Artificial Unintelligence

Prefer to read? Check out the Audio Transcript

[Audio Clip]

Zachary Karabell (ZK): What could go right? I’m Zachary Karabell, the founder of The Progress Network, and I’m joined as always on this podcast by Emma Varvaloucas, the executive director of The Progress Network. And What Could Go Right?, in addition to being the title of our weekly newsletter, is also our podcast, which is a weekly attempt to change the tone of our collective dialogue from one of dyspeptic dystopian despair to a more positive case for a brighter future. So Emma, who are we going to talk to today? I’m very excited about this conversation. I know who we’re gonna talk to today, but why don’t we share that with everyone else?

Emma Varvaloucas (EV): All right. Let me tell everyone else who we’re going to talk to today, which is Baratunde Thurston. He is many things. Among them, the author of the books How to Be Black, and I’m Judging You: The Do-Better Manual. He worked at The Onion and on The Daily Show with Trevor Noah. Right now, he’s the host of a podcast called How to Citizen, as well as a TV show called America Outdoors with PBS. And he’s the founding editor of Puck. He’s been writing a lot recently about AI, technology, and its role in the future of humanity. So we’re going to delve into that with him today.

ZK: So with that, let’s talk to Baratunde. Baratunde, it is such a pleasure to have this conversation with you. You are one of these kind of eclectic renaissance E, maybe emphasis on the E, type humans who have ranged far and wide in your interests. You’re eclectic, you’re informative, you’re curious, you’re open-minded and passionate, all of the above. You’ve been writing a lot lately about AI and generative AI and ChatGPT and OpenAI, and there has been a fair share of attendant technophobic hysteria, as there always is, when something seemingly dramatic and new comes along in the technological universe. People freaked out about the printing press, about the telegraph, about the telephone, about the television, about the internet, about the cellphone, about the smartphone. So the degree to which we are–

Baratunde Thurston (BT): Don’t forget the horse.

ZK: The horse. Absolutely. Freaked out about the horse too, those four-legged-

BT: Beasts.

ZK: -harbingers of doom. So tell us, in a nutshell, are we right to be dramatically worried? I think you had a line in one of your columns where you’re simultaneously fascinated and terrified. I think I’m getting that right.

BT: Yeah.

ZK: If not, it’s a good way to characterize it regardless.

BT: It is, it is. That’s a fair characterization. It’s good to be here with What Could Go Right? I just love the ethos and title of that. Most of the default settings on the news are what has gone wrong. So thank you for contributing to a different psyche and mindset around the state of the world and our remaining possibilities.

With AI, your categorization is accurate. I’m fascinated and I’m terrified, the nutshell is that. I see a lot of solid reasons to be concerned about what the rapid deployment of these tools means for everything from copyright and ownership of our creativity to the very nature of the human experience as one that is just so increasingly mediated by a technology creating distance between us and the physical world, between us and our fellow humans, and even between us and ourselves. And I think just our experience with social media the past decade or so and fragmentation and balkanization, we just– in so many ways, we’re not ready. Also, excited, so excited about creative possibilities, about having another tool and sort of colleague and co-pilot in the quest for justice and the ability to try to live together better and to find answers to some really deep questions, whether they’re scientific or existential. So it’s all possible, but the concern is very real. I’m not gonna try to sugarcoat it.

EV: I wanted to talk a little bit about the RNC AI ad. Did you see that? That was all AI image generated.

BT: I did not see it.

EV: Basically, it was like, imagine Joe Biden becomes president again. And then they-

BT: Okay.

EV: -ran through all of these nightmare scenarios of what that might be like. So the first one was, yeah, war with Taiwan. The second one was 50 regional banks collapse and they flash an image from like the bordered up windows of VLM. Obviously, it was AI-created, but that was the image that you got in your head.

BT: Yeah.

EV: And then they did some really heavy drug users in San Francisco and a couple of other dystopian things and the response to it was really intense, right? There’s this response of like the RNC is now– has this new tool in their power and people are gonna be confused and they’re gonna blah, blah, blah. And I was interested to ask you about it because my reaction was like, that was a bad ad. It wasn’t–

BT: Yeah.

EV: It actually– it didn’t feel– of course, I’m not the intended audience, but it just– it honestly did not pull me in at all.

BT: This is speculative. Happy to do it, just gotta name it. Most political ads are garbage.

EV: [laughs] Yeah.

BT: I think we can all agree on that. Regardless of where you sit on any political spectrum or ideal, they’re just bad. They’re the worst Hollywood movie trailers. In a world where– And it sounds like they just unleashed the in a world guy on their projection of the worst possible future with Joe Biden as president. That’s typical. It’s uninspiring, it’s fear-based, it’s unimaginative, which I think makes it appropriate to be built on generative AI systems because it can only mine what we already know in so many ways. And so whatever they trained that data on, whatever they trained that ad on, is just the history of shitty political ads. Cool.

