The Reboot Podcast

We read a lot of AI books so you don’t have to

It’s been approximately three years since the launch of ChatGPT vaulted “A(G)I” into public consciousness. No coincidence that, around the 2.5-3 year mark, a bunch of AI books have now hit the market…. Jasmine, Jacob, Shira, and I talk through as many as we can get to in this long(! sorry) podcast. In reverse chronological order: * The Scaling Era: An Oral History of AI, 2019-2025 by Dwarkesh Patel and Gavin Leech (October 2025) * If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All by Eliezer Yudkowsky and Nate Soares (September 2025) * What Is Intelligence?: Lessons from AI About Evolution, Computing, and Minds by Blaise Aguera y Arcas (September 2025) * The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want by Emily Bender and Alex Hanna (May 2025) * Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI by Karen Hao (May 2025) An abridged transcript is below, or jump to the bottom of this email to get our “buy/borrow/skip” (spoiler: unfortunately, most people will probably only find around 1.5 books worth reading). As always, audio version is more than a little spicier than the transcript. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit joinreboot.org [https://joinreboot.org?utm_medium=podcast&utm_campaign=CTA_1]

By And For Technologists

TL;DR: Reboot has a new mission: We are a publication by and for technologists. We are also open for pitches [https://forms.gle/1g76BBXaB1ygUhbw8] (and now pay $750 for newsletter essays!). Keep reading and listening for more context on how the editorial board came to this decision—this talk was first given at the Kernel 5 [https://shop.kernelmag.io/products/kernel-5] magazine launch in San Francisco—and for examples of the kinds of pieces we’d love to have. Talk: By and For Technologists Hi! My name is Jasmine, and I’m the director and cofounder of Reboot. Thanks so much for being here. Kernel launch parties are always one of my favorite parts of what we do. Online writing often doesn’t feel real until everyone shows up in physical space together. Reboot turned five years old earlier this year, which is pretty crazy. Lots of things have changed since jessica dai [https://substack.com/profile/2572689-jessica-dai] and I started it as undergrads in 2020—in our lives, in the tech discourse, the industry writ large. Back then we were all talking about Facebook and the end of democracy, or freaking out that OpenAI wasn’t really open because they didn’t release GPT-2. At the time, Jessica and I started Reboot because it felt urgent to articulate a vision of technology that wasn’t about total refusal or hype. We wanted pragmatic, clear-eyed optimism; and we wanted a community of fellow early-career technologists to think through hard questions with. A recognition that tech is part of our strategy for achieving the goals we want, whether reproductive rights or more fun telephone poles [https://joinreboot.org/p/tall-dead-trees] in our communities. In 2021, when we were putting together Kernel’s first issue, I holed up in a lodge in Asheville, North Carolina and wrote a manifesto—“Take Back the Future! [https://joinreboot.org/p/manifesto]”—about what a “progressive techno-optimism” could look like. Well, a lot more people are talking about “techno-optimism” these days, and tragically not in the way that we meant. We waged a noble battle to reimagine the term, but unfortunately, Andreessen Horowitz has far more money and more Twitter followers than we do. Now, the tech industry has followed Marc’s lead and taken a turn to the right. Log onto x.com [http://x.com], and you’ll find infinite e/acc memes about how everyone who mentions ethics or safety or sustainability is automatically a doomer decel. According to Marc Andreessen’s techno-optimist manifesto, if you’re getting in the way of pure acceleration and profit—no matter the reason why—you are the enemy. And as I’ve spent more time reporting on Silicon Valley culture this year, one of the trends I’ve been most surprised and disturbed to observe is not merely a shift to the right, but the emergence of a nihilism [https://x.com/jasminewsun/status/1948192092939022348] about whether tech should serve humans at all. Here’s something I hear reasonably often: AGI is going to be so much smarter than us, so we should just hand over the reins and make them our worthy successors. If LLMs can now ace the IMO, why not make them president and CEO too? They should run the institutions, not us. Relatedly: the idea that Mars colonization or Cluely or whatever is some kind of natural, inevitable endpoint to humanity; that regardless of whether a product is something we want, there is a moral