Fourteen Podcast Episodes about AI
Yuri Balasanov and I are teaching a course on what engineering managers need to know about AI. A student noticed that I referred to a few podcasts in the lectures. So, she asked for some recommendations.
I have a lot more podcast recommendations coming. But, to stick with AI, here are some that I’ve liked that cover different aspects.1
ARXP (the AI X-risk Research Podcast) Episode 20 - 'Reform' AI Alignment with Scott Aaronson. Scott Aaronson is a leading researcher in quantum computing. I’ve learned a lot from his blog and interviews. He has spent the last year at OpenAI researching AI safety. This is a must-listen if you are curious about AI safety and threats. He is spending a lot of time on watermarking. Besides AI, he provides a glimpse into science— we should work on problems where we can make progress.
Talks at Google. Prediction Machines; Avid Goldfarb and Ajay Agrawal. This podcast is based on my 2nd most recommended AI book, Prediction Machines. The podcast is about four years old, but the ideas have aged perfectly.
Lex Fridman. I’ve listened to many of these 2-3+ hour interviews, mostly the AI or science guests. I usually learn a lot. Here are a few AI ones:
#333 Andrej Karpathy: Tesla AI, Self-Driving, and more. Good information on using deep learning for self-driving cars. Good overview of deep learning. I especially liked the details on the engineering challenges in creating a self-driving car.
#252 Elon Musk: SpaceX, Mars, Tesla Autopilot, Self-Driving, Robotics, and AI. Good discussion on the size and quality team needed for self-driving cars. It was interesting that they wrote a C-compiler to gain speed and that he talked about deep learning replacing a lot of code (this seems to be a general trend).
#217 Rodney Brooks: Robotics. Rodney Brooks is my favorite AI skeptic. He will keep you grounded on what is possible and what is hype. This is a fun conversation where he tries to convince Lex that we aren’t as close to self-driving cars or AGI as the hype would lead us to believe.
#258 Yann LeCun: Dark Matter of Intelligence and Self-Supervised Learning Yann works at Meta/Facebook. This was recorded before the release of ChatGPT, but the conversation on LLM was good. Also, a discussion on how far away we are from getting machines to learn as we do. For example, a kid can be shown one picture of a toy giraffe and then recognize one in the wild, while machines need millions of images of both kinds.
My First Million. This podcast is mostly about entrepreneurship and breaking down businesses. I enjoy the energy and insights of the hosts, and their guests are people I don’t encounter on other podcasts. But unfortunately, they only occasionally talk about AI and then in the context of new business ideas. Here are two good ones:
Episode 438, Brainstorming ChatGPT Business Ideas With Billionaire Dharmesh Shah. Dharmesh is the founder of HubSpot. He knows the podcast listeners are a general audience, and he talks about turning text into vectors and how this creates many new business opportunities. He is also putting up his money: he bought chat.com for over $10M (and prompt.com for over $1M).
Last year, before ChatGPT, they did a podcast on the AI Gold Rush. I wrote a blog on it. But, again, you could feel the excitement for new business opportunities.
Embracing Digital Transformation podcast. Episode: WaveForm Artificial Intelligence. This is an interview with the co-founder and president of DataShapes. DataShapes analyzes data from sensors and waveforms for the defense and entertainment industry. Interestingly, analyzing sensor data is still in the early days. DataShapes is not using deep learning but has innovated on traditional algorithms like K-Nearest Neighbor (KNN) to make good progress on this problem. This reminded me of the power of KNN and traditional machine learning algorithms to do powerful analyses and explain the results.
Gradient Descent: Jordan Fisher of Standard AI. Jordan is the Founder and CEO of a company doing image recognition for self-checkout. The podcast covers some of what they’ve learned about deep learning. But, also, lots of practical advice on focusing on the problem (not the technology), advances in deep learning, sticking with what works, technical debt, business challenges, and much more.
a16z. This is the podcast of the famous VC company. I’ve listened to many good episodes on start-ups, trends, and business. Here are a couple of AI episodes.
AI and the Creator Economy with Karen X Cheng. Karen is an artist who has embraced the latest AI tools. She will get you excited about embracing all this new technology. And this episode came out before ChatGPT. So I would bet she is even more excited now.
The 1000x Developer. This is about Replit and their AI co-pilot. I was more impressed with the promise of the podcast than when I spent a few hours working with Replit. But I’m willing to blame myself for not giving it enough time. I am convinced that tools like ChatGPT and the AI co-pilot the economies of scale on coding.
Invest Like the Best with Patrick O'Shaughnessy podcast, the Alexandr Wang - A Primer on AI episode. Alexandr is the founder and CEO of Scale AI. This episode gives you a peak into the massive amount of data needed to run deep learning models— Scale AI is a company devoted to this. (My friend and former colleague Alex Banks-Watson recommended this one.)
Podcast.ai’s famous episode: “Joe Rogan interviewing Steve Jobs | AI generated." This episode is not about AI, but AI generated it. It came out in Oct 2022 and made a big splash. If you listen to it, there will be a few places where you can detect it is generated by an algorithm. However, If I were listening casually, I wouldn’t have guessed. Of course, Drake is in the news this week because people have created fake songs with his voice. Unfortunately, this will only continue.
Stay tuned for more podcast recommendations.
I’d love to hear your AI recommendations.
Two notes. First, I don’t specifically seek out AI podcasts, so I’m surely missing a lot of good content. Second, there isn’t always a natural link for podcasts. I listen to podcasts only on Spotify. I’ve linked to various sources above to help you find the podcasts.