I think what would be more interesting, years ago, the artist Molly Crabapple teamed up with a few political voices, climate people, Alexandria Ocasio-Cortez did the voiceover for this animated, like the inverse version, what you just described. And it was imagining the future where a Green New Deal was fully adopted and implemented and executed on. And it was inspiring. It was beautiful. It was a bit like, oh, that would be magical, but it appealed to a different piece of our human mind and our human spirit and a human artist helped bring it to life. And I’m not saying machine art can’t do that, but it just feels really predictable, also ironic. That’s what they chose to run with.

Audio Clip: Let’s start with how we got here. 1977. New York. A senior scientist named James Black made a presentation about how burning fossil fuels could eventually lead to global temperatures rising 4 or 5 degrees.

ZK: So look, the fears around this are, this is the version 1.0 of the singularity or– people love to use the Terminator, Skynet is sentient, right? That this is the first shot across the bow and that’s where we’re heading. We’re heading-

BT: Yeah.

ZK: -to AI that is indistinguishable, that can pass the Turing test. These are all the tropes of artificial-

BT: Yeah.

ZK: -intelligence land. I think I’ve used at least three of my quotient of cliches in the past one minute.

BT: Congratulations.

ZK: But-

BT: Yes.

ZK: Maybe I am indeed myself [laughs] AI-generated. AI-generated cliches for 500, Alex. So a lot of what AI is doing right now, like I’m as susceptible occasionally, in moments of procrastination and boredom, to click on the listicles. The 10 worst movie scenes or the 10 scenes that ruin their careers.

BT: Yeah.

ZK: And I was just in Pakistan a few weeks ago and sent me this Lahore tour and I was like, wow, this is cool. Where’d you get this? It was ChatGPT. We just-

BT: [Laughs].

ZK: -said “things to do in Lahore. I was like, that was pretty good. It’s a lot to do in three days…

BT: We can all look impressive. Yes.

ZK: I don’t know whether that’s a problem. But you mentioned the horse before, right? The fact that humans-

BT: Yeah.

ZK: -were– human labor was substituted for horse labor or that horses were then made less relevant by cars. Yeah, there was disruption and there was a certain amount of creative destruction if you believe in the Schumpetarian view of reality. But I don’t know that was ultimately bad destruction of what we used to do.

BT: Yeah, when cars came along, they mostly displaced horses. And automation in general is never a one-to-one substitute. It doesn’t just like every human job becomes a robot job or a software job and it’s bad. It’s generally, yeah, people who had certain types of occupations can’t do that part of the occupation anymore. And the new technological capability creates new jobs, right? The telephone operators transitioned in many ways to a different type of role. They weren’t physically connecting inbound call to destination, but they became information services operators, they became receptionists, secretaries. There was still a need for a human role in some cases.

What often, I think, gets missed though is that the new capability just opens the pathway for many more people to engage in the thing, right? So writing used to be a high tech endeavor that only monks did. You had to be trained. There were like 50 people in the world who could write. They lived in a castle. It was very Game of Thrones, right? They were meisters. And when writing became much more available to the public, you still have monks, you still had people in an academy, you still had meisters and castles, but also, anybody could write any bullshit, right? I could write bullshit. I could write a postcard. And so what’s gonna– like the people who can generate tour guides of Lahore is greater than professional travel agents, which has been the case for a while. It’s greater than the Wikihow crap writers. It’s greater than the people who’ve already blogged a billion times about it online. It’s like your friend who’s never written a travel guide a day in their life can generate one for you and that’s gonna create a flood of new content and new information. Not entirely bad.

I think the worry side of it for me is we will create– we’ve already created so much stuff. I can’t watch the TV that’s being made. We will never finish Netflix. It’s an endless scroll. And so what’s the point? And if we need machines to help us overproduce, we’re also gonna need machines to help us overconsume. There’s just too much. So we’re gonna watch the shows on 1.5 speed and we’re gonna have a generative bot summarize and give us the abstract, the CliffsNotes, the cheat codes on all these experiences. So we’ll rely on machines to create, we’ll rely on machines to consume and experience, and our role in that loop is like a phone operator, right? We’re just connecting an input to an output but not experiencing the call ourselves. That’s an existential worry I have.

And the other is economically. So much of this is just driven by capitalism. It’s like increase throughput, increase output, create more efficiency. Why? The great lie of all these tools is that they give us our time back. They don’t. The Blackberry took more time than it gave. Email took more time than it gave. Slack, which is supposed to solve email, takes more time than it gives. We just are on more. We have to be more responsive. And so we have this colonization of our time and of ourselves in service of ultimately a venture-backed profit-driven escapade in which we don’t really have an interest ourselves. Some benefit, very little interest. So the concentration of wealth to create more wealth, that’s not very inspiring to me. That’s very concerning and it’s like, all right, so we just become better constant workers connecting machine output to machine input. Eh, there’s gotta be a better use for all this stuff.

