Close All Tabs - Introducing: Dream Machines from KQED

Episode Date: August 3, 2026

We’ve got guests in the feed all August — a new series from KQED called Dream Machines! Today Morgan talks to co-hosts Alexis Madrigal and Robin Sloan about this podcast project that looks at AI ��� how it works, what it means, and how it feels to be living right in the epicenter of it all: San Francisco. More episodes will be dropping in the feed, every Monday in August. Guests: Alexis Madrigal, co-host of KQED’s Forum, author of The Pacific Circuit, and former senior editor at The Atlantic, covering technology.  Robin Sloan, author of New York Times bestseller Mr. Penumbra's 24-Hour Bookstore, along with the novels Sourdough and Moonbound.  Learn more about your ad choices. Visit megaphone.fm/adchoices

from KQED. Hey, Tabbers. It's Morgan, coming to you on a Monday, which is a little unusual for us because we've got an announcement. For the next four Mondays, we're handing over part of this feed to a new series from KQED, the station that makes close all tabs. It's called Dream Machines, and it's a podcast about AI, how it works, what it means, and how it feels to be living right in the epicenter of it all. San Francisco.

Don't worry, we're not going anywhere. you'll still get your regular dose of close all tabs every Wednesday. This is just a bonus drop for the next month. The hosts are Alexis Madrigal and Robin Sloan, two people who've spent their careers thinking about how technology and humanity are intertwined. AI is already reshaping our world, and the stakes are high. So ahead of the miniseries, I sat down with Alexis and Robin, to ask them how they got into this field,

which of them is the AI doomer, and why, even if you're a huge AI skeptic, I mean, guilty, it's worth digging under the hood to understand what this technology actually is and where it's headed. Here's my conversation with the hosts of Dream Machines. Alexis, Robin, welcome to close all tabs. To start off, Alexis, KQAD listeners might recognize your voice, but what is your relationship to AI? So, you know, I'm co-host here at KQED, a show called Forum. That's the big long-running current affairs program here, talking about all kinds of San Francisco Bay Area civic life stuff and arts and culture and all these things. Before that, I was a longtime technology reporter and edited technology coverage at the Atlantic Magazine for a long time.

So I was pretty immersed in the world of technology, even before this round of AI and all the things that have come around. Robin, what about you? Can you introduce yourself, your work? And yeah, your feelings about AI, big or small? Yeah, I'm a writer first and foremost. Novels starting with Mr. Penumbra's 24-hour bookstore and then a couple more after that that tend to be sort of on the edge between San Francisco Bay Area

magical realism and science fiction. And along with the writing and the imagining, for me there's always been just super healthy dose of technological investigation, which involves programming and building things and making websites, and then for about the last 10 years, getting, along with many other people, really, really interested in AI and in language models particularly. Okay, well, tell me about how you guys met. Oh, in the most perfect way for two friends in the 2000s, it was a...

Two nerds. Yeah, in San Francisco. It was a blog meetup, a blogger. You know, in this realm, Clay Shirky, in this realm of technology and society was coming to San Francisco and posted, hey, anybody out there want to meet up? And these two guys sure did. It sat next to each other. Yeah. And then later on, you know, we were just remained friends for a long time and then work together in a little industrial space. It was an old former olive canning plant down where Berkeley, Oakland and Emeryville all come together in the East Bay. And we can. called it the Murray Street Media Lab, which you could see it is hilarious. And that was cool. That really was a kind of a crucible, particularly for this conversation, because I had just gotten into this technology, you know, these nascent little mini-language models that were just kind of

becoming a thing around that time. And so you'd be over there working on your laptop, and I'd call you over and say, hey, come look at this thing. Hey, look at this weird paper. Yeah, yeah, yeah, and we would. And you'd look at stuff, we'd talk about it. We'd run little experiments. I mean, for a while there It really was a strange little media lab. And we kind of never stopped talking about it after that. True. Wow, what a beautiful meet cute. Emphasis on cute.

Okay, so YouTube have been friends for a very long time. And you've been thinking about these topics and talking about them together, both as friends and colleagues. And you agree on a lot about AI. Can you tell me about that? I think we both agree that this is a really important set of technologies. I think there is still a set of people out there who are like, no, this is all bullshit. This doesn't matter.

You can not pay any attention. I think we think to understand kind of the latest versions of these things because the baseline has really been shifting with their capabilities. So if you are saying a lot of stuff about AI having played with it in 2023, doesn't really make a lot of sense to people who are using the tools in 2026. I also think we're kind of rooting for something more interesting to happen than oligal. by a couple of companies. Well, the company, I mean, that's what I was about to say that I think there's a lot of people who are interested in this domain. But in that real businessy tech titan, who's up, who's down, who's going to IPO sense. Obviously, that is interesting and will be important, particularly to this city and its future.

But we are so much more interested, I would say, in, like, what is this? What's happening inside these models? You know, the questions of, like, what does it mean to think? What does it mean to write? What does it mean to be creative? Those are interesting questions anytime. They were interesting questions 100 years ago, a thousand years ago. They are in some ways sharpened by what we're seeing in these new machines.

That's so interesting. Okay. So this is something because I love drama, you know. I live for it. I love gossip. I love drama. I love all of that. Tell me about what you don't agree on when it comes to AI.

