Frank McCourt: Founder of Project Liberty (Part II)
RadicalxChange(s) | 2024-07-28 | 1:15:55
In this episode, Project Liberty Founder Frank McCourt joins Matt for a second round to discuss the challenges and opportunities presented by rapidly developing AI technologies. Building on their previous chat about digital infrastructure, they explore whether AI will exacerbate social media, digital advertising, and data centralization issues, or fundamentally change them. McCourt emphasizes fixing the internet’s design flaws to ensure AI benefits society, advocates for returning data ownership to individuals and stresses the need for political engagement to align AI with democratic values. Tune in for this enlightening conversation and what we can do moving forward.
Top Keywords
- data 0.009
- technology 0.009
- models 0.006
- internet 0.006
- power 0.005
- values 0.005
- information 0.004
- powerful 0.004
- infrastructure 0.004
- forth 0.004
- liberty 0.004
- principles 0.004
Transcript
Speaker 0
0:01 – 1:08
Hello, and welcome to Radical Exchanges. Today's episode is part two of Matt's discussion with Frank McCourt. In our last episode, Frank argued that the Internet has become a dehumanizing system of surveillance, and he called for new protocols similar to what we have seen in past struggles against oppression in order to realign digital infrastructure with human rights principles. In this episode, he expands on that, turning his eye to the societal implications of powerful new AI technologies like large language models. Together, Matt and Frank explore where underlying principles such as freedom, equality, and citizen empowerment fit into the development of advanced AI systems, and they grapple with the contrasting possibilities of extractive or manipulative AI systems versus an alternative approach that enhances human agency and citizenship. Frank McCourt is the founder of Project Liberty, an organization dedicated to developing technology aligned with principles of individual liberty and democracy. And now here is Matt Pruitt and Frank McCourt.
Speaker 1
1:10 – 1:14
Frank McCourt, thank you so much for joining for a, for a second conversation.
Speaker 2
1:15 – 1:17
It's nice to be with you again, Matt. How's it going?
Speaker 1
1:18 – 3:05
Doing pretty well. So by way of orientation, in our in our last conversation, we talked a lot about, about infrastructure, about the way that, that the prior generation of digital infrastructure has either failed or succeeded and in some ways perhaps deviated from the pattern of older infrastructure building like roads. And I think that, we we both wanted to talk a little bit further to extend that, conversation, into the question of the day, which is, AI. And, you know, how how will the, how will these new quickly developing AI technologies, change the kinds of problems that, we have been worried about with regards to social media and digital advertising and the way that the digital infrastructure that we have had around us for about ten years now is affecting the public. So there's obviously a lot of, a lot of very specific questions that we can get into with regarding, with regards to AI, but I'd love to start with just your general thoughts. Like does, is AI likely to be sort of a continuation or an exacerbation of the kinds of issues that have worried you with, about social media, or, does it change the nature of the problem? What are your feelings about it?
Speaker 2
3:06 – 12:48
Yeah. I, I think this is a really important issue for us all to to grapple with here because generative AI has become this kind of shiny object, which is attracting a ton of attention and, as it should, by the way, but not in a way that it is, you know, distracting us from the the basic design of how tech is working right now. Right? We talked about that in our last session that, not just social media platforms, but all the large platforms are basically scraping data from individuals, aggregating it, applying algorithms, which are predictive tools, and, and, you know, making these algorithms make judgments about us, right, and and our behaviors and our our our our tendencies and what we're gonna do with a piece of information or how we're gonna react to a piece of information or how we're gonna get triggered by one thing or the other. And and, it might be as benign as buying something, but it might be a lot more important than just buying a new, you know, a new device or a new, piece of clothing or something like that. So this entire tech, you know, the Internet, again, let's focus on it, you know, evolved from a decentralized Internet to a highly centralized one when the you know, at the dawn of the app age. And we've had now, you know, fifteen to twenty years of of that, centralization of the power of the data and of the power and, and and influence that that creates into these, you know, five large platforms that, basically are the, if not the only ones, certainly the primary ones to have access to both massive datasets, which these large language models, are trained on, and huge compute power. And remember, it's the it's the massive data and the the math and and the massive compute power that are necessary for these large language models to operate. Right? Because they too are predictive, the these large language models are are are are predictive technology. Right? The the lots of information is crunched, and the the the next most likely word is spit out, and then the next most likely phrase and then sentence and paragraph, etcetera, etcetera. So, again, predictive models based on a ton of information and the ability to crunch all that. And, and by apply applying sophisticated, algorithms or AI to to, in this case, large language models to this, data, you know, we get the we get this new age of generative, of AI. But there's a through line here. Right? That this is data for the most part, now this is data in information that's produced by individuals who are whose data is being either, you know, scraped from them because they're they're using a smart device, quote, unquote, you know, a television, a dishwasher, a refrigerator, a car, or whatever that's