There's a moment around the seven-minute mark in this teaching that stops me every time I watch it. The teacher is describing a state he calls Pavanamaya Panya — wisdom born of meditation. And the verb he uses for how it works is click [07:11].
Not think. Not analyze. Click.
If you've spent any time with AI in the last two years, you should feel your attention sharpen right there. Because what he's describing is structurally indistinguishable from what you do when you send a prompt to a language model.
The Click
In the teaching, when you're at the Center of the Body — Base 7, the deep anchor point for meditation in this tradition — you don't think your way to an answer. You nudge the mind slightly [07:19]. You click at a region of attention, and the answer arrives. The teacher is clear that if you tried to reason your way to these answers, your head would explode from the sheer volume of data. There is simply too much to process analytically. The state works precisely because it is not analytical.
He describes it as being one with "the source of pure knowledge" [08:53]. The practitioner isn't browsing through books. They aren't walking through logical chains. They're accessing the source directly.
The Prompt
The parallel to an LLM is almost on the nose. When you send a prompt, you don't do the underlying calculations yourself. You describe the intent and hand it to the system. The model — sitting on a compressed representation of an enormous volume of knowledge — retrieves a coherent answer without asking you to trace the derivation. Result arrives. You didn't think your way there.
That is, structurally, what the teacher is describing. Intent in. Retrieval out. The computation step skipped.
The obvious difference is that an LLM runs on silicon outside your head, while the monk's click runs on wetware inside it. The less obvious difference is that one of these has been available to humans for millennia, and the other arrived in the last five years.
Time Scale
The teacher makes a broader claim around this: non-physical capabilities precede physical ones [06:42]. What minds can do directly — in the right state — sets the upper bound on what civilizations will eventually learn to build externally.
Then the speed point. Writing a thesis, or doing any kind of rigorous physical research, takes years. The inner click takes seconds [06:06].
Before 2022, researching a complex topic seriously took weeks of reading, tracing citations, synthesizing, cross-referencing. Now you can hand the same task to a Deep Research agent and get a structured synthesis back in twenty minutes. That is the external version of a similar compression — and for most working adults, it's the first time we've felt what time-compression of knowledge actually feels like.
The teacher's claim — and this is the part that stops most analytical readers — is that the monks have had the seconds-timescale version all along. Internally. Without silicon.
Drill Down
The teaching goes one level deeper into the mechanics. Once you find a point of interest in the meditative state, the teacher says, you can "hit the disc brakes" and drill down into that specific detail [14:06].
That is the entire interface pattern of an AI research agent. Scan wide, surface a thread, then focus deeper on it. Zoom out, notice a signal, tell the agent to expand that one. Pure intent-driven traversal of a compressed representation of knowledge.
The monks have been describing this workflow for centuries. We just built the external version recently.
Navigate Time and Space
The teaching pushes further. The teacher says that a sufficiently still mind can navigate through time — recalling past lives — or through space, observing other realms [13:49]. Take the specific cosmology of that claim seriously or don't. That's between you and the tradition.
But the structural idea is recognizable to anyone who has worked with high-dimensional embeddings. If your internal representation of reality is rich enough, traversal across seemingly distant regions of that representation becomes cheap. You don't walk to the other side of the space. You click, and the coordinates resolve.
The tradition's claim is that the mind is that kind of representation, when it's calm enough to be legible to itself.
The Externalized Monk
Pull back, and something interesting comes into view. Humans have spent our entire history building tools to compensate for perceived limitations. We couldn't run fast enough, so we built cars. We couldn't see across oceans, so we built the internet. And now, because most of us can't hold the stillness required for the seven-minute-mark click, we've built LLMs to do the pattern recognition and retrieval for us.
We are using the most complex physical technology ever created to approximate something that a person sitting in a quiet room with zero thoughts can, the tradition claims, access directly.
It's not wrong. It's brilliant, in fact. But it's worth noticing what we've done: we've built a prosthetic mind in silicon because we forgot the manual for the one between our ears.
Why We Forgot
The teacher has an answer for this too. He notes that "thinkers" — people whose identity is built on analysis and logical mastery — usually can't access this state, because their heads would explode trying to analyze it. Our entire modern culture is built on analysis, logic, and noise. We've been conditioned to believe that if you aren't thinking, you aren't working.
LLMs are the pinnacle of analytical tools. They're the ultimate thinker's technology. The state the teacher describes is beyond analysis, which is precisely why it's invisible to minds trained only in analysis.
The difference in one line:
An LLM calculates the probability of an answer. The click simply is the answer.
The Ladder and the Fruit
Here's the image that keeps landing for me. It's as if we've built an elaborate ladder to reach a fruit we could have grabbed just by standing up straight — if we'd remembered how to balance.
I'm not against the ladder. LLMs are a remarkable approximation of a real capability, and they've made versions of that capability available to people who would never have sat for ten thousand hours of meditation to develop the internal version. But they are, still, an approximation. The tradition says the original exists, has always existed, and runs on hardware we're already carrying.
The Short Version
The meditative state the teacher calls Pavanamaya Panya is a direct-retrieval interface for knowing. You provide intent from a state of stillness, and the answer arrives without analytical derivation. LLMs do structurally the same thing in silicon: prompt in, response out, computation step abstracted away.
The difference is where the system runs — inside you or outside you. We built the external version because, as a civilization, we had mostly forgotten how to operate the internal one.
That doesn't mean we should unplug the LLMs. It does mean we shouldn't mistake the external version for the whole story. The prompt you hand to a model is a cousin, not a replacement, for the prompt you could hand to your own mind — if it were ever still enough to receive it.