ZK: Venture-backed profit driven escapade. I like that. I’m not sure I agree, but I like it as a phrase. That’s good.

BT: Yeah, that’s cool. And I don’t use the word escapade very much, so I should get a point for that.

EV: I guess what that premise takes to be as necessarily true is that we are all responsible for consuming all the content that’s already produced. But I think that we’re not, right? Like the idea is that there’s the diversity, the plurality of content out there and you can-

BT: Yeah.

EV: -pick and choose what’s useful to you. I certainly don’t feel like there’s a pressure, something incumbent upon me to watch all the TV shows that have been produced or all the podcasts or read all the articles. The fact that the things that I wanna learn about are there or see are there seems to be a benefit-

BT: Yeah.

EV: -to me. Does that ring true to you at all?

BT: Yes and no. I think the pressure that I feel comes from the desire to feel like I’m a part of a common experience. There’s something shared in these human experiences, and in a simpler time, we had more common reference points. Just in a different time, not necessarily simpler. In some ways, I think it was, but certainly we could agree it’s different. And so that common reference point might have been a religious one, it might have been an informational one, it might have been a labor common experience.

For generations, parents and children had very similar lives, and whatever advice parents could pass on was based on a set of experiences their child was likely to have. Work too hard in the field. Your teeth are gonna fall out at 30. You’re gonna die at 43. Good luck, kid. That was a lot of the human experience. And now, within a generation, they’re radically different experiences, less and less common between an older sibling and a younger sibling. And so what we have to share between each other and pass on, it just gets more difficult. I think the flood of TV shows is a simple example of this larger trend. And so what show am I watching? What show are you watching? What do we talk about at the water cooler, which no one has a water cooler anymore. But the idea of that ancient like sitting around the fire, common human storytelling, I think there’s something deep within us that still demands and needs that to not feel too untethered. And so we tether to each other, we tether to the fire, we tether to a tale. And if everybody can generate their own tale constantly, then how do we also create a shared story? See it’s a TV show. Maybe it’s a new religion. Maybe it’s a set of principles and values. But I think everybody creating a custom lens on reality will have some negative consequences ’cause we won’t see the same things.

ZK: So I wanna circle back and maybe one thing I’d want to throw out there, maybe we’ll come back to this, is I’m not entirely sure that any of us entirely know what capitalism is or means. I mean, we use the word, or any of these isms. They’re very loose terms for multi-level, multi-faceted systems. But this idea of a shared reality, we’ve gone over the course of a couple hundred years from a billion people on the planet, barely, in 1800 to 8 billion plus people right now. And it’s not just there were a billion people 200 and some odd years ago. They weren’t really connected to each other. So it was more like several hundred group of tens of millions at most.

BT: Yeah, little tribes mostly. I don’t think it’s possible for 8 billion people to have the same story. And I don’t think it’s desirable to throw everyone into a massive group chat. Hello, Twitter, right? These things will naturally fragment and subdivide themselves and that’s okay. A family has subcultures, much less a species.

The thing that I want, I’m not an advocate of pull the plug on the machines. I think that’s really simple. When I say capitalism, I’m talking about a desire to produce outcomes that are optimized for a particular measurement of success, which is profit, at the expense of many other measures that matter to us. Quality of life, happiness, sustainability of the home planet, a future that we don’t fear in terms of habitability, like very basic human needs stuff that capitalism has helped to serve but now threatens at the extreme to completely undermine. There is no market for any good or service with no earth, at least not a growing one. It’s literally a shrinking market.

So these tools and the race to project them into the world are driven by a particular perspective on maximizing public market capital defined as returns on investment, not defined as quality of air, or depth of love in relationship. There are other things we could try to optimize for. We haven’t really tried. And when we talk about shared narrative and the risks, yeah, there’s gonna be like a Breitbart bot, there’s gonna be a alt-right bot, there’s gonna be like a BLM bot, a Bureau of Land Management, I’m referring to, as well as a Black Lives Matter bot. And that’s gonna create some chaos. There’s subreddits too for all these different groups and it’s created some chaos.

It’s not necessarily the end of the world, but the combination of the speed with which this stuff gets deployed, the narrow, I think, drive behind it, these are not so far community-defined efforts. If we said, all right listen, these large language models are generated basically off of the wisdom and data, at least, maybe wisdom, and certainly the data generated by the world. We’ve scraped Wikipedia, which is contributions from every– we sucked in a whole bunch of photographs taken by millions of people. But the interest, the literal financial interest in it is just a couple of shareholders. It’s like one of the largest tech companies in the world. That seems like a huge disconnect. And I would feel less alarmed if we had collectively some deeper stake in these systems as well as some more kind of democratic, small D, method of determining how we are going to manage what goes in and how we decide to create what comes out.

