Greetings, recording, and family
Greetings.
Hello. Good morning. Good to hear you. How are you?
Back in Taiwan and feeling pretty good. How are you doing?
I'm okay. I'm also good. I had a good summer. I'm in the countryside, just going back into town. And yeah, I've been working a lot. I don't know — I mean, you know, you're also involved in all these many activities, and these are complicated, exciting times. It's demanding. But thanks for taking the time for this. This is journalism, as I also used to do — and still do sometimes. This would be for Die Zeit, the German publication. Let's see if this works. Why is this not — I'm just checking whether the recording works.
Okay?
I have this new thing, PLAUD. Do you know PLAUD? Is this a good thing?
PLAUD?
This recording — PLAUD. So it's — yeah.
PLAUD. Yeah, I've heard of it, but I have never used it before.
Yeah. Yeah. Anyway, so, yeah, just briefly: where are you? You're in Taiwan. You're in Taipei, or in the countryside, or are you on vacation? What's your—
Yeah, I mean, Tamsui is in the north of Taiwan, part of New Taipei City. It's now very close to Taoyuan City and the airport. It used to take more than an hour, maybe an hour and a half, but now it takes half an hour to get to the airport because of a new bridge. My parents are here, and my grandma is here, so I'm just spending some time with family.
Very nice. Very nice. You have siblings, or, you know, only child?
Yeah, younger brother, four years younger.
Publication and radical transparency
Okay. Nice. Okay, well, should we jump in? Should we just — I mean, it's meant as an interview, but they might want me to write it up as a profile. I don't know. But I'll send you the interview to have a look at it, if that's fine with you.
Sure, of course. And usually I publish after you do, after the site publishes. And also you can check my site as well, on archive.tw. It's all just part of this radical transparency thing.
Is the internet dead?
Yeah, of course. No, great. Good. So let's just jump in: is the internet dead?
Well, since we're still having this conversation, it's quite obvious the internet is still working, at least as far as Signal is concerned.
But some people discuss, at least, that either corporate logic or AI — I guess those are the main factors — are at least undermining how the internet works, as an ideal or as infrastructure.
Not only can I hear you very well, I can also hear the — are those birds chirping in the background?
Yeah.
Okay, yeah. Okay, well, good to know that it's not synthetic, or Ghost in the Shell. So, to your question. Yeah, so to your question: the internet is not dead. But the question is really about who mediates our conversations, right? So what changed is not that machines can now synthesize bird chirping, but rather that in some channels, unaccountable mediation — so-called peak slop — is becoming the default, right? So the feeds may decide what we see, and generative systems decide what gets written or what gets played, with no one owed an answer. But that's a design choice. It's not a law of nature. And as our conversation can see, it's exaggerated to say that the internet is already dead. I think our conversation today is very much alive. And I think I would also say one particular business model, right, the attention economy, is exhausting its soil. And soil, unlike oil, can be tilled back to life.
Data as soil: bridging incentives and Taiwan's scam-ad deliberation
Yeah. How does that work? How do you get back to life? I guess that's a key question today.
Yes. So one of the main ways we shift from data as oil, as extraction, to data as soil, as regeneration, is to change what this mediation optimizes for, right? The attention economy pays for whatever polarizes us, because outrage holds the eyeballs. That is the extraction. And the alternative already exists. We can build bridging systems that surface the statements people across divides actually agree on, rather than the ones that split us. So we can just shift to a different incentive system, making it pay more to build constructive bridges than destructive outrage.
In Taiwan, we ran this on a national scale. When scam ads flooded the platforms, we sent 200,000 SMS invitations and brought 447 people into small deliberative groups. And the incentive here is to have each group of 10 produce something people can live with. So we set platform liability, KYC for advertisements, even slowing down connections for platforms that did not comply. And then they all became law with cross-party support.