I'm more gullible. I think that's kind of. Yeah. You know, in a weird way, I don't know if you would agree with this, Alexis, but I, I got started with it earlier, and maybe as a result, I think today you are a little more enthusiastic about it. I mean, you're an enthusiastic user, an appreciative user of some of the tools. I mean, for practical purposes, you've built stuff for your life and, you know, different undertakings. And I might be a little bit more of a domer than you.

Do you think that's true? I do think that's true. Yeah. Yeah. I have a lot of, like, wonder and awe about it that I think, I just can't believe how much we encoded in, The fact that they can get so far by just like wamping the entire internet of text into these systems and then setting them off on their own after some,

you know, abstract mathematical reward and then what pops out on the other side is the thing that's like, I can help you with that. You know, like that to me, it's just, uh, it's unexpected relative to, you know, how people thought these things might work for most of computing history. And I think I would, want to believe it says something about the way our own brains work. And I think Robin disagrees with that largely. That's funny. I'm seeing our own, like our core temperaments

emerge because we both are very conscious of that mystery. And by the way, just to dwell on that point that there's two potentially trillion-dollar companies that will go public at some point in this city, like two miles apart at most. And at the core, the innermost heart of their technological engines is fundamentally a mystery, a black box, that even those programmers and engineers don't fully understand or understand much at all. So we're both aware of that. My, again, the temperamental thing, I am someone who kind of wants to like know what's going on and sort of be like, let's like plan out the steps. And so I really do focus sometimes on what I perceive as the recklessness of some of these companies and their releases. Okay, so on close all tabs,

we do tend to cover a lot of the concerns and dangers of AI. And, yeah, often the way people talk about AI is either, like, it's either side of the extreme, right? It's like super boostery. It's super like, this is going to save humanity or it's super doomery where it's like, this is going to destroy humanity. How does your approach on your show differ? And why do you think it's important for people to finally understand, like, the intricacies of AI and, like, where it's going regardless of how they feel about it? It's so interesting to hear you kind of frame it that way because I agree that's, I think that characterizes most coverage of, and maybe add in the business part that's kind of, you know, about, you know, IPOs and trillions of dollars. I would say that one of our moves is almost to say, okay, people, other people across the media spectrum and the political spectrum have got AI good or bad covered.

We want to, and we don't ignore those questions, but we want to focus on just this question of like, what are we talking about? How has it changed and like what is it becoming? There's also just so many other feelings that AI generates. You know, I'll just tell like, quick story for me, which I think is one of the origin points in my own interest in doing this podcast. I was in Ashland, Oregon, which is one of like the hippiest places of all time. And I happened to duck into their bookstore where the hippiest woman in the hippiest place in the entire universe had co-written a book about, you know, all of her. spiritual beliefs with chat GPT, who she had developed like a really intense, like, relationship and writing relationship with. And I thought, man, if the hippiest people in the hippiest

place on earth can be sucked into this technology, the feelings that these things are generating are obviously like much more complicated than we're giving it credit for, you know. And I think that's the thing. There's a whole range of human emotions that are being generated by interacting with another kind of mind in language for the first time. I don't know. To me, that's like new parts of human experience are being unlocked. And those seem more interesting to me than just declaring good or evil. Okay. Well, I want each of you to make your pitch. 30 seconds on the clock on why someone should listen to dream machines. I will set a timer right now. And Robin, Let's start with you.

Oh, easy. You know, I think Dream Machines is going to be a unique contribution to this conversation. One, because it's grounded here in this place. And we frankly have access to people here in this city who are, they're making this happen. And they're going to be here in the studio, you know, talking to us directly. And I think we approach this whole world with a mix of curiosity and caution that is pretty unique. And I think it's exciting. This is the podcast I would listen to if I wasn't hosting it.

All right. Alexis, it's your turn. I think people need a more humane podcast to help them orient themselves to something that is both being shoved down their throat. And also that is incredibly powerful and interesting. And that has been sort of obscured by all the money that's like flying around it. So that's us. We're human.

We're friends. Fearlessly independent. Beautiful. Okay. Well, those are all my questions. Any last thoughts on dream machines, the state of the AI industry, you know, the kind of generational panic right now around this industry? From my perspective as someone who spent a lot of time around historians of technology, it is really true that every technological innovation, whether it was things more like the railroad, or things more like semiconductors or social media

have been so deeply shaped by the times in which that technology was developed, the political and social systems around the technology and that we're informing it. And sometimes it's hard to see at the very moment of creation, but it really does matter. And so I think even if people aren't technical and they're never gonna, you know,

open up the hood on one of these like AI systems, they really can have in, influence on how these things develop in these other realms by knowing about them and kind of pressing on them from the outside. And for me, the upshot of that is that this is something that is very productive, maybe even kind of urgent for everybody to be thinking and talking about. I mean, I think everybody who's a citizen of this democracy has a stake in this. And it's not just for the experts. It's not just for the researchers. It's definitely not just for the venture capitalists. So I think part of what we're doing here is inviting people into the conversation

and saying, like, this is yours as well. Yeah. Well, thank you so much for joining us. Yeah. Thanks for having us. That was Alexis Madrigal and Robin Sloan, hosts of the new podcast series, Dream Machines. You can find Dream Machines right here on the Close All Tabs feed every Monday for the next month, starting August 10th. If you'd rather watch than listen, the video podcast will be up on YouTube. Search KQED News or go to YouTube. YouTube.com slash at KQED News. Close all tabs. We'll be back on Wednesday with a brand new episode.