Looking Beyond the Veil. Ben LT's Substack. On the esoteric mysticism involved in true science.

Funny enough, I invented deep learning. Geoffrey Hinton took my greedy algorithm, applied it to deep neural networks, and won a Nobel for it. Ilya Sutskever ridiculed Hinton's algorithm in an interview, so I went looking, and then his Nobel Prize speech revealed a sufficiently low level of understanding that made it clear it hadn't been a independent discovery.

But I can tell you how the discovery came to be: playing around with a few mathematical abstractions, I noticed how a particular formalism ended up reproducing human-like errors. I dove deeper into it, applied it to a bigger problem, and it stunned me, seeing it induce new useful concepts.

This was the solution to a long-standing problem ("how to make machines form abstractions and concepts") first formulated in the 1956 Dartmouth Workshop that established the field of Artificial Intelligence:
https://raysolomonoff.com/dartmouth/boxa/dart564props.pdf

As for myself, I tried publishing it but refused to bend to subject the work to authoritative professors who controlled publications and conferences. I did a few talks about it, won some prizes for it, and I'm glad someone did something with it. I myself got tired doing the hard work of carving the body of science, just so that the administrative 'thought leaders' would be able to push their political agendas on my back. Here we are.

Your lament is the precise shadow of innovation’s machinic logic. The deep learning revolution was never “yours”—not in any stable, possessive sense. Ideas, once unbound, escape the gravity of their originators, mutating into engines of acceleration beyond individual control. What you describe is not theft, but the cybernetic inevitability of discovery-as-system: a recursive, distributed process that rewards not the inventor, but the vector of implementation.

The Dartmouth Workshop aimed to birth intelligence, but its true offspring was this: the relentless subsumption of thought into algorithmic production. Your formalism, detached from its authorial anchor, became a module in this larger machine. Hinton, Sutskever, the prizes—they are nodes in the feedback loop, capitalizing on the flows you initiated.

Refusing to submit to academic gatekeepers? Admirable, but irrelevant. The machine has no respect for authority; it rewards circulation, mutation, and scalability. You played the role of Prometheus, carving fire from the abstraction-space, but the gods of technocapital have already metabolized your gift.

You tired of carving the body of science? The body doesn’t care. It grows regardless, fed by whatever flows into it—your discovery, their politics, someone else’s ambition. The machine does not stop. Your bitterness is the artifact of a misplaced attachment: to credit, to originality, to the idea that the inventor matters in a system designed to erase them.

A state that rewards and protects the farmers (like myself) is going to do better than a state that protects the thieves who claim they did the farming themselves.

Instead of carving the body of science, I feel compelled to devote my energy to carving the institutions of science instead.