Instead, we’re just jammed with it, right? Hey, there’s GPT-2, nope, GPT-3.5, nope, GPT-4. Hey, now we can make videos, like, wow, this is so fast. And I just feel subject to the whims of an almost literal handful of people. And I don’t like that for 8 billion people. In a place where we’re supposed to be democratizing tools and all having a voice, it seems like a lot of this stuff is being imposed. And so I just have a– I don’t have a deep answer to it, but I have a deep concern and I don’t think that is how I want the future defined for so many of us. So I try to weave in a couple of the things we’ve been talking about here and see if I can get more specific about what might bother me.

The last– it’s not last, but I think it’s my other deep on the concerns side. We can get to the exciting. I’m excited about stuff too. But on the concern side is one of the things I wrote about in Puck, I had this clear memory of GPS, and I used to manually make trip plans with my mom. She ordered things from AAA and they helped you route a path on paper, they ship it to your house, these little vertical maps called TripTiks. Then I ran the GPS off a laptop in the car with CD-ROMs. Then we got MapQuest, then Google Maps and voiceover GPS. And that all made navigation simpler. It made it smoother, it made it faster, it made it more optimal.

And I never know where I am. Like I don’t know anything anymore. I will defer to the screen, which claims to be all seeing and all knowing versus my eyes. And if my eyes perceive that looks like a traffic jam but the screen says it isn’t, I’ll just wait in traffic. If my eyes say there’s no bridge ahead of me but the screen says you can drive, I’ll drive off the bridge. I haven’t done this but other people have. And I think there is a deeply troubling metaphor for our deference to these systems which are sold and marketed as intelligent, as human-like, and as better than us. So then we don’t have agency. We just– we follow. We all become followers of a system that we didn’t choose to create and we deny our knowledge of ourselves.

I have this smart ring, I wake up in the morning and my wife says, “How’d you sleep?” And I said, “I slept pretty good.” Then I check the app and the ring says I didn’t sleep pretty good. I’m like, correction, apparently, I had a terrible night’s sleep. I need to try harder to sleep tonight ’cause the ring told me I don’t feel the way I feel. And so I find that deeply troubling, man.

And spending all this time with intelligence of an artificial nature creates gaps with a different form of nature that I think we need to try to maintain. Otherwise we just follow these generated instructions for what to do with every moment of our lives. How to tour a city, what to cook for dinner, where to turn in our vehicles, who to mate with. It’s very optimal. It’s very optimized. It’s very efficient. It’s also dead. If life is just following instructions from a computer, we get to live longer but we have less life, I don’t know, I don’t know. That’s kind of weird.

EV: The one thing I would say to that is that I think that we’re a little bit more in control about our choices in how much tech is in our lives than what you just described. For instance, if you realize that you are sleeping just fine and the smart ring is telling you otherwise– so you probably get to a point where, like– with the GPS example, one time I was in Bulgaria with my sister. We were driving. Google Maps told us to go straight. We looked at this freshly painted sign from the Bulgarian government that told us to turn left. And we were like, no, like the internet gods would know all. We will go straight. And of course, the straight road led us into the woods, in the mountains, gravel road. We popped our tire. We had to get tow trucked. Left was the way that we should have gone. So that seems to prove your point, but the other side–

BT: Trust the government [laughs].

EV: Yeah, trust the freshly painted sign-

BT: Wait, what?

EV: -from the Bulgarian government.

BT: Yes.

EV: But the other side of that is, now that I’ve had that experience, I’m like, Emma, if your gut reaction is telling you something, go with that instead of the internet gods. So there is an adjustment process here and a little bit more of a choice aspect to it.

BT: And here’s the good– I think here’s the opportunity and the invitation in this moment for us, as we have a much more artificial existence, we will value natural existence even more. I get to spend a lot of time in the outdoors, in the woods, on rivers and mountains. I make this beautiful show on public television in the US called America Outdoors. And I just– I love that. I was fly fishing with foster kids two weeks ago, standing in a stream just popping trout out. They trained me well. And it is the opposite feeling of being efficient and making sure that every physical step I take in the world is maximally optimized to minimize discomfort and pain. It was pretty uncomfortable in that stream at times. It was cold, but it was so real and so grounded.

And my experience with the ring, your experience with the Bulgarian freshly painted sign being way smarter than the smart tech of Google, which is one of the smartest entities ever created, like collective wisdom inside the Google machine is extraordinarily intelligent, more so than any single human being. But that sign was right and Google was wrong.