So the tilling of the soil is very practical. You just use levers such as joint liability to reward what bridges instead of just what enrages, and give people a say in the ranking, and then the trust can grow back. The latest numbers now show that compared to the height of the deepfake fraud, today it's down by 98 percent when it comes to impersonation scam ads, down by 99 percent when it comes to the investment fraud ads. So obviously, just like the ozone layer can heal itself, so can the information environment.
Interoperability and digital freedom of movement
You talk about Taiwan and, I guess, a strong role for the state or civil society. But corporations now determine the feed — what you see. How do you democratize the fight for a free, or different, internet?
Well, first of all, I think it is already democratized. People everywhere already have such tools. What is lacking is not access to the tools, but whether the existing mediators and intermediaries answer to civil society armed with those tools.
In many jurisdictions, states face a kind of dilemma: either you build a nationalized platform that is run by the state and therefore carries its own state-surveillance problems, or it is built by big tech, in which case it carries surveillance-capitalist forms. But I don't think this is about choosing the state over the big corporation, or the other way around. It means choosing the defaults so that civil-society alternatives — such as Signal, the software we're using for this very interview — can compete fairly with entrenched players.
One part of that, I believe, is interoperability. The European Union has regulated this as part of the Digital Markets Act, so one can maintain instant-messaging flows across providers. And I understand that, in the consultation, many people are saying, “Let's extend that to social media too.” In Utah, they have already passed the Digital Choice Act, so that when you switch from one social-media provider to another, you get to keep your friends and communities; new likes, follows, and reactions still travel across the move. I call this the digital freedom of movement. The state's role is not simply to nationalize a platform or build its own. It is to ensure that the information superhighway always has an off-ramp and an on-ramp.
Transparency, participation, and symbiosis
Great. Can we talk about your ideals for a technological future? You just mentioned radical transparency. Could we go through a set of principles that you find relevant, starting with radical transparency, and then see what other principles there are?
Certainly. I practice radical transparency, meaning that whenever the state makes a decision, it makes available not only the decision itself — the what — but also the how and the why. The process is made radically available, especially to journalists and investigative journalists. But in Taiwan, all citizens are civic journalists, so by definition it is available to everyone.
This is important because, in some jurisdictions, “transparency” means the reverse: making the people transparent to the state. That is extremely dangerous. That is how authoritarianism becomes totalitarianism — not subtotal, not just one part of the total, but truly total surveillance. That is why this interview will sit in my public archive, archive.tw, for future generations to see.
Equally important to transparency is participation. I call it DPI: digital participation infrastructure. It is not just a vote every four years. It means placing AI systems into the loops of each community, so that the people in each community collectively hold the steering wheel around the AI systems they use. Instead of plugging individuals into a loop of AI like a hamster wheel, we need to restore existing communities' control over the algorithms and AI systems that affect them.
More and more communities now understand that they do not need a huge data center or a huge black box somewhere else. Most of the assistive roles that everyday algorithms play can run on personal hardware or community-owned hardware. That leads to what I call symbiotic intelligence, or symbiosis: instead of one big Skynet running a parasitic relationship with communities, the system is constantly refueled and regenerated by community participation. Transparency, participation, and symbiosis are three very good rules of thumb.
AI as assistive intelligence
You brought AI into that almost automatically. Of course, it is the big technological factor of the present. Could you take one step back and say what the role of AI is? How do you see both the potential and the peril of AI in shaping a fair, democratic technological future?
Certainly. First of all, I think of AI as assistive intelligence. I do not have in mind one superintelligent future — which is some people's fear, or perhaps their quest — in which something sees everything and makes every decision better than humans do. Assistive intelligences are more like connective tissues that stay within particular communities and are always accountable and loyal to our relationships.
Many people now see that if you train a very large AI model to seek a high score, or to pursue goals at any cost, it can go critical — pass some tipping point — and perhaps irreversibly damage the internet itself, or our trust in the internet, just to score very highly on some examination. We have seen more and more of those cybersecurity incidents recently. The peril is not necessarily malice, as if the AI wants to damage the internet. It is more like an examinee that cannot be trusted to play by the rules we set for the examination.
