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
Science Is an Error-Correcting Mechanism
So getting back to good explanations, where do these explanations come from? There’s currently an obsession with induction. Induction being the idea that you can predict the future from the past. You can say, “I saw one, then two, then three, then four, …
A Brief History of Dogs | National Geographic
Long before we raised livestock and grew crops, humans lived side by side with dogs. It’s widely accepted among scientists that dogs are descendants of wolves. In fact, their DNA is virtually identical. But how exactly did a fierce wild animal become our …
Comparing animal and plant cells | Cells and organisms | Middle school biology | Khan Academy
So, let’s play a game of spot the difference. Now, if you were asked to spot the difference between these two pictures, you’d probably laugh and say that’s too easy because it’s obvious that this picture of a lion on the left is nowhere close to looking …
Disability and long term care insurance | Insurance | Financial literacy | Khan Academy
Talking about insurance is never a lot of fun because you’re talking about thinking, or you’re thinking about things that most of us don’t want to have to deal with in life. I’m going to continue that trend by talking about two new scenarios of insurance,…
Is Pluto a planet?
Pluto: Planet or not? Before we can answer this question, we need to know what the word “planet” is for, and that takes us back to the ancient Greeks who called Mercury, Venus, Mars, Jupiter, Saturn, the Moon, and Sun planets. Basically, if it moved acros…
Worked example: Product rule with table | Derivative rules | AP Calculus AB | Khan Academy
The following table lists the values of functions F and H and of their derivatives f prime and H prime for x is equal to 3. So, let’s just tell us when x is equal to three, the value of the function is six. F of three is six. You could view it that way: H…