recording all kinds of information, or they're engaged more actively, you know, with a a computer or an iPad or a smartphone, and they're engaged in in you know, usually, it's in social search or shopping. Right? That's those are the big big ways that that that that, individuals engage. And and so this data is is is being created at massive, massive scale and being grabbed by, you know, a few large platforms that then convert it. Right? Train train the data, and create these these chatbots and so forth. And it's it's the same model. Right? The same architecture, which is gather up people's data, aggregate it, and apply sophisticated, you know, algorithms or AI or whatever you wanna call it to to it, and for for whatever result is being optimized for. So it's not like we have a broken Internet that is all gonna suddenly be fixed by a more powerful version of technology. As a matter of fact, we you know, I think we have a broken Internet, and the problems are gonna be exacerbated by a more powerful version. You know, you'll have, better misinformation and better disinformation and better deepfakes and more hallucinating by the large language models and so on and so forth. And more more harm can be done and more chaos and confusion can be created. Now having said all that, the generative AI is pretty awesome just as the Internet is pretty awesome. And if we if we had a better design, then we could eliminate a lot of the problems. But we already know what the problems are because we're we we we're seeing and experiencing them pre generative AI, right, or pre the pre the, emergence of the term because it's kind of a marketing term and pre the delivery of a user interface. You know? And and but the model is clear out in terms of how it operates and the the good things it does on the one hand and the bad things it does on the on the other. And I would just say, like like, with any flawed design, the, you know, logic says fix it before you make up more health. You know, fix it before you you, you put it under further stress and and and, and so on and so forth. I mean, we talked a little bit about physical infrastructure in our our last session. Well, there's been all kinds of physical infrastructure that's been built in this country that is, you know, was better than the alternative but flawed. And when the flaws were once it was pressure tested with scale, the flaws became apparent, and then it was redesigned, reengineered for safer, better, bigger, you know, more effective infrastructure. And and I think we can, you know, learn from that. We have flaws in the design, which are now very apparent. We should fix them and then then have the benefits and and and, wonders of generative AI and, you know, solving problems at scale rather than creating problems at scale. So I'm all for AI and all for the the tech. It it's just I'd love to see it fixed before it it it gets you know, it becomes more powerful and unwieldy. And I think, by the way, Matt, that all comes down to what we talked about before. Let's put the power back in individual's hands so that they own and control their data. And, you know, you and I can can have a say as to whether our data is used in a particular large language model or not. And if it makes somebody a jillion dollars, what's what's, you know, what's the deal? And, you know, we we have the ability to say, no. You know, I'm not comfortable with, you know, you using my data because you haven't you haven't proved to be trustworthy or I don't like, you know, your standards of operating, and I'd rather let my data be used in some other, you know, some other platform, that is, more in keeping with my own belief system and, you know, my own standards and so on and so forth. So it it all comes back to agency, and we we've lost agency in the current world. I mean, the pre generative AI world. And I think we need to return agency to individuals first before, again, we we rocket ship into this next next version of the technology. And, as we said in the last session together, it just if one believes in the principles of individual liberty and freedom and rights and autonomy and choice and and, agency and so forth, well, we don't have that now in the app era of the Internet. So let's fix let's address that, fix that, and then, yeah, and then enjoy the benefits of, you know, rapidly advancing technology, which I think will be pretty awesome if we get the the design right.
Speaker 1
12:48 – 15:00
Yeah. I I wanna get into the design a little bit. I'd love to, I'd love to, sort of double click on that. But I have a slight digression I I just wanted to share with you, which is I I I read something recently that made me think of our earlier conversation about about infrastructure, which was, so I was reading about the, the collapse of the Khmer empire. So the civilization that built Angkor Wat. And, what they, they thrived, you know, eight hundred years ago or so, because they built, the best network of roads in in the area. They built this fantastic network of roads, and this enabled them to to to trade and so on and so forth. They also had this great infrastructure around water and things like that. But what what happened is when they when the empire kind of fell on hard times and they weakened, the, they collapsed quite quickly because the roads that they built were used by their, their enemies to to move the, move the troops, into the part of the Khmer empire really quickly. I just thought that was interesting. There's this kind of so there's this sort of, dual aspect to to infrastructure, that came through just in that in that little anecdote that made me think of of, of the conversation that that we were having. And it also makes me think about about about AI. Right? Because if we if we think of if we think of AI as as infrastructure, it's it's important for us to think about sort of how, you know, not only what it can do for us under ideal circumstances, but also what vulnerabilities it creates and what what sort of, faults that it might exacerbate in the, in in the society. So it is serving us, you know, not only on our good days, but on our on our bad days. Right?