duty to bring it into existence—to enact the market’s and technological history’s will. This style of thinking is quite common among high-up people in Silicon Valley. But I think it’s low-agency and anti-human, to say the least. Reboot’s editorial board has been talking about how our publication should position ourselves in this strange moment. And the forcing function came to this: How many more times do we want to repeat, “Not that kind of techno-optimism”? I have always defined “techno-optimism” not as an uncritical belief that more technology equals more good—but rather optimism as agency, a faith that humans, as the builders of tools, can shape these incredible forces to achieve the values and goals that we define. Sand does not think until we make it. Modern civilization has always been about finding social and technological solutions to bring out the better angels of our nature—to transcend our monkey-brains and pursue our higher values and aspirations. For Reboot’s next era, we want to re-center the human and the intentional act of creating. Technology is something we do to the world, it is something we choose, and we humans are responsible for those choices. That leads us to a new mission: Reboot is a publication by and for technologists. I view this as less a shift than a clarification. In short: a technologist is anyone who exercises agency to shape technologies toward their goals. It’s a mindset, not a job title; an orientation, if you will. It includes many software engineers and founders, of course, but also makers of home-cooked apps [https://www.wsj.com/tech/ai/your-next-favorite-app-the-one-you-make-yourself-a6a84f5f] and clever Zapier workflows, hacktivists and traffic cam artists [https://trafficcamphotobooth.com/]. It’s an orientation of active play, not passive consumption. It combines the critic’s eye for spotting the flaws in a system with the artist’s or entrepreneur’s creative solutions. It’s not just posting about the problem but doing something about it. The technologist says: These systems were made once and they can be remade again. The world is a museum of passion projects. I will not accept things as they come out of the box. As Kevin wrote in his Kernel 5 editor’s note [https://joinreboot.org/p/where-do-we-draw-the-line], technologists are players of infinite games. Reboot will continue publishing essays, interviews, and other creative works by technologists. We believe in lived experience and tacit knowledge; the deep understanding that comes from the personal experience of being “close to the machine,” as Ellen Ullman described in her memoir [https://www.goodreads.com/book/show/486625.Close_to_the_Machine?ac=1&from_search=true&qid=J5ZbHIJ90w&rank=1] of the same name. As editors, we’ve noticed that both hype and doom deal in vague, sweeping proclamations. Most people who believe AGI will cure cancer or start WW3 tomorrow have worked neither in medicine nor in military strategy. Thus, we view the specificity of technologists’ experience—the fact that they know intimately where tools work versus don’t, how to tweak them to work a bit better—as a potent vaccine against bad ideas. As always, we are especially excited to work with people who are not professional writers. We want to develop ideas from practitioners: people doing stuff on the ground. Field-building manifestos, essays about projects you’ve built, and interviews (anonymous or otherwise) with the people doing the most interesting, challenging work in the space. We are also more than doubling our newsletter pay rates, so do pitch us [https://forms.gle/1g76BBXaB1ygUhbw8]! Writing is not quite as lucrative as a $100 million comp package from Meta, but we hope it will be at least somewhat more fulfilling. And again, thanks for being with Reboot, whether you’re an OG who subscribed in 2020 or a new reader who stumbled through the door today. I’m keenly aware that the market does not reward reflection on why we build what we build, which makes it all the more meaningful that you have decided to do it anyway. Thank you to Gray Area for hosting us here—they’re an incredible art and tech venue in SF, and do lots of other great events—and to all the incredible writers and contributors to Kernel Issue 5. Have some drinks! Buy some magazines [https://shop.kernelmag.io/products/kernel-5]! And thank you all for coming. We’re all super excited about this new direction—which emerged from lots of rich discussion and debate—and hope that you are too. You can pitch us here: Thanks for being here in year five! — Jasmine & Reboot team This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit joinreboot.org [https://joinreboot.org?utm_medium=podcast&utm_campaign=CTA_1]