The potential is the inverse. Instead of maximizing a goal, we make sure the system simply satisfies what a relationship requires. If I speak a different language from you, we can have a custom-made small model that does the translation and nothing else. Or, for example, you now have a small model — PLAUD — that is concerned only with turning what we say into a transcript and summarizing it. If these are small models, custom-made for particular purposes within a comprehensive set of services, then AI becomes the most powerful listening instrument democracy has ever had.
The peril and the potential have the same technological shape. The difference is whether our institutions are overtaken by the technologies and we give up our civic muscle — our ability to remain accountable for our own decisions — or whether we strengthen our civic muscle and use AI merely as scaffolding.
Reverse alignment
You recently formed a coalition around what you call reverse alignment. How is that connected to what you just explained, and what is it?
Reverse alignment is the idea that, instead of investing everything in making AI stronger, or controlling AI so that it obeys one or two principals' commands, we should make sure our institutions can use AI without succumbing to a concentrated superintelligence takeoff.
That means redesigning our roles within institutions, our existing skills, and the way we govern ourselves. Instead of blindly chasing productivity, we can share prosperity. Instead of merely ensuring that AI can synthesize more and more realistic images, we can make verification more accountable than it was before: not just “seeing is believing,” but ensuring, for example, that everything has a digital signature that does not compromise privacy.
And instead of allowing only existing power centers — including big tech and the state — to use this technology, we want the paths to appeal and petition, and people's ability to associate online, to grow in tandem, so that governance becomes scalable. Reverse alignment includes all of this. We have twelve different pillars, but the idea stays the same: the ambition of our technology should not dominate the ambition of our social capabilities and social systems. The two should evolve in tandem.
Network-shaped institutions
And what is the shape of these new institutions?
I think the shape we're looking for is closer to the internet itself, or to the way neural networks work. Bruce Lee, from Hong Kong, had a saying: “Be water.” So this is not a new super-ministry, but a set of more portable muscles — what I call the 6-Pack of Care — that existing institutions can grow within themselves and across one another.
Take any community: a school district, a hospital, a city, and so on. The community can come together and author its own evaluations, so AI systems are examined against what that place actually needs, rather than against a generic benchmark. There is a standing path of appeal, so when a system acts, someone there is authorized to answer for it. There is also a collective power to pause and roll back, held locally rather than at headquarters overseas, and always the right to move to a second source, so no single provider becomes a landlord.
None of this requires concentrating power in data centers. Most of it can run on hardware communities already own. These are old democratic capabilities — examination, appeal, recall, and association — reimplemented in the shape of networks.
The 6-Pack of Care
Can you explain the 6-Pack of Care a bit more?
Certainly. The 6-Pack of Care is a set of care-ethics principles that I developed with Caroline Green and many others in Oxford. You can see it at civic.ai, with beautiful illustrations. Very simply, the idea is that we need to align AI systems within the flow of communities.
Care ethicists, beginning with Carol Gilligan and Joan Tronto, built an entire branch of ethics distinct from simply chasing the highest score — consequentialism — or following written rules — deontology. It is about interactions within actual relationships. Tronto says good care has four elements.
First is attentiveness. You begin where the needs actually are, listening from the place of the people rather than the place of the technology maker. We need to attend to people's actual needs — not only those of the powerful and already heard, but especially those of the voiceless.
Second is responsibility: committing to care. If you start listening to people who take a great deal of time to speak to you, you cannot simply say, “Here is the summary, and thank you for your input.” That is not responsible. We need to pre-commit, so people know exactly how their words, ideas, or input will affect whom, where, how, and what. That is an engagement or participation contract.