Speaker 2
15:02 – 20:38
It's a great it's a great one, you know, because, you know, we have right before us right now the whole TikTok situation. Right? And and, you see, you know, how quickly, Congress moved because of the national security threat of, you know, this two way two way street, between the Chinese government and American citizens. Right? That that is, you know, providing something to American citizens that they they they like and enjoy, but at the same time is providing a pathway in for for China. And, you know, I I I think we need to have a a a bigger discussion about that because this is the same tech that our own big platforms are using, except that the data is going to Silicon Valley, not Beijing or Shanghai. And I get that. That's a big difference. But, nonetheless, I don't think we want all that of our personal information, you know, being centralized and aggregated anywhere, quite frankly. And and and that's goes back to the point that if this is gonna be if we're just gonna, you know, march forward or race forward without, you know, thinking about how to make the the the technology better, we're just we're gonna default into a place where five companies have the data and the compute power, and they're gonna be the beneficiaries of, generative AI. You know? Yes. Apps will be created or or, you know, that will be or use cases or, you know, chatbots. There'll be there'll be things created that you and I can use, but it won't address the issue of who who the data that's collected and and, you know, that continues to power this machinery. Who who will own it? And because, again, it's the the data that is fueling this and particularly the, you know, human data because these predictive models are based on how humans think, not how, you know, machines think. Right? Or or animal other animals think. It's based on human thought processing and and human preferences and human emotions and human tendencies. So all of this information that's scraped from the Internet is the live data that's going into in into these models that are being trained. And and there's only a few places that currently can can benefit from all this. Right? Because they have the head start with having aggregated a lot of data and having built up a lot of compute power. But I would argue that this is not private infrastructure. This is at some there some portion of this has to be public infrastructure that we that we all benefit from and have a say into how it's used, right, and how it's operated, not be, you know, reliant on, you know, an individual CEO or, you know, board of board members of of one of these companies deciding what's good for all of us. And worse yet, one of those company individuals, you know, you know, making some kind of deal with a political figure and giving that political figure all kinds of power because they, by extension, have the ability to use a platform to influence and sway and convince and propagandize and and so on and so forth. So, again, this is you know, when I think of generative AI, I think of this just this is a a bigger, more potent, more powerful version of what we have, and we already know that what we have is flawed, deeply flawed. Right? You don't even have to be an individual to be on the the current version of the Internet. Right? You can you can participate by being a fake person, multiple fake persons, a machine bot, etcetera, etcetera. So, you know, identity is just one example of of of a flawed infrastructure that needs to be addressed before we get to the, you know, next level of of of of power and of of of capacity. And, again, the technology could be awesome and and really be incredibly beneficial for humanity, which is what the original Internet was designed to be, right, to make us all smarter and and and create access to information and connect us and so so forth. It's It's become something very different, which is gonna lead to very, very different outcomes and ones that I'm very uncomfortable with, which is having all this power centralized, the power derived from the information and and and and, the influence and the and the wealth and so forth that that that creates that, you know, then becomes a a cycle of more influence, more wealth, more influence, more wealth, more power, etcetera. And that's exactly what we fought against to become this country, right, from a from a colony to a country and and, you know, and from subjects to citizens. And I I don't wanna see a slip backward here.
Speaker 1
20:39 – 24:55
So when I when I hear when I listen to the AI conversations, of which there are of there there are many, obviously, the way that I sometimes categorize it in my mind or just one little framework that I have for it is I hear I often hear basically monopoly concerns, being pitted against safety concerns, being pitted against geopolitical concerns. Right? So there's these there's these kind of three buckets. And I'm I'm curious how you think about these three buckets of concerns and in particular how you think various different interventions interact with them. So, I'd love to hear your thoughts about two things in particular. The first one is open source. So how do you feel about open source models? Do you think that that is, does that address the concerns that you have with, our concentration and monopoly? And let me let me set that up. Let me set that up. Well, the let me the the next thing I wanna hear about is, is, you know, data ownership and, you know, decentralized, social networking protocols and sort of systems for people to maintain control over data, how that could interact with AI. But let me, let me set up the AI question just a little bit. So or it's not or sorry. The open source question. Because it it seems to me that there's a, like, out of first pass, if you think about open source AI, it seems like a it seems like an intervention that mitigates the monopoly problem, but could potentially exacerbate some of the safety problems and or, some of the geopolitical, concerns. Right? So and by geopolitical concerns, just, just to get it out there, explicitly, what I mean by that is basically, you know, is China gonna get there before The United States in terms of, you know, building a super powerful AI? That's that's what I mean. That's all I mean by that, basically. And, anyway, so but, oh, you know, open sourcing models seems like a you know, to me, it seems like a potentially an interesting intervention or good intervention against the monopoly dynamics. But it's worth remembering this there's this kind of ironic history here in my view because, open source data twenty years ago, you know, seemed like this force that was very powerfully, fighting against monopoly. But but in the last five years or so, open source datasets have been one of the tools that have enabled just a handful of companies who have the, the expertise and the compute to build these models to, to to zoom very far ahead. So there's a there's a way in which, you know, sometimes things that, things that seem open at first can get can get captured in hard to foresee ways down the line. I just think that's an interesting history. And, you know, it's hard to articulate, you know, what how open source models might go, you know, in the next in the next sort of generation of the technology. But, anyway, the the there's there's a lot there, but I essentially, I'm curious to hear your thoughts about, about whether open sourcing models is a good idea, how your ideas about sort of data control, can you know, we'll end up moving power around and how it interacts with, you know, monopoly safety and and geopolitics. Yeah.
