7 Essential AI YouTube Channels
How to stay updated with the latest and coolest machine learning advancements
Machine learning has been one of the hottest topics around for several years. Everyone is talking about it; organisations of all sizes are embracing these technologies.
Harvard Business Review describes the role of a data scientist as “The Sexiest Job of the 21st Century”. They are critical for business decision-making, and they are desperately in demand.
Spending on machine learning research is growing, from $1.58 billion in 2017 to a projected $20.83 billion in 2024. On the academic scene, researchers publish approximately 100 new machine learning papers every day on Arxiv.
One paper every 15 minutes!
This chart shows the level of excitement and progress in machine learning. How could anyone keep up with advances in this field? 🤔
In this article, I will introduce seven YouTube channels that share our excitement. These channels aim to uncover the latest and coolest developments in machine learning.
#1. Two Minute Papers
First on the list, the one I love the most, Two Minute Papers by Károly Zsolnai-Fehér. His videos are entertaining, engaging, and an absolute joy to watch. He selects and reviews the most exciting research developments. OpenAI on DotA games, detecting DeepFake, simulating physics, or cloning your voice. He has it all in his channel.
Károly will describe what is so brilliant about the research he is reviewing. Even for the most advanced and recent AI research, watching his videos do not need a PhD in Computer Science. His demonstrations are so clear that someone without machine learning knowledge can understand.
Since this is my favourite channel, let me show you one of the videos, OpenAI Plays Hide and Seek and Breaks The Game! 🤖 I’m sure you’ll enjoy it.
It’s about two AI teams (hiders and seekers) playing hide-and-seek against each other. When one side learns a new strategy, the other team will adapt to counter back. The agents played until a point where they learn to abuse the game physics engine. Bravo!
YouTube: Two Minute Papers
#2. Lex Fridman
Lex Fridman conducts interviews with the artificial intelligence industry superstars. Elon Musk, Yann LeCun, Yoshua Bengio, Eric Schmidt? Yea, Lex has already interviewed them.
These interviews can give you a glimpse on the future, as they will share what they are working on. Finding a cure for cancer using AI; Adobe is working on AI; Neuralink; new iRobot vacuum capabilities; YouTube algorithm. You can find a vast selection of topics.
Lex Fridman is a research scientist at MIT. He conducts deep learning and autonomous vehicle lectures. If attending MIT deep learning lectures interests you, Lex shares them on his channel too.
YouTube: Lex Fridman
#3. Henry AI Labs
Henry AI Labs. This channel is rich! Natural language processing, reinforcement learning, generative adversarial network, weakly supervised learning. All the hottest topics covered.
Henry summarises new papers that research labs from companies such as Google and Facebook are publishing. Definitely useful if you want to stay up to date with the latest AI developments. He gives an overview of the newest research topics and real-world applications, and you don’t have to digest the full paper on your own.
YouTube: Henry AI Labs
#4. ArXiv Insights
Research papers can be a bit dry & take a while to read, right? ArXiv Insights by Xander Steenbrugge is here to help! He aims to simplify and explain technical details in machine learning papers.
At the time of writing, this channel has only 11 videos. But his high quality and reliable content have attracted 45K subscribers. He breakdown complicated topics like generative adversarial networks, into step-by-step bite-size explanations.
While watching his videos, I was thinking, “man, how I wish my lecturers would explain concepts like him”.
YouTube: ArXiv Insights
#5. Yannic Kilcher
Yannic Kilcher focuses on explaining deep learning research papers. He introduces the paper and describes the novelty behind it. He reads the paper meticulously to you, and he tells it like a story. He covers topics on deep learning architectures, natural language processing and reinforcement learning.
The level of details is incredible. Yannic breaks up a dense paper into parts and uncovers each idea paragraph by paragraph. He explains and scribbles on the screen to walk you through the thought process. When there are concepts not covered by the paper, he references additional resources to reveal any concepts outside of the paper.
At the end of the video, it feels like I have a full understanding of the paper as if I have read thoroughly for at least three times. How I wish he reviews the papers I am reading.
YouTube: Yannic Kilcher
#6. Leo Isikdogan
Leo Isikdogan showcases and concise research topics into practical, real-world application. He covers topics on machine learning and computer vision. He also creates educational how-to machine learning tutorials.
Leo is a researcher at Intel, and he introduces his work in the channel. One that I like is the research on Eye Contact Correction using Deep Neural Networks. In his video, he explains that maintain eye contact during a call requires looking into the camera rather than the screen. His work is to correct the gaze, to improve the quality of video conferencing experience.
I love it when research has real-world application and impact.
This channel is also great for anyone starting on deep learning, so much to learn from. Check out the playlists in his channel, the deep learning crash course series and hands-on with TensorFlow coding sessions.
YouTube: Leo Isikdogan
Last by not least, Kaggle, Your Home for Data Science. It is the largest data science community. Users can find datasets and code within the Kaggle platform. They can publish notebooks to share codes, collaborate and learn from each other. There are competitions with cash prizes; competing is a great way to learn, and cash is definitely a great motivation to do well.
There are “how-tos” coding tutorials on deep learning, building dashboards, and data processing. They have live coding sessions on how to solve data science problems.
For updates on the latest NLP research, there is a playlist, Kaggle Reading Group. She reads and discusses NLP research papers. She will explain every single paragraph, diagrams, code, mathematics equations to you. If you are a new researcher or new to academic papers, watch these videos, and you will learn how to dissect papers.
There you go! I hope you will enjoy these channels. Let me (and other readers) know in the comments if you have channels to recommend.
I will end here with this quote by Roy Bennett.
There are five important things for living a successful and fulfilling life: never stop dreaming, never stop believing, never give up, never stop trying, and never stop learning.