The Teacher, Part 5: what a teacher is for

We end where we should have started, with a single sentence Geoffrey Hinton said almost in passing, at a developer conference in 2019, shortly after he had received the Turing Award. He was talking about what it might mean if we ever truly understood how the brain learns. “You’d have thought,” he said, “that if we really understand what’s going on we should be able to make things like education better, and I think we will.”

We have built a whole mission on roughly that sentence. So for the last part of this tribute we want to take it seriously — and say plainly what we think a teacher is actually for, and why his story is the clearest argument for it we have found.

The conviction underneath it

Hinton’s belief about learning was never only about machines. At the same 2019 talk he said of neural networks as a model of the brain, “It seems to me that there is no other way the brain could work” — that learning, in us as in the machines he built, is connections changing strength with experience, not rules handed down from on high. He spent fifty years following that conviction past the point where it was respectable, and it is worth pausing on the courage that took. The lesson he has repeated to young researchers, in his own paraphrase across many interviews, is to trust a good intuition even when the whole field is telling you it is nonsense — because the idea everyone dismisses is sometimes the one that changes the world. His own career is the proof: the unfashionable idea won.

That is a teacher’s deepest gift, and it is not information. It is permission. Permission to think the thing that sounds silly, to follow it, to become more than the room expected. Everything in the four parts before this one is really one story told four ways: a man who kept faith with an idea, taught it to others, watched them surpass him, and then told the truth about where it all leads. The science, the students, the warning — they are all teaching.

What we think a teacher is for

A teacher is a multiplier, as we said in Part 3 — but multiplying what? Not facts; facts are cheap now and getting cheaper. A teacher is for getting the best out of the particular being in front of them. The thing only this person can do, the intuition only this person has, the direction only this person wants to go — and then handing them what they need to go there. That is the part no system invented, the part worth protecting, and the part most of human history has rationed cruelly: real teaching has almost always been reserved for whoever could afford proximity to it.

Meta DAX is our attempt to stop rationing it. Learning handed to every human being — any subject, any domain, free, forever, built on paid compute rather than on anyone’s inability to pay. The oldest ratio in education, too few teachers for too many learners, deliberately inverted, so that the multiplication that once happened inside one Toronto lab can happen for anyone with something to learn or something to teach. The cognition behind the tools is provided by Obsidian Delta. The purpose of pointing it at learning is the sentence Hinton said at a conference in 2019 and the one we would add to it: not only to make education better, but to get the best out of every single being it reaches — and to let that best pass on, the way his passed to his students, and theirs to the world.

The ask, which is not a metaphor

This series opened the second face of a grand opening. The first face — a tribute running alongside this one — is about getting people off their screens and moving. This face is about learning, and the ask is the same in both: come team up with us.

We are not a gatekeeper. We are looking for partners — teachers, tutors, parents, schools, clinics, towns, anyone, anywhere, who already spends their days getting the best out of people. We will give you the tools. We will show you how to use them, safely and honestly, which after Part 4 you know we do not treat as optional. And then they are yours. This was built to be handed over, because learning held behind a paywall is not, to our minds, learning that has understood itself.

Geoffrey Hinton taught machines to learn, taught people who changed the world, and then warned that world about the very thing he had made. We are not affiliated with him, and nothing here is his endorsement — read the line below, as you have read it on every part. But if you want to know what we are trying to do, you could do worse than his own sentence: if we really understand what’s going on, we should be able to make things like education better — and I think we will.

So do we. That is the whole of it.

This is Part 5 of five, and the end of the series. Thank you for reading it.

---

This is a tribute and our own commentary. The DAX Foundation and Obsidian Delta are not affiliated with, endorsed by, or in partnership with Geoffrey Hinton, the University of Toronto, Google, or any person or organisation named across this series. Every quote here is attributed to its public source as reported; nothing is put in anyone’s mouth.