The Future is Constructed, Not Predicted

Shortly after the AI 2027 report was released, my friend Saffron posted a tweet [https://x.com/saffronhuang/status/1907863453009867183]/mini-blog in response: Looking to “accurately” predict AI outcomes is… highly counterproductive for good outcomes. The researchers aim for predictive accuracy and make a big deal of their credentials in forecasting and research. (Although they obscure the actual research, wrapping this up with lots of very specific narrative.) This creates an intended illusion, especially for the majority of people who haven’t thought much about AI, that the near term scenarios are basically inevitable--they claim they are so objective, and good at forecasting! Why implicitly frame it as inevitable if they explicitly say (buried in a footnote in the “What is this?” info box) that they hope that this scenario does not come to pass? Why not draw attention to points of leverage for human agency in this future, if they *actually* want this scenario to not come to pass? I, too, was somewhat confused about the report, to put it lightly, and wanted to talk through it together. We try to understand the report and forecasting in general, and our conversation turns out to be less of an AI 2027 hate train than we initially thought! By the end, we end up coalescing around these three ideas: (1) the very act of prediction has an impact on the future; (2) ideally, forecasts should be empowering, rather than disempowering; (3) evaluating forecasts is messy business, so the intentions of the forecasters matter. — Jessica Listen here on Substack (web [https://joinreboot.org/p/ben-recht] or app [https://substack.com/app/app-store-redirect]), or subscribe on Apple Podcasts [https://podcasts.apple.com/us/podcast/reboot/id1614092110] or Spotify [https://open.spotify.com/show/7b99Odl5zCcLf8wHm3fLPv]. A transcript and takeaways will be published with each episode. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit joinreboot.org [https://joinreboot.org?utm_medium=podcast&utm_campaign=CTA_1]