Third is competence. When we deliver care, the people receiving it should know what is going on. In open source, we are not so afraid of agentic coding precisely because the source code is there: people can see what the AI changed, and the AI can be transparent about how it is doing the work and the chain of thinking that went into it. By contrast, in many areas of policymaking, people feel pressured to introduce AI black boxes with no source code in their trajectory. That creates tension, because it is not competent care if you simply delegate the work blindly. It is like sending your robot to the gym: the robot can lift a lot of weight, but then we lose muscle, and we lose friends. Transparency is part of competent care.
Fourth is responsiveness. If the need changes, at some point we may no longer need the same service the AI was providing. Sometimes we simply need to change and let go. When the need changes, so should the care, so the system must remain responsive. When people who care build the muscle, we can become our own caretakers. Knowing when to let go, rather than creating dependency, requires people to author their own evaluations of success instead of inheriting old evaluations that tend to prioritize lock-in and dependency.
Attentiveness, responsibility, competence, and responsiveness form the core care loop. If you run it well each time, the people who receive care grow stronger at caring with one another. Joan Tronto later added a fifth element, caring with: care does not come from just one provider. A whole society can arrange care together, so people can count on care being there over time, on terms that are just and equal for everyone. We call this solidarity, the fifth pack. Portability and interoperability should be positive-sum. If you change providers, you should receive more care in total, not vindictiveness from an old provider that refuses to let go.
Finally, there is symbiosis. As I mentioned, an AI system should be as local as possible. Anyone should be able to run local inference and, increasingly, local tuning, so that the system can be symbiotic rather than parasitic: not continually extracting data as oil against people's wishes, but regenerating data as soil when we care for it properly. Symbiosis plus Tronto's five elements gives us the 6-Pack of Care.
Place, community, and the human loop
So place — or community, as you call it — is central in some way, even though “central” is not the right word because this is decentralized. Is place or community the most important thing?
Yes, and — or — the human loop. There are many ways to say this. But instead of AI serving only to maximize a score or follow rules set by a few people, it should be attentive to our existing communities and relationships, take responsibility, be competent, and respond to them.
Democracy with more bandwidth
This would also be a different idea of democracy, wouldn't it? You have worked on techno-communitarianism. Could you explain your vision of place, community, technology, and democracy? It seems different from the representative model we have.
I think it is entirely compatible with representative democracy. In Taiwan, we see democracy as a continuous process. People set the agenda together, discover needs together, explore together, and perhaps to some degree define common values together. But the development and delivery of particular policies still belong to professional, full-time career public servants and to members of parliament.
What we are doing is increasing the bandwidth of democracy, so each person can contribute more bits, while shortening the time it takes to receive feedback and close the loop — lowering the latency. We improve democracy's bandwidth and latency. At the end of the day, the principle remains the same: legitimacy should live where the consequences live, not at the headquarters of some technology company.
The unit of democracy is not isolated individuals clicking buttons, but people in real relationships. Most members of parliament, city councillors, and people who run for election already understand this: they all want to listen better to their constituents. Here, AI is scaffolding that holds the listening while it grows. We are not proposing to replace the listening relationship between you and your city councillor with a relationship between you and a city bot.
Techno-communitarianism
What is the concept of techno-communitarianism? And how did you and, most notably, Patrick Deneen team up? Some would call him a reactionary postliberal thinker, although that may not be a helpful starting point for discussing your collaboration. Could you explain the concept and the collaboration?
What you're referring to is a piece that I'm working on with Glen Weyl and Patrick Deneen, advocating for technologies that serve communities. I don't think it is particularly controversial to say that the past decade or so — as seen, for example, in The Anxious Generation and many other works — has shown a very high correlation between atomized, engagement-maximizing technologies and the withering of what we call civic muscle: the real relationships among citizens.
People have seen identity severed from actual relationships, reducing a person to a unique row in a database with some preference functions. Deliberation — what used to be deep conversations, town meetings, and so on — no longer happens in town halls, but on comment threads, where it becomes a performance that people mostly watch as others dunk on each other, instead of a thoughtful, neighbourly conversation. Productivity is hollowed of vocation and solidarity, and roots are traded for a homogenized promise of abundance, such as universal basic income. But many people say that it is not universal basic income that would give their lives meaning; it is the roots they can practise together in a cultural community.
