Want to Learn About AI? Follow These 5 Robotic YouTube Stars
When you think of people learning about artificial intelligence, deep learning and image recognition, you probably picture lecture halls crammed with laptop-toting computer science majors and a lecturer in front filling a digital whiteboard with forbidding jargon and code. You wouldn’t be wrong.
Universities across the country from MIT to Stanford are producing bumper crops of AI-savvy grad students, and even undergrad programs such as one at the University of Tennessee at Knoxville have introduced B.S. degrees that focus on AI. If you’re reading this, though, you’re probably a subject-matter expert in another field that’s not AI—but you’re aware of how AI could transform your field. So you want to learn about it without getting bogged down.
Enter YouTube. To make getting up to speed on AI as simple and accessible as possible, we’re giving you the lowdown on some great YouTube AI resources that you can get started with as soon as, well, right now. No tuition necessary.
Created by Canadian data scientist Jagannath Rajagopal, the DeepLearning.TV channel explores deep learning, the area of AI that’s involved in teaching machines how to see and make sense of the world around them. Across its more than three dozen videos, DLTV uses clever animations and everyday language to provide answers to questions such as “What is a neural network?” and “What is a Deep Learning library?” Even better, no video runs longer than 10 minutes and most clock in at about three or four—and none of them get mired in math or code, instead opting to use conversational English to explain the broad concepts rather than the computer-science nitty gritty. Watch even just a few of these videos and you’ll be smarter about some of the basics of AI.
> Best for: Learning about AI in broad strokes.
The Artificial Intelligence Channel
Think of the Artificial Intelligence Channel on as an eclectic, curated stream of videos from all over that have to do with the intersection of AI and society. Recent examples: Wired magazine contributor Tom Simonite’s appearance on C-SPAN’s Washington Journal show, in which he explained how AI is being deployed at Google and Facebook; a conference presentation by AI pioneer Stuart Russell titled “Building Artificial Intelligence That is Provably Safe & Beneficial”; and a video of the very first Congressional hearing on AI, at which scientists from NASA and Microsoft explained AI in terms that, well, even a congressman could understand. Unlike DeepLearning.TV, you’re unlikely to want to watch all the videos on this channel; instead, subscribe to it and sample the videos that pique your interest.
> Best for: Exploring the public policy implications of AI.
Artificial Intelligence (AI)
Another channel that curates videos from all over, Artificial Intelligence (AI) seems to go a little farther afield in sourcing its clips. For instance, a video titled “AI Robots Passed the Chinese College Entrance Exams” is a report from the China Global Television Network, “How Artificial Intelligence AI is Changing Financial Services” is a recording of an IBM Watson webinar, and “How Netflix Leverages Artificial Intelligence AI and Big Data” is video of Brian Sullivan, director of streaming analytics at Netflix, speaking at a conference in Vancouver. If filtering out the useful links from the nonstop barrage of algorithmically-generated “AI” Google Alerts is getting to you, this nicely curated collection of videos will come as sweet and informative relief.
> Best for: Hearing from leading AI experts at global tech companies.
Two Minute Papers
Two Minute Papers serves up two new videos per week about “awesome scientific works,” and though it doesn’t focus exclusively on Artificial Intelligence, it’s obviously one of the channel’s core obsessions. Recent videos—which actually typically run a little longer than two minutes—include “AI Creates Facial Animation From Audio,” “AI Learns to Synthesize Pictures of Animals” and “AI Learns To Improve Smoke Simulations.” If you want to see how AI is leaping from the theoretical to the practical, this channel’s a great destination.
> Best for: Sourcing gee-whiz inspiration, particularly when it comes to computer-vision-related applications of AI.
Siraj Raval: Artificial Intelligence Education
On his YouTube channel, San Francisco-based developer Siraj Raval says that he’s been called the “Bill Nye of Computer Science” and the “Kanye of Code.” Given the quirky sense of humor he displays in his videos, it’s entirely possible that Raval was the first to call himself those things. But no matter—his frequently updated library is highly watchable, particularly the videos that spotlight his more analytical/philosophical musings, such as “Why is Elon Musk Connecting Brains to the Internet?” and “How to Prevent an AI Apocalypse.” Be forewarned, though, that even though Raval epitomizes the YouTube creator aesthetic and vibe with his lively style, many of his videos, such as “How to Make an Image Classifier: Intro to Deep Learning” and “Deep Q Learning for Video Games,” quickly get into the weeds, so this channel is definitely not for beginners only.
Best for: Pure entertainment value and YouTube-creator-style charisma, even if Raval is a little (or a lot) wonky.
Illustration by James Gilleard