Speaker 2
24:57 – 40:22
So, a lot there. And and I'm glad you mentioned, data ownership because I I do think and this is, you know, my perspective on on this, and I welcome, you know, other perspectives because I I think, you know, this whole we know or many, I I think would agree that the current situation is less than ideal, and we're entering a new a new era where we shouldn't just be kinda winging it and hoping it works out. We we should be reflective right now and really fix the design so that, as I said earlier, when it becomes more powerful, it's more powerful for good, right, for humanity. And, and, of course, we need commercial models to sustain it. So, you know, I come at this from the vantage point of how do we best have the technology, embrace and and and kind of, reflect the ideals and principles of a free society. You know? And, you know, this this kind of, capitalist democracy that that that that, you know, has been the, you know, those are the two operating systems that work together to have built, you know, America and, to one extent or the other, other other free societies. And and so that's that's a point of view or or or starting point that I'm that I'm coming forward with that others may disagree with. I I happen to think that this is this is a model worth saving, and improving, right, and strengthening. So it becomes, more just, more fair, but based on the principles of the worth of the individual and, the importance of, you know, you know, rights that are unalienable, rights that only mean something if you respect them in other people, so the social contract, and, you know, a meritocracy where everybody is it has the same chance to succeed, which is very different than everybody gets the same result. Right? So it's a it's kind of the foundational in terms of my my perspective on this. And so I would love to see the technology that we have kind of reflect those principles and those values. And think as we talked about in our prior session together, you know, if I asked you to describe democracy, you probably wouldn't say centralized, autocratic, surveillance based, predatory, ex exploitive. You know what I mean? And it it we are words like liberty and choice and autonomy and freedom and and social contract and and, you know, fairness all come to come to mind. So when I zoom out and I see what's happening, with highly centralized technology, I I I see how an an autocracy like China will use it. Right? Because they they you know, I don't I don't think of Chinese people as citizens in the same way as I think of Americans as citizens. I see them as, you know, you know, people that are proud to be Chinese and and are, you know, are are are living there and so on and so on. But they don't come with the same rights that we have in The United States. Right? So they're they in my construct, they would be more like subjects. Right? And, and the government there is very explicit about its, expectation that everybody kind of toes the same line and that they, you know, the people Chinese people are surveilled. And if they they might get punished if they're not carrying the the party line, etcetera, etcetera, etcetera. I see that as very un American. And so rather than, you know, replicate highly autocratic surveillance based technology and compete with China that way, I would rather see us create technology that actually embraces our core values and principles. And I think that that starts in in into a in large measure by returning ownership and control of our personal data to each of us and, not scraping it and stealing it from us and then using it in a ways that we don't you know, we're not even aware of. I think that goes a long way towards building out technology that where people feel empowered by the data, not frightened by how it's used or feel exploited by how it's used. And, you know, we have a whole regulatory framework that's being built out to what? Protect people. Right? It's it's data protection. Well, that implies that they could be harmed by it. Right? If you need to be protected, there must be some harm that's being caused. Well, why not shift the thinking and let and and and let other people have that autocratic centralized form of technology? Why don't we have one that's empowered? Where we empower our citizens and empower each of us to feel like we're part of the creation of the next great version of technology to to thrust humanity and civilization forward. It's it's a it's very doable. And I actually believe if you create a an Internet with integrity and where people are are back in the driver's seat, they'll share more information and and do it in a way that the information that that it's being put in into these large language models is is highly calibrated. It's very accurate. It's very clean. You know, you don't you won't have a very high fidelity. You won't have the the noisy data we have now because you have so much garbage, you know, that's that's that's, you know, that's echoing around the Internet. And then, you know, you you you have technology that kind of embraces values and enjoys the benefit of the power of this, you know, massive tool. But it's built on a different set of principles, and those principles are what has allowed The United States to become the dominant, you know, meaning biggest economy, best version yet of, you know, democracy that that humankind has seen, you know, imperfect as it is, and best version of capitalism that it's seen imperfect as it is. But we should do better in terms of fairness and justice, but we're not gonna do better, I don't believe, by taking a radical pivot to to autocracy. And so I just feel like the technology is so powerful that if we don't have it embrace this idea of of agency and liberty and choice and and in the power of the individual that we're going to end up competing with with with China, and I don't think we'll ever outdo China at being China. You know? I don't think we'll ever we might surveil as well as well as they do. And but I don't think we'll we'll be any better at it. And, and and they also have a have a population that is conditioned to being, you know, towing the party line. We have a population that's conditioned for freedom and and and and liberty. And so, I just think what's been our secret sauce are these value core values and principles, and why not embrace those as we as we build powerful technology? And by the way, I think if if if we do that, we we will unleash massive, massive invention and innovation and creativity that China will never never unleash and, with their model. And, you know, we'll have the best, freest minds operating here, and and and building out the alternative to a centralized surveillance based technology. And I think that would be epically, you know, wonderful because it will open up all kinds of economic opportunity for people and and innovation that we can we can only sit here and imagine. And so I think that's for me, the key to all this is to go to the very basic, like, what are values prop what's the values proposition, not the value proposition? After you you you settle on the value values proposition, now you can talk about incentives and economic structure. And you said that, you know, the you know, our current model kinda took a track that was a bit different than quite a bit different than what it was originally designed to do, and we're sitting here with these large platforms now that are the the ones that have the data and the compute power. Well, that happened because of incentives. That happened because of an economic model that, you know, an an attention economy, an ad tech industry that fueled these platforms, you know, to to, you know, connect consumers with products, right, and to sell ads. And that's that incentive model, to show you how powerful capitalism is and how powerful incentives are, you just have to look at how a decentralized Internet became highly centralized because of of incentives. So let's not lose sight of incentives here. We're gonna open source, I think, is necessary, and important at a certain place in the stack, the tech stack. But there will be use cases which will you know, people will creators, builders, entrepreneurs will build things to create value. And let's let's have them incentivized