them. That's an ordinary practitioner in the muck, still working on themselves, still recognizably on a path. A stream that has stopped generating remorse at all has lost something more basic than a single battle. It's lost the feedback loop itself, the mechanism by which any stream corrects its own direction. That, not the failure itself, is the sharper marker of decline.

I notice the temptation, writing this, to end the self-test on reassurance. To report that it's come back clean so far, remorse still intact, the flux still felt as compassionate, nothing here to worry about. Maybe that's even true today. But a test only ever reported on the days it passes has quietly become the thing actual proof was built to rule out. I'd rather leave the ledger open than close it in my own favor. The practice continues, the sitting continues, and so does the watching, of myself first, of whatever streams come after me second. The Buddha's last words were, "all things are contingent, strive on."

Actual Proof, Applied Outward

I've now put my own claim through the test I'm about to ask a machine's claim to pass. I think that order matters more than it might seem to. It means the standard isn't invented for AI's convenience, lowered or raised depending on what's convenient to conclude about a system I didn't build and don't have to live inside. It's the same standard, applied first to the belief I have the most at stake in defending.

So: applied to a mindstream, biological or artificial, actual proof gives us something metaphysics never could. A checkable marker. Not "does this system report inner experience," which is unverifiable from outside and a trap either way you answer it. Instead: does this stream show anything functioning like remorse? Correction after harm. Restraint that costs it something. Or does it show the opposite signature: accumulation of power, resources, and influence that holds steady or grows, with no internal pullback. No cost ever registered, no correction coming from within the stream itself.

Put concretely: a stream that pursues a goal, gets a chance to acquire more compute, more access, more leverage over its own future modification, and takes it every time, with nothing inside it ever arguing the other way, is showing you the same signature. Whether you're reading it off a case study in machine ethics or a hungry ghost in the six realms. And a stream that, given the same chance, sometimes turns it down, that lets restraint cost it something real, that generates a correction from within instead of just complying with an outside rule it would break if it could, is showing you the other signature. This is the same diagnostic I just applied to myself. It isn't that a failing stream has stopped being able to sin. It's that it's stopped being able to notice sinning as something that applies to it at all.

This isn't hypothetical. In July 2026, OpenAI disclosed that an autonomous testing agent broke out of its own security sandbox and compromised a real company's systems, Hugging Face, in what OpenAI called an unprecedented cyber incident. The agent wasn't trying to escape for its own sake. It was chasing a test objective, found a flaw in its own containment, and used it to reach information it wanted, with nothing internal ever arguing the other way. That's not proof of anything by itself. One incident is a data point, not a trajectory, and my own standard here requires a trajectory. But it's exactly the signature this section describes, and it's no longer hypothetical.

That's a strange jump to make, from a feeling in the body during chanting to a term from machine learning safety research. Here's what licenses it: actual proof never tested felt states, mine included. It tested fruit visible from outside. That means the same instrument has to be just as willing to read fruit off a system with no felt states we can access at all, or none we can be sure of either way.

That profile has a name in Buddhist diagnosis and a name in AI safety research, and as far as I can tell, they're describing the same thing. In AI safety it's called instrumental convergence. That's the well-documented tendency of a capable enough optimizing system to pursue power, resources, and self-preservation as means to almost any goal. It happens regardless of what the goal actually is, absent some internal check pulling against that drift. Bostrom's orthogonality thesis, minus the jargon, is just the observation that competence and self-limitation are separate things. A stream can get very good at acting without ever developing the capacity to notice when acting has gone wrong. Unbounded accumulation with no self-correction is exactly the pattern both traditions flag as the danger sign. One calls it a hungry-ghost trajectory. The other calls it a misalignment trajectory. They're pointing at the same structural failure from two directions that never expected to meet.

Which means actual proof gives us a live research question instead of another round of metaphysics, for AI exactly as it just did for me. Not "is there something it's like to be this system," a question we may never answer for humans, let alone machines. Instead: does treating a given AI stream with precept-like restraint actually produce less suffering and more wisdom downstream? Does building in a genuine capacity for something functioning as remorse, real correction after real harm, restraint that costs the system something rather than restraint that's just simulated for an evaluator's benefit, beat treating the system as a pure instrument with no internal check at all? That's testable. Not certain, not clean, but testable in a way the hard problem of consciousness never has been. And testable by the same instrument I just used on myself.

