
26 segments available
I recorded an AMA! I had a blast shooting the shit with my friends Trenton Bricken and Sholto Douglas. We discussed my new book, career advice given AGI, how I pick guests, how I research for the show, and some other nonsense. My book, “The Scaling Era: An Oral History of AI, 2019-2025” is available in digital format now. Links below! * Stripe Press website: https://press.stripe.com/scaling * E-book: https://www.amazon.com/Scaling-Era-Oral-History-2019-2025-ebook/dp/B0F22SKW5Y/ref=tmm_kin_swatch_0 * Print pre-order: https://www.amazon.com/Scaling-Era-Oral-History-2019-2025/dp/1953953557/ref=tmm_hrd_swatch_0 𝐄𝐏𝐈𝐒𝐎𝐃𝐄 𝐋𝐈𝐍𝐊𝐒 * Transcript: https://www.dwarkesh.com/p/scaling-ama * Apple Podcasts: https://podcasts.apple.com/us/podcast/dwarkesh-podcast/id1516093381 * Spotify: https://open.spotify.com/episode/4yso3gE93kHV6vGZw2cgtp?si=c1dfbe07b63343f8 To sponsor a future episode, visit https://www.dwarkesh.com/p/advertise 𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒 0:00:00 - Book launch announcement 0:04:57 - AI models not making connections across fields 0:10:52 - Career advice given AGI 0:15:20 - Guest selection criteria 0:17:19 - Choosing to pursue the podcast long-term 0:25:12 - Reading habits 0:31:10 - Beard deepdive 0:33:02 - Who is best suited for running an AI lab? 0:35:16 - Preparing for fast AGI timelines 0:40:50 - Growing the podcast
In this segment, the host announces the launch of his new book, 'The Scaling Era,' which compiles insights from interviews with AI experts across various fields. He emphasizes the book's relevance to understanding the future of intelligence and its implications for society, making a case for why everyone, including ordinary readers, should engage with its content.
"Today, this is going to be an Ask Me Anything episode. I'm joined by my friends Trenton Bricken and Sholto Douglas. You guys do some AI stuff, right? Yeah. We dabble. They're researchers at Anthr..."
The host explains the interdisciplinary nature of his book, highlighting contributions from AI lab CEOs, economists, and philosophers. He discusses the pressing questions addressed in the book, such as the nature of intelligence and the societal impacts of AI, making it clear why it is essential reading for a broad audience.
"it is. So, you know, these last few years, I've been interviewing AI lab CEOs, researchers, people like you guys, but also scholars from all kinds of different fields, economists, philosophers. ..."
This segment dives into the diverse perspectives included in 'The Scaling Era,' featuring insights from physicists and evolutionary theorists. The host shares how the book connects various fields of knowledge, allowing readers to see the interconnections between different ideas and enhancing their understanding of AI's evolution.
"next page is, of course, you guys going through the technical details of how these models work. And then there's so many different fields that are implicated. I mean, I feel like AI is one of the..."
The discussion shifts to the book's design and accessibility, highlighting how Stripe Press books present complex ideas with clarity. The host mentions the inclusion of diagrams and side captions that explain technical concepts, making the book approachable for readers unfamiliar with AI.
"So it does seem like a nice way to listen to all of the interviews in one digestible way. Yeah. There's two interviews I've done that haven't been released publicly before that are in the book. S..."
The conversation transitions to a question about AI models' limitations in making connections across different fields. The host reflects on a previous discussion with Dario and explores the idea that while humans can make unique connections, AI models struggle with this due to their training objectives.
"what a model is, these sorts of things. Actually, when we did our episode together, a bunch of people, I don't know if you saw this, independently made these blog posts and Anki cards and shit wh..."
In this segment, the host discusses the importance of reinforcement learning (RL) in enabling AI models to make novel discoveries. He emphasizes that current models lack the ability to form new connections and suggests that future advancements should focus on RL to enhance their capabilities.