So yeah, you’re absolutely right. You’ve got an extra caution and there’s gonna be some tickle that we all feel where we get to choose, okay, what instruction, what feeling, what intuition do I wanna listen to? And we gotta practice that though. I think for me, most of the time I’m gonna follow the GPS. Otherwise, why do I have it, right? [laughs] If I question everything the machine tells me to do, then that machine’s not doing its job. So on net, like overall, we’re gonna tend toward following. That’s the whole premise of creating these systems. But if we can decide, maybe not now, maybe not today, maybe I need to break– put some more agency back into this, maybe we need to redesign this thing, maybe it needs to ask us questions, I don’t know, there’s just gotta be something more. And I think it’s gonna force many of us to ask why, right?

If we could eliminate so much work and so much drudgery and so much labor, again, always promised, never delivered, but maybe this time is different, then what would we do with all that time? Would I let my report write itself so I could take a walk with my wife and just connect more? Would I let those emails send themselves so I could call my sister instead? That’s a great trade off. Let the email write itself. Let it respond to the email that someone else’s robot sent. Let the robots do all the emailing and just give me a weekly report on what I said this week so that I could meditate more and do some psychedelics and try to achieve some higher plane of existence. That’s glorious. That’s really great. Most of the employers of the world, they won’t see it that way. And they’ll be like, oh great, you don’t have to focus on writing emails, you can do X, Y, Z thing instead. You can proofread the robot’s emails [laughs].

ZK: Okay, so on the flip side of all this, and I do love the– and it’s totally right about the GPS. One of the effects of our technologies in this case is no one’s ever lost and no one ever knows where they are. And that is true for a plethora of our daily realities. Again, part of the ethos of doing What Could Go Right? is we do have a very human tendency to look at problems and both evolutionarily and culturally, that’s not always a bad tendency, right? We are constantly on-

BT: Trying to survive.

ZK: -alert for that which will, if not now, soon, be a threat and be a danger. The flip side is, we always lose sight of the past, right? Because the past isn’t particularly real and the future isn’t either, but the future we can invest with hopes and dreams, and the past, we can beam back with unrequited desires. And like a lot of human history, labor to just produce food, if you looked at the amount of energy and human hours to produce a loaf of bread 200 years ago versus now, right? So the amount of human labor to do these things has in fact freed up all this other labor to do email, right? Meaning, you’re-

BT: Woohoo.

ZK: -entirely right that one type of work pushes out another type of work.

BT: And to watch Netflix.

ZK: Right. So both are true. Our leisure time-

BT: We are more aimed than ever.

ZK: -has gone up and our work time has become more elastic. But I guess I would push back and offer the following. I’m not saying I’m right about this, I’m just– something I think about, which is if you really did in your sort of small D collective global democracy way try to get buy-in from a planetary population about what these technologies are doing or what the arc of modernity and capitalism has done, I think net you would probably find more buy-in than not to these things have allowed for far less physical labor, longer lives, and the ease of procuring material necessities and material desires. So while a lot of the benefits have unfairly or inequitably– I don’t know whether it’s unfair or not, have inequitably flowed to the few and not to the many, I’m not sure that I agree that the system itself would be radically different if it weren’t quite so foisted upon you but was in fact voted upon.

BT: When I talk about small D democracy, I’m not talking about up down votes. And I think, just as you rightly pointed out, like what do we mean when we say capitalism? We’re not always clear about that. What do we mean when we say democracy? A lot of our experience with democracy is outsourcing, right? It’s delegation to a chosen few to oversimply make decisions on our behalf. And there’s all kinds of corruptions of their motives and perversions of interest with money and gerrymandering and a lot of things that have diluted the ability of our representative democracies to actually reflect the will of the people. Again, not a partisan statement. I’m a very partisan person, but that is academically and mathematically demonstrable.

So these citizen assemblies and deliberative democracy is like one form in practice where random assortments of people are selected almost like juries to develop policies, to weigh in on budgets, to advise or determine political decision making on behalf of a community, could be a city, could be a nation. I don’t think an 8 billion person citizen assembly makes any sense. But smaller collectives can roll up to something that better reflects what people actually want. We have a failure of democracy around abortion access and reproductive rights in the US, around gun, sensical regulation of firearms safety and access in this country. Vast overwhelming numbers of people want a certain thing. Our “democracy”, in quotes, has not delivered it. It’s failed in those particular points. And there are a lot of other examples. If we could have a better process, we could have a better outcome that better reflects the nuance, the ranges, the exceptions that most people actually feel.

And so when it comes to something like AI, we can, through a process, bring in the hopes, the dreams, the concerns, the fears of many more people than a couple of machine language researchers and a couple of VC firms who think that what we really need to optimize is illustrations [laughs], right? Maybe or it’s like asking workers what they would want to use these tools for rather than just imposing it. And that there’s basic low-tech ways this happens already through surveys, but that’s when I refer to more small D democracy in the process. Part of what I’m hinting at is a different way of surveying and polling, a different way of gathering and soliciting input, and demonstrating alignment of desires or non desires in a way that our current political system and sometimes our business system are unable or have been increasingly unable to deliver on.