Mr. Beast Saying Increasingly Large Amounts of Money

Mr. Beast Saying Increasingly Large Amounts of Money By Morry Kolman [https://wttdotm.com/] Mr. Beast Saying Increasingly Large Amounts of Money is a compilation of over 2800 clips from 206 Mr. Beast videos. This is the abridged 12 minute version—both this and the full length hour-long cut are also available on my YouTube [https://www.youtube.com/@WTTDOTM] and my website [https://wttdotm.com/mrbeast]. The work is intended to distill the content of the most popular YouTuber in the world down to one of its core motifs: the promise of the next number being even bigger. Mr. Beast, born as Jimmy Donaldson, tries to present himself as apolitical. But when you convince 4% of the world to hit the red Subscribe button, you get the politics for free. These clips are those politics. Donaldson has often explained that the over-the-top excesses of his content—both in conspicuous consumption and even more conspicuous philanthropy—enable him to snowball money and influence that he can then leverage to make even more positive change in the world. This framing is overwhelming. Donaldson has been in the hot seat for allegations of worker safety violations, hawking moldy knock-off Lunchables, and exploiting the poor, disabled, and destitute for views. Nothing, however, can avoid getting subsumed by the bottom line. “Is Mr. Beast good?” quickly becomes a proxy for “Should Mr. Beast's videos exist?” and if there is anything Mr. Beast does well, it is documenting the size, appreciation, and impact of his content at an incomparable rate. To argue against Donaldson is to rebuke the planting of 20 million trees, wish 100 people remained blind, and contend that those homeless people should not, actually, have been given $10,000. Mr. Beast Saying Increasingly Large Amounts of Money is an attempt to critique him on his own terms. In their proper context, Donaldson's absurd monetary figures not only make some level of sense, but engage the viewer on an emotional and entertaining level. Boiled down to concentrate and injected intravenously, though, they are a hypnotic experience of whiplash. It’s so easy to watch the numbers go up, increasing in opulence. At the same time, what’s actually happening—the deployment of real wealth and capital in service of making the world’s already largest creator get even more views—is uncanny. In incessant and mesmerizing form, this video portrays a 2020s version of the American Dream. Whether through extreme challenges, complete luck, or simply being a good supporting character, the beneficiaries of these videos receive houses, cars, and shiny briefcases of cash. In the video, though, this “philanthropy” is contextualized by the spectacle of consumption around it. Donaldson gives $50,000 to teachers, then dishes out $70,000 on a golden pizza. He spends tens of thousands to blow up fireworks in the sky, and drops hundreds of thousands back down from planes. He shells out $1 million on groceries for the hungry—and wastes the same amount on lottery tickets, all for the love of content. With his (often literal) piles of money, Mr. Beast wields the ability to change lives at will. Unfortunately it is a power he uses indiscriminately, self-servingly, and ostentatiously. Jimmy Donaldson does not perform acts of kindness, he purchases views, and in this video we watch that transaction happen several thousand times. Mothers cry, children scream, and a guy named Mac gets buried alive. In return, he has not dipped below 100 million views in over four years. Methodology First, for the main source of data, I chose all Mr. Beast videos with uploaded (ie. non-auto-generated) transcripts—a total of 229 out of 837 published videos on his flagship channel. This gave me a source of processable ground truth about where money was mentioned and also limited the videos to those published the last 6 years, which make up the majority of his meteoric rise. Then, I downloaded the videos in 360p and scraped their transcripts for every occurrence of a dollar amount, logging each mention with its sum, video, and context in a database that I would build on top of as I nailed down the exact timing. I used those contextual timestamps to make rough clips that I fed into the open source AI tool Whisper [https://github.com/ggerganov/whisper.cpp] to (a) get a more precise measurement of where “X dollars” was actually said and (b) standardize and double check that my first scrape had gotten the amount correct. Finally, as many of the clips were still off by a few annoying and noticeable fractions of a second in any direction, I made a script that allowed me to go through each entry individually, trim or extend the clip on either end, and modify the amount one last time if my first 2 methods had failed. After all 2800+ were processed—a task that took weeks—I made a final set of clips out of higher quality versions of the videos and used Premiere to make the film’s final dizzying supercut you see before you. 90% of data science is data cleaning, and I have kept this overview pretty high-level in the interest of making it accessible to a wide audience. A much longer and more technical dive into the steps needed to go from a raw YouTube archive to this video—including everything from token suppression, the comparative benefits of transcription libraries, counterintuitive ways to standardize and parse numbers in natural language, and debugging audio desyncs in clip concatenations - may appear in the future on my website. Reboot publishes essays on tech, humanity, and power every week. If you want to keep up with the community, subscribe below ⚡️ 💝 closing note This project took literal hundreds of hours to complete, so thank you for watching! It could not have happened without Sam Lavigne’s Infinite Video [http://lav.io/notes/infinite-video-showcase] class at the School for Poetic Computation [https://sfpc.study/], which gave me the opportunity to develop a proof of concept last year. There’s a lot of content out there, and critique-through-clip-compilation is a fun medium. I encourage you to give it a try :) Your fellow brains in rot, — Morry & Reboot Team This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit joinreboot.org [https://joinreboot.org?utm_medium=podcast&utm_campaign=CTA_1]

Probability at the End of the Information Age

Surprise, another podcast! (Apparently Kelin’s [https://joinreboot.org/p/kelin-zhang] was the “first”, so this is the “second.” But, from 2022-era Reboot, we have at least 3 older audio posts [https://joinreboot.org/podcast] — I still think they’re excellent, so do check them out if you enjoy audio!) On certain slices of the internet, Ben Recht might be known as Substack power-user [https://argmin.substack.com/] (“that one prof who blogs about math”), professional hot-take-haver [https://x.com/beenwrekt], or “recurring podcast guest [https://goodbye.substack.com/p/do-we-like-living-in-dataworld-and].” He’s also a computer science professor and one of my PhD advisors at Berkeley. I recently sat down with him to talk about… Nate Silver’s new book, understanding the world through statistics, the academic rat race, and the psychology of it all. Listen here on Substack (web [https://joinreboot.org/p/ben-recht] or app [https://substack.com/app/app-store-redirect]), or subscribe on Apple Podcasts [https://podcasts.apple.com/us/podcast/reboot/id1614092110] or Spotify [https://open.spotify.com/show/7b99Odl5zCcLf8wHm3fLPv]. A transcript and takeaways will be published with each episode. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit joinreboot.org [https://joinreboot.org?utm_medium=podcast&utm_campaign=CTA_1]