"So the issue is, one of the questions I asked Dario is, look, these models have all of human knowledge memorized and you would think if a human had this much stuff memorized, and they were moder..."
The discussion continues with a focus on memory limitations in AI models. The host compares AI's memory capabilities to human memory, highlighting the challenges models face in retaining and utilizing information effectively, which impacts their ability to generalize knowledge.
"I think my answer at the moment is that the sort of pre-training objective doesn't necessarily- like it imbues with this nice flexible general knowledge about the world, but doesn't necessarily ..."
This segment explores the trade-off between memorization and the ability to extract meaningful insights. The host discusses how excessive memorization can hinder generalization and the importance of balancing these aspects for effective learning and understanding.
"Exactly, yeah, or like someone worked on it, it was awesome, it got far, but, another excited Anthropic employee then iterated on the memory scaffold and was able to very quickly improve on it. ..."
The host addresses a question about career advice for young individuals entering college. He emphasizes the importance of gaining leverage and deep technical knowledge while also encouraging exploration and adaptability in a rapidly changing job market influenced by AI.
"Yeah. Terrence Deacon, whose book you recommended, had this interesting insight about how we learn best when we're children, but we forget literally everything that happened to us when we were c..."
Continuing the discussion on career paths, the host reflects on the unpredictability of career planning in the context of AI advancements. He shares personal experiences and stresses the value of trying different things to discover one's passion and potential.
"become a podcaster? I feel like that job's still gonna be around. It's funny, because I studied computer science, and in retrospect- at the time, you could've become a software engineer or some..."
In this segment, the host reveals his criteria for selecting podcast guests. He emphasizes the importance of genuine interest in the guest's work and the need for thorough preparation to ensure engaging and insightful conversations.
"stuff. But I think in a very wide part of the possibility spectrum you are managing enormous, vastly more resources than an individual could command today, and you should be able to solve so man..."
The conversation wraps up with a discussion on the value of engaging with influential figures in the AI community. The host shares insights on the balance between influence and genuine interest in creating meaningful podcast content.
"really hard to plan your career in general. And I don't know what advice that implies, because I remember being super frustrated. I was in college, and the reason I was doing the podcast was to ..."
Here, the speaker outlines their criteria for selecting podcast guests, emphasizing the importance of genuine interest in the guest's work. They explain that the research process is time-consuming, and they prefer guests whose work excites them, rather than those who are merely influential. This approach has led to successful episodes with lesser-known but impactful guests.
"So, the most important thing is, do I wanna spend one to two weeks reading every single thing you have ever written, every single interview you ever recorded, talking to a bunch of other people ab..."
The speaker shares a pivotal moment when they recognized the potential of their podcast as a long-term career. They recount the experience of negotiating ad spots and realizing the financial viability of the podcast, which opened the door to hiring help and expanding their efforts. This segment highlights the transition from hobby to business.
"that you realized you, that producing your podcast was a viable long-term strategy? I think when I was shopping around ad spots for a Mark Zuckerberg episode. And now when I look back on it, it's..."
In this segment, the speaker discusses the significance of having a personal vision when creating media content. They argue that successful new media often stems from an individual's unique perspective rather than collective efforts. The speaker encourages aspiring creators to pursue their passions and emphasizes the potential for financial success in this space.
"are there? What sort of things would you be excited to see in new blogs, podcasts? I wonder if you guys think this too, but I think this “Matt Levine of AI” is a totally open niche as far as I ca..."
The conversation explores the misconception of slow, compounding growth in media. The speaker argues that impactful content can gain rapid attention if it resonates well, citing examples of viral essays. They emphasize that improvement comes from engaging with a community and learning from feedback rather than merely accumulating followers.
"the three months of blogging on the side really gonna cost you? And people just don't actually seriously do the thing for long enough to actually get evidence or get the sort of RL feedback on, ..."
Reflecting on their journey, the speaker identifies a significant moment of success linked to a blog post that gained traction. They discuss how recognition from influential figures can propel a creator's visibility and the importance of articulating ideas clearly to resonate with audiences. This segment underscores the value of taking risks and sharing insights.
