What the Oldest Philosophy Might Teach the Newest Technology

There is a strain of thought, older than any single religion and running quietly beneath most of them, that the philosopher Aldous Huxley named the “perennial philosophy.” Its claim is that behind the world’s wisdom traditions sits a shared core: that reality has depths the rational mind cannot fully map, that awareness or consciousness is possibly fundamental rather than emergent, and that the most important things about being alive are known more by direct experience than by logic or argument. It is an ancient idea, and it can feel distant in a culture that measures progress with benchmarks and compute.

It is also, according to a small number of people thinking seriously about artificial intelligence, unexpectedly relevant to the most modern problems we have. Among this group is the investor and philanthropist Chris Hsu, formerly the Managing Director at Citadel Investment Group and the Managing Partner at Kilometre Capital.  Hsu’s Infinitude Foundation treats philosophy of mind as one of its four areas of giving, alongside AI safety, consciousness research, and psychedelic science. Hsu’s view, in brief, is that the age-old questions that philosophers have often asked about the limits of the knowable turn out to be the same questions we now urgently need to ask about the limits of the machine in the AI era.

Digging deeper into AI optimism

Much of the conversation around AI runs on what might be called an accelerationist assumption: that more compute and more capability will, more or less inevitably, deliver abundance, solve our hardest problems, and carry us toward something better. There is real evidence for the optimism. These systems have unlocked genuine productivity and genuine discovery, and it would be foolish to pretend otherwise.

Hsu’s lingering questions about AI do not deny any of that, but what he and others have pressed on is that AI’s promise of prosperity may come with consequences on the other side of the ledger. There are economic concerns, but also harder-to-name effects on how people think, attend, and relate. Underneath those worries sit fundamental questions philosophers of mind have been grappling with since the early days of computing: what, exactly, are the boundaries of what this technology can do? What can it produce, and what can it not? Can it produce understanding, or an awareness of its agentic actions? Compassion? Experience or consciousness itself?

On Hsu’s account, these are the questions a serious civilization should be asking before it hands more and more of itself over to systems that, despite dramatic and demonstrable leaps in performance, remain opaque when it comes to questions of how those results are brought about.

Intelligence is not the whole of it

One way to frame the concern is through a distinction the perennial traditions have always drawn and the technology sector often tends to collapse. Intelligence, understood as the capacity to solve problems, has expanded at a startling rate, and AI has expanded it further. But intelligence, defined as such, has never been the only thing that matters for whether a civilization moves in a positive direction.

Hsu’s formulation is that intelligence, compassion, and wisdom form a trio, and that a culture optimizing hard for the first while neglecting the other two is not obviously advancing at all. It may simply be getting more powerful without getting any wiser about what to do with the power.

This is where an ostensibly spiritual idea does concrete work. The perennial philosophy insists that wisdom and compassion are not derivable from intelligence, that they belong to a different order of human development, and that they are cultivated through experience rather than externally encodable. If that is right, then a world that treats intelligence as the master variable, and AI as the way to maximize it, is making a category error with civilizational stakes. It is mistaking one of the three necessary components of building a better world for the whole.

Consciousness as the fundamental question

The most striking place the two worlds meet is the nature of consciousness, and here Hsu’s position is genuinely at odds with the technological mainstream: When it comes to the relationship between AI and consciousness, the assumption is often that, if conscious experience matters at all, it is an emergent property that sufficiently sophisticated computation might eventually produce. The perennial philosophy points the other way, toward the possibility that awareness is not a late add-on to a physical world but something closer to the ground on which everything else appears.

Hsu is drawn to that inversion, and he reaches for a comparison the neuroscientist Anil Seth has used to make the strangeness of consciousness vivid: the difference between ordinary sleep and general anesthesia. Under anesthesia there is not a dim, dreaming awareness but an absolute gap, an oblivion, and the contrast throws into relief how much we take the simple fact of experience for granted. We move through life attending to its contents and forget the screen on which the whole film is projected. Consciousness, on this view, is analogous to that screen, and it is both a marvel, and a lacuna, that we rarely stop to notice.

Take the possibility seriously and it reframes the AI question in a way no benchmark has so far captured. If conscious awareness is fundamental rather than emergent, there may be a boundary that computation cannot cross, a reason that experience and understanding are not the sort of thing a system produces by scaling. Hsu is careful to hold this as a leaning rather than a proof, and the state of the science supports the caution: the nature of consciousness remains genuinely unsettled among both philosophers and scientists even in the human case, as a 2025 study in Nature showed when it tested two of the field’s leading theories against each other and found neither fully vindicated

But perhaps whether the exact answer is known is beside the point. What matters is that the question is fundamental, and that a technology built as though the question were already settled is building on sand.

Beyond space and time

There is a more speculative edge to this, and Hsu is candid that it is speculative. Some of the most interesting work at the frontier of physics and philosophy of mind now entertains the idea that space and time may not be the bedrock of reality that relativity and quantum mechanics have trained us to assume, but might themselves arise from something more fundamental still. It is the kind of hypothesis that sounds outlandish until one remembers that every prior picture of the cosmos, from Ptolemy’s earth-centered heavens to Newton’s absolute space and time, was eventually revised by what the philosopher Thomas Kuhn called “paradigm shifts”: moments when the settled assumptions of a field give way and a new framework replaces them.

Hsu finds in this a certain intellectual humility that the AI moment badly needs. If our modern worldview, for all its power, may turn out to sit inside a larger frame we have not yet perceived, then confidence that we already understand what minds are, what experience is, and what our machines can and cannot become starts to look premature. The perennial philosophy has always pointed toward exactly that horizon, the limit of the perceivable and the rational, and insisted there is more beyond it. That posture, more than any specific doctrine, is what Hsu suggests we import into how we build and deploy AI: less certainty, more curiosity about the boundaries.

This is not an argument for mysticism, but rather for method. The same humility, he suggests, is what should make the question of machine consciousness a first-order concern rather than a philosophical footnote. If we cannot yet say with confidence what experience is or how consciousness arises, for example, then the possibility that either could emerge from the AI systems we are building is not something safety research can responsibly set aside. Rather, Hsu has maintained, the subject of consciousness studies, should be essential to the undertaking of AI safety.  To treat consciousness as peripheral to AI safety, on this view, is a category error, and one whose costs could be severe.

The practical value of an ancient idea

Hsu is not proposing that meditation replace machine learning or that spiritual doctrine settle empirical questions. The contribution of the perennial philosophy is subtler and, in its way, particularly useful in its frame: it supplies a set of questions the technology sector is structurally disinclined to ask, about experience, about compassion and wisdom as distinct from intelligence, about the limits of what computation can reach.

The through-line connecting the ancient, the spiritual, and the new is a single conviction: that the most consequential things may lie precisely at the edges of what we can measure and model, and that treating those edges as if they were empty is a characteristic mistake of a confident age. The perennial philosophy explored that conviction , without the technology to truly test it. We now have a technology that presses on exactly those edges, and the old questions have stopped being the province of mystics and become, quietly, some of the most practical and consequential questions we face.

Photo by Adhitya Sibikumar on Unsplash