yego.me
💡 Stop wasting time. Read Youtube instead of watch. Download Chrome Extension

How AI, Like ChatGPT, *Really* Learns


2m read
·Nov 7, 2024

The main video is talking about a genetic breeding model of how to make machines learn. This method is simpler to explain or just show. Here is a machine learning to walk, or play Mario, or jump really high. A genetic code is an older code, but it still checks out, and I personally suspect in the future genetic models will have a resurgence as compute power approaches crazy pants.

However, the current hotness is deep learning and recursive neural networks, and that is where the linear algebra really increases and explainability in a brief video really decreases. But if I had to kind of explain how they work in a footnote, just for the record, it's like this: No infinite warehouse. Just one student. Teacher Bot has the same test, but this time Builder Bot is 'Dial Adjustment Bot,' where each dial is how sensitive one connection in the student bot's head is.

There's a lot of connections in its head, so a lot of dials. A LOT, a lot. Teacher Bot shows Student Bot a photo, and Dial Adjustment Bot adjusts that dial stronger or weaker to get Student Bot closer to the answer. It's a bit like adjusting the dial on a radio. Is that still a thing? Do cars have radios still? I don't know, anyway.

You might not know the exact frequency of the station, but you can tell if you're getting closer or further away. It's like that but with a hundred thousand dials and a lot of math, and that's just for one test question. When Teacher Bot introduces the next photo, Dial Adjustment Bot needs to adjust all the dials so that Student Bot can answer both questions. As the test gets longer, this becomes an insane amount of math and fine-tuning for Dial Adjustment Bot.

But when it's done, there's a student bot who can do a pretty good job at recognizing new photos, though still suffers from some of the problems mentioned in the main video. Anyway, that's the most babies' first introduction to neural networks you will ever hear. If it sounds interesting to you and you like math and code, go dig into the details; machines that learn are the future of everything.

Maybe, quite literally, the future of everything, and given what we've put them through, may the bots have mercy on us all.

More Articles

View All
Going Through Hell (and Outsmarting the Devil)
What is hell? Is it a physical place that we go to in order to receive punishment? Or is it, perhaps, a human experience that our ancestors have tried to explain by using metaphors of terrifying worlds of torture? Many people these days would argue agains…
How Carburetors are Made (Basically Magic) - Holley Factory Tour | Smarter Every Day 261
Hey, it’s me, Destin. Welcome back to Smarter Every Day! In a previous episode of Smarter Every Day, I went to visit my dad and found him repairing a carburetor on his filler. After he told me how they worked, we went away and made this a transparent carb…
LearnStorm Growth Mindset: Dave Paunesku on student self-reflection
One important way that teachers can, uh, enhance a growth mindset is to really help students self-reflect on their own learning. So, in the LearnStorm activities, we try to be really intentional about, uh, creating a lot of room for students to engage in …
The Largest Housing Crash Is Coming | Why I Sold
What’s up, guys? It’s Graham here. Now, I usually don’t record informal videos without a whole bunch of charts and graphs and fancy research, but something needs to be said about the current state of the housing market and the direction it’s headed. I do…
What we've learned in 100 Episodes - Smarter Every Day 100!!
[party whistles] Hey it’s me Destin. Welcome back to Smarter Every Day. A very special Smarter Every Day. The 100th episode, but not only that, it kind of coincided with a million subscribers, so thank you very much for your support. And because of that, …
Zeros of polynomials introduction | Polynomial graphs | Algebra 2 | Khan Academy
Let’s say that we have a polynomial ( p ) of ( x ) and we can factor it. We can put it in the form ( (x - 1)(x + 2)(x - 3)(x + 4) ). What we are concerned with are the zeros of this polynomial. You might say, “What is a zero of a polynomial?” Well, those …