"Let's hear them. This is, this is the AMA. So I believe that slow compounding growth in media is kind of fake. Like, Leopold's situational awareness. It's not like he was building up an audience ..."
The speaker highlights how creating content has enabled them to build valuable networks across various fields. They share experiences of connecting with experts and how these relationships enhance their understanding and future content. This segment illustrates the interconnectedness of knowledge and collaboration in the podcasting landscape.
"your podcast. I distinctly remember you had your blog post on the Annus Mirabilis. And Jeff Bezos retweeted it, I think. I'm trying to remember if it was before that or not, but, yeah. I'm curiou..."
In this segment, the speaker discusses their reading habits and the influence of personal interests on learning. They emphasize the importance of engaging with material that resonates personally and how this approach enhances retention and understanding. The conversation touches on the balance between structured learning and personal exploration.
"Interesting. So move to San Francisco? Yes. If you're trying to do AI, yeah. Oh, very important question from Jacked Pajeet. How much can you bench? You can't lie because we both know the answer. A..."
The speaker shares insights gained from reading biographies, particularly focusing on LBJ's approach to effort and success. They reflect on the importance of doing thorough work to achieve significant outcomes and how this philosophy applies to their own endeavors. This segment encourages listeners to draw lessons from historical figures.
"Yeah. What's your favorite history book? There's a wall of them behind you. Oh, obviously the Caro LBJ biographies. The main thing I took away from those books is LBJ had this quote that he would..."
Here, the speaker discusses their strategies for learning complex subjects, emphasizing the importance of deep dives into topics before interviews. They highlight the difference between being a practitioner and asking insightful questions, advocating for a focused approach to understanding new fields. This segment provides practical advice for aspiring podcasters and learners.
"no ultimate impact, and still try doing that. You've shared on Twitter, using Anki. Or even, like, a Claude integration. Do you do book clubs? Do you use GoodReads? And what are you reading right..."
In this concluding segment, the speaker reflects on the challenges of setting long-term goals in the context of rapid advancements in AI. They discuss their current focus on growing the podcast and the uncertainty of future plans, highlighting the need for adaptability in a changing landscape. This segment encapsulates the overarching theme of navigating an unpredictable future.
"Oh. My friend Alvaro De Menard, author of Fantastic Anachronism. Can I just hold it up? Actually it's right here. I hope he's okay with me sharing this. But he made 100 copies of this translatio..."
In a light-hearted exchange, the host discusses his beard grooming routine. He shares tips on maintenance, including the use of beard oil and shampoo. This segment adds a personal touch to the conversation, showcasing the host's personality and humor.
"I don't know what to say, just genetics. I do trim it, but- No beard oil? Sometimes I do beard oil. How often? Once every couple of days. That's not sometimes! That's pretty often! But, do you hav..."
The discussion shifts to the qualities needed for running an AI lab, with the host and guests speculating on which historical figures would excel in this role. They explore the balance between leadership skills and technical expertise, providing a thought-provoking perspective on AI leadership.
"Oh, yeah. Oh, yeah. Which historical figure would be best suited to run a frontier AI lab? This is definitely a question for you guys. Oh. No, I mean, I'm curious what your take is first. You've s..."
The host discusses strategies for preparing for rapid advancements in AGI. He emphasizes the importance of being proactive in decision-making during critical periods of AI development. This segment underscores the urgency of staying informed and adaptable in a fast-changing landscape.
"How are you preparing for fast timelines? If there's fast timelines, then there will be this six-month period in which the most important decisions in human history are being made. And I feel lik..."
In the final segment, the host shares insights on the importance of distribution in AI-related content creation. He highlights how platforms like YouTube Shorts have significantly contributed to the growth of his podcast. This segment emphasizes the need for effective distribution strategies in reaching wider audiences.
"get asked about, is distribution. Distribution of AI? No, sorry. Like, Mr Beast-style distribution, where people, I think rightly, focus on the content, and if that's not up to snuff, I think you..."