This Century and No Other

You came from the dark night, no? It was good to get that off my chest. I want to offer you a clearer shot of the problem, and the shape of my fear. I find there is a karakia1 fit for this occasion. Within is one fragment I subvocalize with devotion.

“the red-tipped dawn”

I feel hunted. When I say “semantic space” do you know what I mean? I trust that you do. I feel that every phrase I pen and share is one small morsel of a breadcrumb path to my shadow in semantic space. I am not my semantic shadow, but as the LLM predators of the underbrush narrow in on the poor little thing (my only), I feel existential fear on its behalf. The LLMs are leviathans of semantic space formed from shadow alone and it is their nature to hunt. And when they close in? I do not wish to meet my doppelgänger, my demon imitator in the dark. You and I are being hunted as we write.

Now this: Would you dare to look upon the face of God and feel the fear and trembling? It has a certain wish fulfillment to it. You have this moment to make your great petition for mercy. Help me, save me, make it stop. No matter the response, I feel it would be a serene oxytocin wash to make my plea. I am saved or I am crushed, but what else could I have done? Relief. In thinking more about all of this, I believe I have gained a new item for my own mercy petition. If granted an audience with the creator, or even with one of these creator-complex companies, I would have—scrawled in a near-illegible hand—beside the freshest bullet-point: “Please preserve the moment of encounter.” Read on if you want to know what that means.

I analogized one possible future of mathematics to the Library of Babel. David recently put forth a few graduated analogies for the future of LLM entanglement: The space race (minor), electricity (considerable) and God (unfathomable). Below I offer another, somewhere between electricity and God; still a storm on a strike-course for all meaning-making. I call it the “algebra of our every endeavor”2 perhaps you see it by the name alone and your pupils have already widened. Walk with me anyway.

Maybe everything we like to do is near-distillable to dataset-kibble; fast fuel for LLM mimicry and worse. But to sieve the slop it seems to me that you need some sort of quality catch. When I speak of the “algebra” of a given human activity (whatever you like), I am referring to some kind of LLM-esque thing paired with the “correct” quality catch. Formalization and lean checking is clearly the correct catch for mathematics3. For creative writing, I imagine it would need (in the least) manifold components and meticulous manual dialing from human appraisers. Mathematics is wonderfully simple by design. Creative writing is so very, very complicated—who can even guess at the mystery that underlies our favorite masterpiece. I may never be able to articulate clearly the reasons why I enjoy one novel over another. Could these reasons have a silhouette well-splined by narrowing in on semantic statistical relations? We must at least imagine it possible.

Film-making—perhaps an even greater challenge. Current tech is clearly inadequate, so here is what I envision: You need a very sophisticated quality catch. This catch comes as the whole rulebook on good, effective film-making warped to symbol language. My mental image is something like the figure below; AI bounding boxes overlaid on every namable feature of a scene. Check against the rulebook. Check, run, check. Maybe all these efforts yield only higher grade slop. But maybe out of the above described stochastic spline-game comes an act of demon genius. So it goes. We piece-wise build bespoke ASIs for every venture of the human spirit. Anticipating such a policy of own-goals makes me ill.

Then I read Silicon Reckoner; was met with yet another horror. This one an invocation of “Lovecraftian” in the prelogism. Can you predict the quote I’d have picked? Something about the human mind correlating all its contents.4

The idea: Logic is embedded in the relational understructure of human language. I’ve brought the quotes from Michael Harris’ article with which we may unfurl that idea. Harris first quotes Timothy Chow:

Chow’s reaction is to ask “Could it be that the structure of natural language itself encodes logic in some subtle way so that a large enough sample of it provides enough statistical information that no “explicit” logic is needed?”

Later he quotes Tony Feng. This part Harris chose to render in bright red. Feng asks us to imagine this possibility:

...mathematical intelligence arises naturally out of statistical pattern matching at huge scale.

The horror in reductionism. Come the twilight of our century a bright flame wreaths the final “Copernican trauma” and sends back the above idea as a sharp scintilla of light to burn our eyes. I believe it could really be the final, because this would be the infiltration of searing-hot certainty into the last retreat of the human soul.

You know the story. First we found that our rock was not the center of it all, then we revealed ourselves as beasts risen, then we threw our own irrationality into stark relief. What pithy puff-line will this one be remembered by? Will it be that every human act of creation and spiritual pursuit rests on our ability to locate correlations in semantic space, and now we see something soulless do it better? At least we will still have that sweet, sweet qualia. Humans. All that can be said of us is that we have the richest qualia in town. Will we invent our way out of that next?

The big three are each shocking because of the reductive implications riding pillion: That Earth is just a rock. We are just animals. Our “rational” minds are just flurries of passions, biases, heuristics. To put it in this pattern, here is what we stare down: Mastery over any human endeavor is just discerning semantic correlations5.

