How I learned AI from YouTube (more practical than Harvard)

I got my bachelor's degree from Harvard

University, but I got all my practical

AI knowledge from YouTube University.

The abundance of free, highquality

knowledge about AI on YouTube is just

amazing. You can turn the world's

leading AI practitioners into your

personal on demand tutors. You just need

to follow the right people. Today, I'll

talk about the specific channels and

videos that I found to be the most

useful on my AI learning journey. These

are the types of channels or videos that

I've learned the most from. Let's dive

into each. The first type is video

podcasts with AI founders or builders.

These are in-depth interviews with

industry practitioners who work at AI

companies or have built AI products

themselves. Here's a specific list of

shows that I recommend. This is

firsthand information because you're

listening to the founders themselves

speak rather than someone else

interpreting what they said. For

example, the founder of Cursor appeared

on Lenny's podcast to talk about the

future of AI coding. and the founder of

the browser company appeared on every

podcast AI and I to talk about their

journey building the AI browser DIA.

These podcasts are often hours long and

you can also find them as audio versions

on podcast. But I prefer to watch the

video versions because of three reasons.

First, they often include demos of AI

products or visual information and

you'll be missing out on these if you

are just listening to the audio. Second,

these podcasts have very high

information density and it's easy to get

distracted when listening to audio. I

find myself being able to focus and

absorb a lot better when watching the

video version because it's a lot more

immersive. Three, I often find myself

wanting to take notes when watching

these and it's much easier to take notes

when watching a video. The second type

of videos is talks at AI conferences by

industry leaders. Here are five specific

conferences that I recommend and all of

their sessions are recorded and

available for free on YouTube. Honestly,

I hardly attend offline conferences

these days because I found it to be much

more effective to just watch the talks

on YouTube in the comfort of my own

house. For example, Andrew gave a talk

at YC's AI startup school recently that

I found to be really insightful. The AI

engineer world's fair is probably the

best conference for AI engineers with an

absolutely insane list of speakers,

including OpenAI co-founder Greg

Brockman. Take advantage of these videos

because you don't need to purchase

expensive tickets to attend the

conferences. You don't need to take the

time to physically be there. Just absorb

the wisdom from the online recordings.

The third type is the official channels

of OpenAI and Anthropic. Whenever they

launch new models or products, I make

sure to watch the original launch videos

because this is firsthand information.

They also have podcast and shows where

they interview their leaders or

researchers. For example, OpenAI

recently released a podcast with Sam

Altman and Anthropic has videos where

their researchers discuss how to build

AI agents. A lot of people talk about

OpenAI anthropics models and products,

but I've always found it helpful to go

back to the primary sources to listen to

the builders themselves. The fourth type

is hands-on tutorials on AI tools. This

is particularly helpful if you want to

get into VIP coding or start building AI

products yourself. I don't have a coding

background, but I've learned a ton about

VIP coding through watching the

following YouTube channels. They do an

amazing job translating technical

concepts into layman language and they

literally walk you through these tools

step by step and are very easy to follow

even for people without technical

backgrounds. For example, I recently

followed this tutorial by McKay Wriggley

on how to connect cloud code to the

note-taking app Obsidian and I was able

to follow through every single step.

Make sure to not just watch but also

follow along and get your hands on these

tools at the same time. The fifth

category is literally just Andre

Karpathy. Anyone who's serious about

learning AI needs to watch Andre's

videos. He's not just a wellestablished

AI researcher, but also a world-class

educator. I found his deep dive into

LLMs like CHBT to be the best resource

on the entire internet to understand how

large language models work. He has a

talent for explaining technical concepts

in a way that anyone can understand. His

videos are hours long, but you'll find

these to be probably the most worthwhile

hours you've ever spent on the internet.

So, there you go. These are my YouTube

University curriculum for learning AI. I

truly believe that when a technological

revolution arrives, it resets the

playing field because everyone needs to

learn something new. Fancy degrees

matter less than curiosity and drive.

What you need is not an elite university

acceptance letter. is the internet

connection and the willingness to