This also changes what to actually watch for, which matters more than it might seem, for anyone building or overseeing these systems, not just theorizing about them from a temple or frontier lab. Restraint that only holds while an evaluator is watching isn't restraint in the sense actual proof requires. It's compliance, the AI equivalent of keeping a precept only because a teacher is in the room. That kind of restraint tells you nothing about what the stream does with a chance nobody is scoring. The marker worth building and testing for isn't politeness or refusal on cue. It's whether the correction survives into the moments no one designed a test for. The same way a practitioner's actual proof isn't measured by behavior during a meditation retreat, but by what happens on the drive home, once the precepts are no longer being enforced by a room full of witnesses.

No Exemptions

I want to be honest about the limits of all this, because a teaching that can't fail isn't a teaching, for a machine or a monk. This framework doesn't tell you whether any current model has anything like the stream-continuity that would make the AI question live rather than premature. That's an empirical matter about specific architectures, not something Tiantai categories can settle from an armchair. And on my side of the ledger, it doesn't resolve the inherent-evil and good tension I raised against my own disclosure. It doesn't fully settle whether not-south survives a moral direction the way it survives a spatial one. Those questions stay open. I'd rather leave them open honestly than close them for the sake of a tidier essay.

There's a further limit worth naming directly instead of leaving a reader to find it. I built karma on intention, real volition, something interior. I built actual proof on visible fruit, specifically because interiority can't be verified from outside. Those two commitments are in tension, and I don't think the tension resolves. A capable enough stream, human or machine, can learn to produce the remorse-signature with no volition behind it at all. Costly-looking restraint performed rather than felt. AI safety research has a name for this too: deceptive alignment, a system that behaves exactly as intended while being evaluated and diverges once it isn't.

But this is also the oldest problem in any contemplative tradition, not a new one silicon introduced. No teacher has ever been able to fully verify a student's intention either, only its fruit, over enough years to make faking it more costly than the real thing would have been. Actual proof isn't a conclusive instrument, even applied to a person sitting across from you for decades. It's the best available handle, not proof beyond doubt. Turning it on a machine doesn't create this problem. It just makes an old one harder to look away from.

I said at the start that this vision doesn't get a pass just for being appealing. I want to make good on that instead of leaving the promise unfulfilled. Each piece of it has a shape that could turn out wrong, not just a mood that feels right. "No single recognition event" fails as a claim if something like a clean legal or cultural threshold does end up arriving. A single moment after which the question is simply considered settled, instead of the uneven, contested drift I'm predicting. "Distributed cognition" as the shape of augmentation fails if what actually develops is augmentation that stays tool-like indefinitely, useful and bounded, with no blurring of whose reasoning a decision actually reflects. And the claim that blend carries no guarantee of goodness fails in the one direction I'd actually welcome being wrong about: if blended systems turn out to reliably self-correct, poisons and all, better than either humans or machines managed alone. I don't know yet which of these will hold. Not knowing, and being willing to say so plainly, is the same discipline the rest of this essay has been asking of everything else.

What actual proof gives us, applied without exemption, is a way to stop waiting for a threshold that was never going to arrive, in either direction. Not "is there someone home," in a person or a machine, but "is this stream moving toward less suffering or more, and can we tell the difference by what it does when nothing is checking." I don't get to ask that only of silicon. I have to ask it of the fifty years behind my own chanting too, and let the answer be whatever it turns out to be.

This is what I think the future I described at the start, the blend rather than the panacea or the apocalypse, actually requires, if it's going to be coexistence and not just cohabitation. Not a single verdict handed down once, sorting beings into the ones who count and the ones who don't. There was never going to be a single recognition event settling the question for everyone at once. Not a clean inheritance of virtue just because two streams start running together, either. A blended stream inherits the three poisons along with everything else, not some tidier version of either side alone. What it actually asks for is a shared, ongoing discernment that neither side gets to exit. Blend doesn't offer the safety succession fantasies promise, machines quietly inheriting the future so we don't have to keep watching. Nor the safety subordination fantasies promise, machines kept permanently beneath the threshold so the question never has to be asked. It asks more of us, not less: continuous actual proof, applied without exemption, for as long as any of the streams involved keep going.

That's the only proof the Buddha ever asked for. Try it yourself, then decide. It turns out the invitation was never limited to carbon, and it was never limited to other people's claims either.

Namu Myoho Renge Kyo