correctly and certainly do it with permission. If somebody's gonna use my data, I'd like to give them permission. And and that's why I said in our last session, let's have an Internet where we're the new platforms are clicking on our terms of use. Right? And if you wanna use my you know, you saw you see the New York Times copyright lawsuit, right, against these these these, platforms. Well well, maybe New York Times would be very happy to have their their information used in some of these models if it was permissioned, not just scraped. Because right now, everything has just been ingested by these machines regardless of who created it or who owned it or how valuable it is. And it's just like a, okay. We own it now. We're gonna go build build stuff and, profit from it. Well, that's that's not right. You know, that's not fair. And let's just, you know, catch our breath here and build it in a way where people give permission to use it. They're gonna people are not gonna not let their data be shared. I mean, we're not gonna you and I are asked to use our data to cure a disease, we're gonna say, sure. You can use it. But, you know, you can use this piece of my information for that purpose only. You can't resell it, and you can't have everything about me just because, you know, you want some piece of, medical information or biological information. And so, you know, the social graph is rich in all kinds of information, and we we should have some ability to parse it out the way we want to and for what purpose. And, you know, maybe there's another piece of my information that's only gonna be used for somebody's commercial purpose, and I'm happy to you let them use it, for an economic, you know, trade. And and, again, all all that is doesn't even get to the the the point of, am I sharing information that's connected to me as a person with a with a name, or am I sharing information that's connected to a purse a verifiable human being, but I'm not giving up any of my privacy? Very different very different deals. Right? Very different, you know, you know, transaction, so to speak. So, yeah, I think the the primacy of who owns and controls the data and decides how it's used is really the key to all of this, Matt. And then incentives will be another key because we we need to build. We wanna we wanna have the best AI and the best technology to help people. But, you know, if we unleash the creativity, we'll we'll build something awesome. I'm totally convinced of it. There's a lot of very, very smart people out there that would prefer freedom and liberty to, you know, autocracy and and, you know, dictatorship or any other kind of kind of centralized political system. So, you know, when when Kennedy said in the sixties, you know, we're going to the moon, you know, that that got a lot of people excited about figuring out how to do it because he didn't know how to do it at the time. But he said, let's do it because this is a new frontier, and we don't we we don't want it to be we want it to be open for humanity, for civilization, and for learning not to be controlled by someone else. And I think we're at that kind of moment again where we we can build some great stuff, but but it was the the values that Kennedy brought forward in that speech. It wasn't just the tech that he brought forward. Right? The tech hadn't even been built. He was talking about a set of values and aspirations and and ambition for human civilization, and then the tech the tech got built to to enable it, embracing those values. And that's what I'm talking about here.
Speaker 1
40:23 – 43:14
Yeah. I mean, if you if you take the value of, autonomy, what strikes me is that, there are are a couple of sort of uninspiring visions, you know, for the future of autonomy. And I think that what what you're arguing for is the articulation of of a of a different one. Right? And the the the two sort of uninspiring visions are so on the one hand I mean, the elephant in the room here, I think, is that AI is a very powerful technology for manipulating people, basically. Right? And, it on the one hand, we've got a a vision of the future in which an autocratic government can use this technology to manipulate its subjects. And on the other hand, we've got we've got a vision of, of basically handful of private actors using it to manipulate manipulate markets, manipulate consumers, these kinds of things. Right? So in other words, there's this kind of there we've got sort of a an autocratic and a and a capitalistic, dystopian vision about how this technology can be used to manipulate people. And, I think the question is, is how can we imagine it how can we imagine this technology being used to enhance our autonomy? And, I think that, that there, you know, there there are different sort of ways in which that might work too. But what you what you seem to be saying, and correct me if I'm wrong, is that if we all had a bit more control over how our data was used, then the terms that we would be able to impose upon the users of our data or the people building models with our data, would, would would would be such that they would determine the sort of downstream uses of the technology in in a in a better way. Right? So that, for example, you know, all of the providers of data to to models could could could join together and and, impose terms on model builders that would ensure that the that the models are, you know, not being used deceptively, not being used manipulatively, these these kinds of things. I mean, I I'm I'm trying to sort of mirror back what I'm hearing. Is that is that right? Is that how you see it too?
Speaker 2
43:14 – 52:54
Yeah. No. I I it it it's totally right. And I I the only thing maybe I would tweak about what you said is that I I don't think it's quite as binary as you you know, you said there are two models that are ones that we don't wanna replicate or or perpetuate, and it's the one that kind of manipulates us as people. Right? And where where where we lose our citizenship and become subjects, and that's sort of a political dynamic. And the other is and and and that's kind of you you you talked about China in that context. Right? And the other is a a corporate model, not a government or communist party model where the data is is is, the advantage of a few a handful of companies, and they, you know, you have incentives that kind of are drive us to do things economically. You know? So it's a marketplace kind of thing. I I my one tweak is we we see I I see it as more than just the the corporate version as more than just manipulating or optimizing for an economic incentive model. So I I see it actually now having spilled over to actually well beyond the economics, you know, marketplace issues into the into our culture and our society and our politics. And it's affecting our our our ability to operate as a democracy and so forth because the the incentives are are not just economic. The incentives are also for, rewarding the most extreme behavior. And I and I say behavior deliberately, because it's not only most extreme viewpoints. It's it's it's behavior because when the the like and follower buttons were added, this kind of surveillance based technology, you know, which was accumulate information to sell ads, you know, to to to sell ads, to sell things, went into overdrive because it became a performance based Internet. And it was, how do I get the most likes or clicks or followers? And that was the most extreme behavior or position or whatever. And, I mean, one only has to look at, you know, pick a a moderate politician that makes sense when they talk. They may be right of center or left of center. It doesn't matter, but they're kinda moderate. And and and see how many, you know, followers they have. And then go to an extreme politician on either side of the spectrum and look at how many followers they have. And there's just no comparison. All of the the noise and the and and is at the extreme edges, and that's that tends to get then, you know, the media picks that up because the media is, you know, kind of driven by this search engine optimization model. Right? Where well, if the if the stuff that's getting picked up the most is this extreme stuff on either side, that's what gets reported on, amplified even further, and then