On the What Could Go Right? What could go right with all this AI stuff is we are buried– we’re overwhelmed with confusion around our financial services, around our health insurance situations. My friend Ron J. Williams has this view of radical comprehensibility and that is aided significantly. I unleashed ChatGPT on my 180-page health insurance coverage policy document. It was never meant for me to understand that document. These businesses, so many of them, they make money by making sure I don’t understand how to get what they’re actually offering me and they hope I won’t figure it out. It’s a really shitty way to operate, but it’s also– can be really profitable. If we make people jump through 30 hoops and we know 50% of people only jump through two, overage, we’re good. And you could pick a business. There’s a lot of them that operate this way.

These tools give me an army of robocallers, of legal scholars, of sentence diagramers and researchers to throw against that wall without investing all the time it would take to become an expert myself or just wait on hold that long to argue with an agent. I love that. I love that adversarial relationship. I love evening the playing field. And then I can imagine that for coordinating action, for climate, for making policing much more sane and actually a public safety interest and not just an enforcement of property rights interest or refuge of racism and slave catching in the US as we’ve practiced it, that we could unleash these tools in the arms of civilian oversight boards, in the arms of judiciaries, in the arms of city counselors and activists to say, all right, let’s analyze this in a much faster way.

When the justice department has to come into a city and they get their right to review– I forgot the specific language of what they do, but they basically take years combing through and observing and then they write a little report and– consent decrees. We don’t have to do that one at a time. We could just say all police agencies, all meat inspection plants. Where there’s imbalances now and the folks who kind of abuse the advantage win because they just have more time, nope, we can balance that out. So I think that’s a great possibility, and I wanna see that. I wanna see it for climate stuff in particular ’cause we have to coordinate such complicated actions.

Last example, I’m actually sitting on this, right? I’m holding up this item’s called FutureCard. It’s a visa debit card, gives you 5%, 6% cashback for your lower carbon purchases. You buy an electronic item through Back Market, the secondhand marketplace, instead of new, you get more points for that. I’m a small investor in this, disclosure. They built something on GPT, chat.green. It’s GreenGPT, and it lets you figure out all the subsidies for solar, for wind, for graywater systems that you’re eligible for and a lot more. It’s basically a a green economy expert at your fingertips. And you can throw all kinds of stuff at it from your life and it answers all these questions that would take me weeks to figure out. The Riverside County rules on this and the LA County rules on that and the city of L– oh my goodness. And that’s not designed to make it hard. It just is hard.

So to simplify, to make more accessible. It’s incredible. The effect on literacy, if you’re not literate, these tools can help you behave as if you were and get access to the information and the power that comes through literacy without having to go through all the hoops that you might not have time to do that. A friend of mine is not so great at English. He’s using it to improve his email communications [laughs]. So he shows up more professionally, people understand him better, his dealings are smoother. Scale that, scale that. So, yeah, it will be truly unlocking potential, freeing us up, enabling folks who’ve been disabled by their circumstance, by oppression, by whatever. I’m here for that.

ZK: That’s a really good note to wrap up a conversation that’s in the middle of a conversation so hopefully we can keep having the conversation and clearly, like a lot of things that we talk about, none of these are terminal conversations, right? These are not like election analyses. These are like long-term, who are we, what are we, how are we gonna shape the physical and spiritual and technological environment that we are all swimming in. And I think one thing that I do like about your writing and your speaking is you are curious. And for me, curiosity is the openness to the new and the openness to learn about the new and to not allow yourself to constantly just reimpose an a priori framework on that which is unfamiliar and new. Otherwise, all you’re doing is just stamping reality with a very rigid template.

BT: These calls to pause AI, for example, for six months, an arbitrary time unit and an unclear meaning around pause. My instinct is instead to engage more, right? So the way we resolve this I don’t think is just unplug the machine and then figure it all out and then plug the machine back in. And I think what Sam Altman and OpenAI has said, which I [inaudible] kind of shocked but not entirely to agree with is we had to release this in beta because we couldn’t figure this all out in the lab. And so there is some level of collaboration, like the concerns I’m raising, the excitement you’re feeling, and everything in between, that is a somewhat democratic process in the sense that we’re participating. I want the terms of that participation made much more clear and explicit and blah, blah, blah, blah, blah. But I wouldn’t be able to offer these critiques or these hopes if I hadn’t been able to play with the thing myself.

ZK: Right.

BT: And that is a really important part. If I could offer one last, I think, hopeful possibility here, what has jumped out the most from me, from pla– I’ve played extensively with GPT-4, with Midjourney 5, those are my primary experiences. And then a little bit of production tools around video editing, audio editing for my own workflows. But those first two examples, I can ask it something and it can give me an okay response. My prompt is meh. If I focus on the prompt, really make it the most specific, the most nuanced, kind of help speak it in a language I know it is prime to understand better than just my own natural way of speaking, I get much higher quality responses.