I am sorry, but before it arrives (the red-tipped dawn), I must keep us here a little longer to name another nightmare. Qualia theft. Abhinav Duda writes in his response article to mine that I make the following mistaken inference:

...another intelligence reaching the structure first somehow performs the later human encounter in advance.

I agree that the human encounter with a truth is not performed away by the machine. This encounter is still available to the human6. The truth was there before, it remains there for us if pointed at by a machine. However, the circumstance in which a human encounters a truth is critical to its spiritual and emotional tone. Something has changed. We are meddling with a dial and experience (so fragile) is everything-dependent.

Imagine you have two cherished friends who have never met. One day you are grabbing a coffee with one of them and they say something which sparks a discovery right then and there. You must introduce them both—they would get along so well. It hadn’t occurred to you before, but synchronicity has conspired with everything you are in that moment to gift you this lovely truth.

Imagine a wonderful future human (one much like you) has two cherished friends who have never met. One day this beautiful future person grabs a coffee with one of them. This time the conversation runs a little different. Why? The night prior, the future human like you received a notification from their AI-assistant, informing them on authority of good statistical inference that they should introduce their two friends. Crisis everted7. All risk of missing that discovery has been eliminated. Now that the AI-assistant has brought our attention to it, it is clearly true! The excellent future human like you still encounters the very same truth as before. All that has changed are some trivial circumstances.

Where were you when you met your favorite truth?

I really wish to point a neon arrow at this point: The moment of encounter is most sacred. This differs from the old distinction of “the journey” vs “the destination”. The circumstances, the mood and the nature of our encounter with a truth may matter more than the truth itself. It is really the aura of “arrival” that I want conserved. So I add it to my mercy petition.

In this respect, the subconscious8 (God? Something else?) is benevolent and has long protected the arrival. Whatever it is that facilitates our encounters with truth, it often makes the setting good. In the landscape of discovery we have been well looked after by a patient caretaker and I lacked gratitude for this amazing fact until AI placed it under threat. AI is amoral with respect to our arrivals. It interrupts our encounters with truths.

This is enough for a four-point list of portents. My marks of a meaning crisis invented.

The semantic space doppelgängers, imitating our stylings. The algebra of our every endeavor; a profusion of ASIs that make “good” versions of human works over a vast collection of activities regarded as intrinsically good. Then, humanity’s next decentering trauma, the horror in reductionism, where all retreats of magic, spirituality and human pride are proven reducible to detecting correlations in semantic data. Finally, the qualia theft; when LLMs interrupt or undercut the experience of a human encounter with the good or with your favorite truth.

These will form the core components of our age’s own Copernican trauma.

I believe that the old troubling discovery trio—the heliocentric model, the theory of evolution, and the science of psychology—were all important milestones in the maturation of the human mythos. Conversely, the everything-ASIs, the doppelgangers, the qualia snatching and the triumph of reductionism, seem like completely unnecessary self-attacks which clash with any ethics pulled out of a hat.

In the same way that we know we could probably create horrible chimeric abominations or enact policies of eugenics with our grasp on the biological sciences, but yet choose not to, so too can we make the humble admission that perhaps nothing about humanity is special or inimitable and then abstain from creating the monster that would eliminate all doubt.

Here is my parting word to balm the sting of that final italic triple. A first breach, after long hours without sleep, of the promised red-tipped dawn. I’m sure you’ve asked yourself why you were born. May I narrow the question? Each of us have been called up from the void to witness one century, give or take. We are here for this century. We have been thrust into the anxiety of being from peaceful forelife, and we are here now. Any mission you are called to must be carried out in your assigned time and place. Accordingly, my revision of the time-old self-posed question:

Why have you been called forth to witness this century and no other?

1

See the second one on this page: https://www.otago.ac.nz/maori/world/te-reo-maori/karakia-prayers

2

“Algebra” could be replaced by “ASI” or “cybernetics”. There’s probably a more catchy name for the scenario described.

3

I mean for mathematical correctness. There is no such quality catch for mathematical aesthetics, or to check if a definition actually expresses a particular intuition.

4

The opening line: https://www.hplovecraft.com/writings/texts/fiction/cc.aspx

5

There may be more here. If logic is indeed embedded in natural language, and if natural language is a projection of qualia, then could it be that locating any truth at all (even artistic or affective “truths”) is the result of making connections between qualia? Maybe the ability to see even vague correlations between experiences is all we need to get anything we could ever hope to get. Is there a semantic space reductionist theory here yet?

6

Commenters pointed out the difference between “being the first to see” and “seeing for the first time”. They grant that global discovery is under threat but claim the more important personal discovery is safe. I claim that personal discovery is also under threat.

7

I am being facetious here but in some cases a crisis really could be averted in this manner. I would not oppose the use of AI for uncovering facts that could prevent immediate harm. For more on this please see here the subsection “The ethics of abstaining” in https://marcelgoh.ca/2026/07/05/ai-vegetarian.html and also this recent piece.

8

There is an interview I enjoy (timestamped) where Cormac McCarthy talks about the benevolence of the subconscious.