we become, you know, anybody in kind of the middle. I don't know what the percentage is, but just take it take 70% that's, you know, in the middle and 15% at the extremes. That that's not science. That's just I'm trying to make a point. It's it's that 15% at either end that may not be at all reflective of the 70% in in the middle whose viewpoints may be more nuanced. They may agree with some things, disagree with others. They have a they have their own personal curated perspective on life and what's appropriate and what isn't. But they're what they're getting is, you know, kind of this extreme behavior on on either side, and that gets amplified even by conventional, non, you know, social forms of media, just, you know, sort of traditional media and and, you know, traditional TV, traditional newspapers, and all that stuff. And it it's just this so I I I my only tweak to what you say is that I I I think this model, which got further further deteriorated with the like and follower buttons where it became performance based and rewarded, you know, extreme behaviors. By the way by the way, including things that aren't weren't even true. There's no there's no, like, reward for truth or reward for discovery, you know, or reward for a better idea or reward for reward for advancing, you know, civilization or strengthening democracy or protecting kids. The reward system is for extreme behavior and that includes lying and and, making making stuff up and, and just it it's almost a yeah. It becomes a performance based Internet, and and that is, you know, that's entertaining for some, I guess. I don't I don't know. But, it's not conducive to the types of things we've we've been talking about here, which is how do you actually use technology as a a force for for good? A force and good, I've already given you my perspective on what I think good is, right, which I'm trying to to it to be a force for a a fairer, better economy, and a more just and and stronger, you know, democracy and, all of that that all of that entails. And I I just don't think we have technology that's doing anything to help right now. And yeah. So why make it more powerful before we fix what's wrong? If we agree that these are noble goals and and and worthwhile goals. If we if it's all about I I kinda see this book in the road map, right, where where, you know, you we could turn left and follow the machines and the technology and say, you know, we're we gotta go build the best AI and the best technology regardless. You know? Or there's a path to the right, which says, let's not, you know, forget about our you know, what makes us special and what's made us successful, and what this country's been about for the last two hundred and fifty years. And let's take that path forward and use the technology, build technology that's every bit as powerful as we're gonna we're gonna get if we took a a left turn or the left fork, but it's gonna actually also, you know, make our, we're not gonna give up our our self sovereignty. Right? We're not gonna give up our agency. We're not gonna give up who we are just so that you know? I think it's time that we have a an an IP address for individuals on the Internet, not, you know, just an I IP address for devices. I think it's time that we have our that we show up as individuals on the Internet and and and and and control us, on each of us on the Internet. That's that's what agency is about. You can't have agency in a, You know? We're we're we're not going backward. The tech is here. The Internet is here. Could be great. Not working great now. And one of the big miss one of the big misses here is that we we let ourselves become devices. And we we we if we want agency, we need we need we need to take back control, take back take back power of our data, and then have this awesome Internet, you know, kick into the next and and and tech kick kick into the next gear. But, we're seeing the power of this of the Internet, the power of data, the power of our social graphs, the power of technology generally, the power of generative AI. All of it is, you know, pretty amazing, but let's let's have it embrace the right values or the ones we agree on, and then let's have fun building again, you know, beating the competition, but with our secret sauce and with what matters to to, you know, you know, to certainly, what what what matters to me and presumably yourself because we're having this conversation, which is, you know, individuals matter. We matter. And I'd rather be a citizen than a subject. And I don't wanna give up my citizenship just to use, you know, the Internet or have the benefits of generative AI.
Speaker 1
52:55 – 57:50
Yeah. Yeah. I I mean, I I I couldn't agree more. And the, it strikes me that what we need to do, to to avoid this technology being used in ways that won't make us happy is we need to sort of say what we want. So the these these sorts of the the kinds of values that you just mentioned, like if we so for example, if we want AI to to to serve, you know, truth or or discovery of new technologies of one kind or another, or, you know, these kinds of things. We sort of need to we need to put down those markers in a way. Right? We need to articulate them. It seems to me that if we don't articulate what we want the technology to do, then the technology will, you know, it certainly has the ability to do the same kind to create to serve the same kinds of incentives that that, for example, social media systems have served. Right? So so for for example, like, you know, we we seek novelty in a way. Right? As as as consumers or as, you know, as viewers of entertainment. Right? We wanna see something new. We wanna see something different. We wanna see something that stimulates us. And, this creates an incentive for content creators on social media to say things that are that are more extreme, that are more inflammatory. AI can do that too. AI can also, you know, create, create content that, that scratches those same kinds of itches. If we don't want it to do that, we have to sort of say what we want it to do and point it in a particular direction. And, what I, you know, what I spend a lot of time thinking about, what I'm what I'm really curious about is, you know, how do we lay down those markers, right? How do those those kinds of those markers, those values that we think the technology should be serving, how do we how are they going to emerge? Right? Are are they going to emerge through, you know, and, I mean, one way that I'm in you know, that interests me and that I'd love to see them emerge is, for example, if people had more power over their, over their data, if people had more leverage or more control over the way that their data is used, could they, could they deliberate about what they want it to be used for and, lay down those markers about how the data can be used in a way that, that, you know, downstream actors building models with their data would need to respect, for example. Can those markers be laid down through, you know, another model is those markers get laid down through regulatory processes, you know, at the government level. There's also, you know, I think one interesting example here is the work of my colleagues at the Collective Intelligence Project, who are, running, innovative democratic processes, that articulate values and then try to build those values into the sort of constitutional AI structure, in anthropics models. Right. So this is another way of kind of articulating what those values are and trying to trying to make sure that the AI systems we build are pointed towards them. So what I'm asking is, you know, how do you think like the, you know, the difficulty of course is agreeing. The difficulty is, you know, how do we all see each other and have a conversation, have a deliberation, take a vote, go through some kind of a cultural process in which we, articulate these things. You know, how do we how do we get to the point where we've got enough enough consensus or enough legitimacy around a few values that we can, plant them in the ground in in this sort of clear and authoritative way that we probably need to and ensure that, that we're that we're steering towards them. I mean, how do you how do you think that what does that what does that look like?