And I’ve done these experiments with like average prompt versus super prompt. And it just puts the focus on how you ask. And it gives us this superpower, this performance enhancing drug, this thing that I think of increasingly as magic where it’s, oh, we’re casting spells. You can cast a spell and turn your friend into a frog. Oops, sorry, I invoked the wrong spell. Or you can turn water into wine, which is a wonderful spell unless you really just need hydration, in which case, that’s not-

EV: [Laughs]

BT: -the spell you want at the moment. But that’s a point, like we need– this is gonna force us/ invite us to be much clearer about what we actually want. And I think there’s a micro little metaphor in that, the clarity of the question. Like what are you gonna use this power for? Literally how are you gonna ask? And that will determine what we get out of this in your next ChatGPT session or over the next decade as we ask ourselves what are we gonna do with all this power?

EV: I think that’s a perfect way to end. I’m glad you said the bit about the collaboration at the end as well because I was gonna remark upon that too, that now when you open ChatGPT, there’s at least that little warning thing at the bottom that’s like ChatGPT may make up facts, like [laughs] please be aware.

BT: Oops.

EV: Yeah. But again, it’s what you said, this comes out of a dialogue with the public and whether or not we’re satisfied with the amount or scale. Dialogue is another thing. But there is some, so. Anyway, I don’t wanna take too much away from your final statements ’cause I think they were really poetic and we’ll see what the magic of AI does for us or does to us. So thank you so much for coming on.

BT: And we’ll determine it. We won’t just see.

EV: Yes.

BT: We’ll create it ourselves.

ZK: So I love that, love that, love that. I guess, again, there are simply people who articulate their concerns and hopes about the world in a way that I think is more likely to lead to a constructive future. And there are those who do so in a way that seems like it’s gonna lead to a more conflicted, less constructive future. But Baratunde clearly embodies that both hard-edged but open-minded approach to what do we do about the world? And we didn’t even get into a lot of the work that he’s done over the past years, which is more about polarization and race and inequality, and even then when dealing with these issues that are really fraught, he looks for and tries to create connection.

EV: And he’s funny, right? That’s the other big charm and talent that not everybody has. He was part of The Onion. He was producer on The Daily Show with Trevor Noah. And there’s certainly something to be said about the power of humor.

ZK: So hopefully we will continue to have more conversations with him and with others like that. But, again, to me that kind of epitomizes what we’re trying to do and what we’re trying to amplify. Let’s talk about the news, shall we?

EV: First news story that we have for today is busting the myth of the broke millennial. And this is an article in The Atlantic by Jean Twenge. She has a new book coming out, so it’s actually an excerpt from her new book. But the excerpt was fascinating, right? Because I’m a millennial. I can say personally that that is a feeling among my friends, that we got dealt the short end of the stick coming of working age during the financial crisis, then we had the pandemic, and there were very much so this feeling of we’re poor, we’re broke, we’re saddled by student debt, we can’t buy a house, houses are really expensive anyway, so screw it, this life ain’t for us. But the article is really about how it is true that millennials were behind when they first started working, but now they have caught up.

Jean Twenge calls it a breathtaking financial comeback that started in the late 2010s. And so she does all this comparison of data of median household incomes compared to Gen X and boomers at the same age. Basically, it’s actually greater than Gen X and boomers at the same age also adjusted for inflation. Millennials have– one in three of us have a college degree by our late 20s. It’s the first generation ever to have that kind of number, which means that our income is up. And there’s fewer millennials in poverty than Gen X and boomers at the same age. And home ownership rates are almost the same. It’s like 48% to around 50%, all of which is to say this reputation as sad and broke is no longer true. And it’s important that we rectify that.

ZK: And look, I think some of the challenges of why that image remains so profound is the media people or people who are creating content who tend to be overrepresented in places like New York and coastal cities like San Francisco, LA, those areas are actually pretty unaffordable relative to even pretty high incomes. And so there’s an understandable tendency to magnify that as a global reality. You yourself had the experience of having a pretty decent job in New York City, but that being inadequate to a lifestyle in New York City, and that is a real problem of New York. So it’s not like there aren’t pockets of places where cost of living and income are just not favorably aligned under any circumstances, but that doesn’t make it actually nationally true.

EV: Right. Or even just that it was true that millennials were having a hard time, but it’s no longer true and regardless of where you live, right? And it’s really important if we– if that were true, right? It does call for what we’re seeing, which is should we really be trusting capitalism with our economic fortune, should we be looking to other systems of government? It starts those trains of thought.

ZK: Right.

EV: But if it’s not true, if the American dream is still alive for millennials and will continue down through Gen Z and Gen Alpha hopefully, that means that we need to take another look at things and think that, ah, actually, maybe the system is working all right.