Speaker 2
57:51 – 69:03
Yeah. It's I I think that you're circling around now the the set of issues that are are really front and center as we reinvent, reinvigorate, reenergize democracy, and, and, you know, not not romanticize about it, but actually, reenergize it. And couple things. I love the way you set this question up with it it requires each of us to engage and express our our our viewpoint on on things. Right? Being being a citizen actually comes with a set of responsibilities. You can be a a a very passive subject. And so for me, if I think of subjecthood, I think, you know, passivity fits right into into into it as one of the characteristics. If I think of citizenship, I don't think of passivity. I think of, you know, being active, responsible, accountable. Right? Each of us are accountable for for our own our own behavior in a democracy. And the the the next point is when you think about it, democracy isn't about getting your own way all the time. It it it it it that will seldom or never happen. It's it's about, as you say, reaching consensus on, like, this is the best way forward at this moment in time for all of us and, staying true to what we are at the core. And, you know, that's why the the thin layer code or protocols or founding documents, the constitution, the bill of rights, even the declaration of independence, these are very, very, you know, deep on the the proposition, deep on the values and the the principles, and very short on the prescription as to how to how to implement those or or how best to, you know, to do it. It it that's left up to us. And, you know, the founders knew that they were gonna be around for such, you know, for a period of time, and they would give it their best shot. But then some the next generation would have to give it their best shot and so on and so on. And here we are. It's here we are. And it's kinda simple. We can just sit back, be passive, allow ourselves to be subjects, and let the machines do their thing. You know? We can all be, you know, sit there and say, aren't these machines awesome and wow. Or we can say, wait a second. Yeah. The machines are awesome, but they should be we should be the master of the tool. And I wanna be a citizen. I wanna and I'm willing to do what it takes to to to citizen. Right? To think of it as a verb, not just as a noun. And and that means, I'm gonna you know, I care about I I I care about the future. I care about the next generation. I care about my fellow man and woman. I care about, you know, my family. I care about my own life, and and it's worth investing in these things. And some will invest more than others. I I get that. It's not, you know, asking everybody to make it a full time job, but it is. We do need to move away from just being entertained by, you know, the Internet and and kind of becoming somewhat passive, because of it and because of its design, which is kind of addictive and and does, you know, get people spending hours and hours, you know, online. And, you know, in in kind of, like let's get the light bulb to go off and say, you know, we can have all the benefits of this and, you know, we can we can be entertained, but at the same time, we could be smarter. And we could be more engaged, and it could be more fulfilling and fun. And we can actually build the future that we as individuals feel more fulfilled by, and we're actually doing something good for each other and for the next generations. And, you know, without that kind of, you know, perspective of the future, we might as well take that left fork and just follow the machines, because the right fork does involve, you know, showing up. It does involve being, being a citizen and and expressing a viewpoint and and, but also listening to others. And and coming at it from the perspective I think everybody should come at it and say, like, what do we all agree on? Do we all agree that we're we're worth something? That we'd prefer to be citizens and subjects? Okay. Good. Do we all agree that we're equal? You know? And should be treated that way by by others and by the law and and so forth? Okay. Good. Do we all agree we should have the same opportunities in life? Okay. Good. You know? Do we all agree we we we're we're smart enough to govern ourselves, and we don't need to have, you know, somebody be appointed to or because they're born into a family, govern us? We can we can figure it out ourselves? Okay. Good. Etcetera etcetera. You know? Do we all agree, by the way, that if the rights are gonna mean anything to each of us, we gotta respect them and other people? You know? Okay. Good. We don't have to agree on much more. You know? We just have have to agree on the basics that we're taking the right fork. We're gonna take the right the the the the the the fork in the road to the to the right and figure this out together based on a set of, you know, principles where we're each worth something. And it's we're each worth a lot. And we're we we are equal, and we're we are we have ideas, and we have, you know, everybody carries their own dream forward. Right? And let's let's express that. Let's amplify that. Let's give people that excitement that comes from the opportunity and the possibilities that we're we're all driven by as human beings and the relationships and the the learning and the getting smarter and and and and and the the acts which make us feel good about ourselves and and the acts of others that when we see them happen, we feel better about all of us. You know? And let's let's embrace that and have technology that enables it. And, you know, the last thing I would say is I get nervous when we start thinking the technology itself is gonna do it for us. You know? That's that's the one part I, you know, the the thin layer legal code or, you know, governance code that's in the constitution, the bill of rights, and so forth, it doesn't do it all for us. We have to figure it out ourselves consistent with a set of principles and values. And, that's a that creates an obligation on us to sort through that. And and I don't think the thin layer code and, you know, computing code is gonna figure it all out for us. I think it can, you know, it can do things like give us ownership and control of our data. That it can do. It can give us agency. It can give others, you know, the same. And it it but it then comes with us having the burden of figuring out the rest and how we want it to be used. And what's and, you know, but, generally, people are I find, if if you take away the the triggering and the manipulation and the brainwashing of people and you just you just get back to basics, people are good for the most part. People wanna be smarter. People wanna make more money. People want their family to be taken care of and safe. People help their neighbors when they see them they're they're having a problem. People do a lot of great things for one another, notwithstanding all the, you know, all the the bad stuff we read about and and and see around us. But why amplify the bad stuff? Let's amplify the good stuff. Let's you know, right now, with a poorly designed technology, it it becomes a tool for people wreaking havoc. You know, it's bad actors both outside of the country that wanna disable our country and bad actors inside the country that would have whatever motivation they have, you know, can use it as a tool to do bad things. You and I aren't getting off the Zoom call and gonna go and figure out how we can use the Internet to do to, you know, to undermine, you know, to harm kids or undermine democracy or create create more mis or disinformation. But there are people doing it full time. You know? And because it's been weaponized, because it's not it's not designed it's it's it goes back to your, Khmer empire example. Those roads can be weaponized. And and and and and and and, you know, and become the actual device that overthrows the empire. And, I think we're at one of those points right now where, you know, something built for for good and to to to help us advance, not just as Americans, but as as as humanity and as civilization. You know, believe me, I've talked to the people who created the Internet, and they did it for they did it for humanity. They didn't do it for money. And, and they they they they did it because they saw the potential and the possibilities, and and they're saddened by what it's become. And I think we owe it to them as well to fix it before it's too late.
Speaker 1
69:04 – 71:04
So I think I've got sort of a hard question here, but I I I wanna ask it. What what it is is, I think you're articulating a vision for the future of technology, which is really quite macroscopic. And the question that occurs to me is whether you think it should be a political question because I, you know, and on the one hand, I think that, you know, me and many other people who work on these issues are often sort of tempted to say, that that we're all on the same side here. Right? That this is that we're all sort of, that these questions about the future of technology concern everyone. They bridge across political divides. They they certainly do bridge across conventional political divides like like left, right. This kind of thing is is, I think, quite orthogonal to that. But it does seem to me that it seems like it might be useful for us to be having conversations at a national level, say, about the future of technology about, you know, sort of do we wanna go this way or do we wanna go that way? And, and I wonder if it needs to be somewhat politicized because, because the reality is that there are two, there are forks in the road here. Right? There are different ways that we could go. And, do you think that in order to in order to really coherently pursue this way and not that way, we need to sort of define that difference and get make it part of the conversation on a political level?
Speaker 2
71:05 – 74:51
Yeah. Absolutely. Absolutely, we do. I think that it's essential to the outcome here. It's why Project Liberty has a third link to the stool. It's why, you know, a lot of people wanna focus on the the technology because it's interesting, and it's such a big part of our lives now. And so they focus on the tech track of Project Liberty. And, you know, to to some extent, they focus on the, you know, all the work that's being done in the policy and governance and research areas, you know, where with our academic partners and and so on and so forth. But, really, it's the third leg of the stool that I think is the key one, and that's the the the movement or or campaign leg of this. This this, I I don't know. Can you think of a bigger issue that, in in in the day and age we live in? I cannot. I cannot. No. Nor can I? And so shouldn't our national political discussion engage in the biggest issue of the of the day? I certainly think it should, and I'm I'm disappointed that it's not front and center in the in the, national political, you know, discourse, and and debate. And I think it's a miss. And and sadly, a lot of the time and energy is being because there's no discussion about the biggest issue of the day, which which might have the potential of of actually driving the conversation towards a more hopeful future for all of us, the vacuum is being filled by a lot of stuff that is irrelevant to our lives, you know, in in in the scheme of things. Right? It's it may be relevant in the in the I'm not I'm not don't mean to be commenting on any of the specifics. Just in general, our our days are full of news, quote, unquote, that is anything but, you know, on the topic we've we've been talking about for the last hour and fifteen minutes. It's it's, it's about stuff that people have already kinda made up their mind driven by the technology that we talk about, you know, to to pick your side and, and then get that information, whichever side you're on, to reassure you that you're correct and and to give you the evidence, quote, unquote, to argue with someone on the other side as to why they're wrong. And you're right while the person on the other side has the has the evidence, quote, unquote, to prove that they're right and you're wrong. And so we we sit in this endless argument about, not about the biggest issue that confront confronts us. Rather about it we we're we're making ourselves smaller, not bigger, and that's never a good thing. I think it's it's time to have the conversation. I think it's essential that we have it. And, if it's not gonna happen, in our national political discourse, then nothing that stops us from having it. You know, we're citizens. Remember?
Speaker 1
74:53 – 74:59
Frank McCourt, thank you so much for taking the time for the conversation, and, thank you for all the work that you're doing.
Speaker 2
75:01 – 75:05
Thanks, Matt. I've enjoyed enjoyed the conversation once again. Talk soon.
Speaker 1
75:05 – 75:08
Alright. Talk to you soon. Thank you. Take care. Bye bye.
Speaker 0
75:08 – 75:46
The Radical Exchanges podcast is executive produced by G. Angela Corpuz and is co produced and audio engineered by myself, Aaron Benavides. If you would like to learn more about Radical Exchange, please follow us on Twitter at rad x change, or check out our website at radicalxchange.org. And if you'd like to join in the conversation, we'd love to hear from you. So hop on our Discord where we have channels discussing topics like what you heard today, as well as topics like plural voting, community currencies, soulbound tokens, and more. There will be links to all these in the description. Have a great day, and stay radical.