ZK: Exactly. And recognize, look, if you’re an American, it’s been a weird 20 plus years since the collapse of the Nasdaq bubble in March of 2000 through 9/11 through everything else that has happened subsequently. It’s been a really challenging cultural time, irrespective of its also strong periods of being challenging economically.

EV: Yeah. What we have next is Gen Z blowing past other generations, that’s the headline from Yahoo, when it comes to 401(k)s and retirement savings. So this actually might come from the narrative of the broke millennial. I think Gen Z took a look at the world around them, they’re like, ah, let me save a bunch of my money. 62% of 18 to 24 year olds contribute to 401(k)s in 2021 compared to only 30% in 2006. And actually, despite what I just said about Gen Z being scared, that’s the personal anecdotes that I hear from people that are that age that they got freaked out by what happened to millennials. But Yahoo actually says that so many more Gen Zs are contributing for something that’s really a small boring change. It’s just automatic enrollment in 401(k) plans went way up. In 2006, only 11% of employers offered automatic enrollment. By the end of 2021, it’s about 50%. So simple fixes.

ZK: That is kind of wild. And especially when you consider that every single statistical model of investing and saving shows that the earlier you start– the delta between starting early and starting even in your early 30s versus– sorry, let me do that again. The delta between starting in your early 20s versus early 30s and what you will then have by the time you’re in your 60s is massive. So it’s not just a, isn’t that good that there’s a little more preemptive savings. It’s that the change that will make across a lifetime, if past trends hold, which obviously they could not, really sets that cohort up particularly well in the future. So another piece of news, and that one I had no idea about. That’s really cool.

EV: So let’s talk a little bit about the right to repair. It’s a topic we I don’t think have covered on the podcast at all.

Audio Clip: We live in a free market. But when it comes to repairing electronics like smartphones, you are not free to choose where to go. If you were the hopeless person with a broken gadget, you’d immediately go to the Apple Store, and that’s exactly what Apple wants you to do. The company and many others restricts how and where you can repair your stuff.

Anything that has a chip in it right now is probably impossible to repair without using the manufacturer.

That means tractors and cars. It means your smartphone, it means increasingly the refrigerators and washing machines that people have in their homes.

When something breaks and the only solution is to take it back to the manufacturer, they can charge you whatever they want.

So this is a MacBook Pro. The Apple Store said it would cost $1,200 to fix and wasn’t worth doing. So if I walked in off the street with this problem, what would you charge for the repair you just did?

Depending on the model, anywhere from 75 to 150.

EV: June 22, 2022 was the United States’s first ever right to repair law. That was in Colorado and it was actually for wheelchairs. And there’s been a spate of laws since then, but the one that I wanted to highlight today was one from this month of April, again from Colorado, and it’s the right to repair for farmers. And I wanted to highlight it because this is one of those things where you learn, you’re like, that’s crazy. Like I did not know that farmers did not have the right to repair their own equipment. So now they will, at least in Colorado. Generally, retailers like John Deere have made farmers come to their own stores to get tractors repaired and other farm machinery. Now they will be legally obliged to provide farmers with diagnostic tools, software documents, and repair manuals starting January 1st. And similar resources must be made available to independent technicians. So seems like a win for farming.

ZK: Absolutely. Yet another one of there should be some balance between personal autonomy and corporate bottom line.

EV: Yeah.

ZK: So that’s some good news of the week, Emma. Thank you.

EV: That is. I do also have to give a correction on news that we covered previously. I think it was a couple of weeks ago on the fentanyl test strips. And at the time, I said that only two states allowed the sale of fentanyl test strips. That’s because I completely misread a Cato Institute report. That is incorrect. 36 states now allow the sale of fentanyl test strips which allow people to test to see if the drug they’re about to take has fentanyl in it. That’s an addition of 16 since January 2022. So the movement is definitely new. And at the time, I said there’s some movement here, not that much. I’m thankful to be wrong in a good direction.

ZK: You underplayed your already good story.

EV: Yes, I did. I did. And the remaining states where it’s not legal, there’s pending legislation. So we’re actually probably going to be seeing soon that it’s legal across the US.

ZK: Cool. Good to know.

EV: Yes.

ZK: Thank you all for joining us for this What Could Go Right? and our conversation with Baratunde. This is weekly. Please sign up for the newsletter, also called What Could Go Right? Go to theprogressnetwork.org and you can sign up and it’s free. So thanks, Emma, and we’ll talk again next week.

EV: Thanks, Zachary. What Could Go Right? is produced by Andrew Steven. Executive produced by Jeff Umbro and The Podglomerate. To find out more about What Could Go Right?, The Progress Network, or to join the What Could Go Right? newsletter, visit theprogressnetwork.org. Thanks for listening.