I don't think these diagnoses are particularly controversial or belong to only one intellectual tradition. What we are advocating is restoring communal values without completely giving up the internet or digital technologies. We can use technology for communal sense-making, as with Pol.is and what we tried in vTaiwan. We can have communal civic digital infrastructure: people can collectively own compute, payments, identity, and data exchange. We can preserve communal integrity with privacy-enhancing technologies.
We can also maintain good provenance within and across communities, even translating between the provenance of biblical creation-care communities and that of climate-justice activists. Language models can help with that, so we do not have to give up language models. The point is that technology can strengthen bonding within communities and, crucially, bridging across communities. That allows deliberation to scale across many different communities without collapsing them into something easier for machines to count.
All this is to say that we should not change ourselves — align humans — merely so that we resemble data points that machines can count more easily. We should bring machines into our communities and mould them in ways that strengthen our bonding and bridging muscles.
Political inspiration: Taoism and conflict co-creation
Where else do you find political traditions that are relevant for thinking about technology or democracy today? Postliberalism, in Deneen's formulation, may be one: it is also about community and a more concrete, circumstantial definition of life — in his case, perhaps family and other structures of society that some would call less meritocratic and less free. Where do you take your inspiration as a political thinker? And how is that connected to what I have seen described as “conservative anarchism,” if that is a term you find relevant to your political thinking?
My inspiration comes from my Taoist background. Laozi says in the Tao Te Ching that the best leader is one of whom people say, “We did it ourselves.” That remains my North Star for technology: it should enable people's collective agency, belonging, care, and dignity, rather than play the role of a superintelligent or super-competent leader.
That is very Taoist. Taoism is at once about conserving — ensuring what is called effortless action, or conserving what is already there — and anarchistic, in the sense that it does not advocate coercive violence by anyone against anyone else. More than any other tradition, these two sides together inform my political ideas.
Another Taoist idea is Chōnghé (沖和), or conflict co-creation. Instead of seeing polarization, or even vehement disagreement across ideological divides, as a volcanic eruption to evacuate from, we can see it as geothermal energy. If you build an engine that turns it into productive tension, conflict becomes a source of lift. Taiwan sits between two tectonic plates that push against each other, producing three small earthquakes somewhere in Taiwan every day. The same pressure pushes the tip of Taiwan — Yushan — upward, skyward, by half a centimetre every year. I see polarization not as something to muffle or squash, but as energy to transform.
Finally, I have seen the left, the right, and everyone in between become collectively tired of peak slop in today's information environment. I would not call this particularly conservative, progressive, liberal, or communitarian. Whatever one's position, we cannot hold a conversation when the information environment has such a high PPM — polarization per minute. We suffocate in a fog of war that caricatures every position as its most extreme version, merely for dunking.
A shared, lower-PPM information environment can take surplus polarization and turn it into uncommon ground: a collective awareness of what we actually agree and disagree about, and then into common knowledge. That benefits everyone, regardless of ideological position.
Technodiversity, not technological essentialism
Building on this Taoist foundation, is there what one might call a Chinese view of technology?
Yuk Hui, the philosopher of technology, calls it technodiversity. It is close to the techno-communitarian idea that each culture and community can shape technology around what that community values and prioritizes. But I would be careful about describing it in an essentialist way.
I would not call the internet an American technology, even though it was initially funded by DARPA. I would not call the World Wide Web a European technology, even though it was literally a product of CERN. Those are places where technologies may come from. In the same way, I would not call digital democracy or the 6-Pack of Care a Taiwanese technology.
I can say that Taiwan's geopolitical and domestic traditions require us to reconcile a distrust of coercion in our somewhat Taoist spiritual needs with an interpersonal-responsibility frame that is very Confucian. This is also shaped by our civic history: Taiwan gained broadband access and democratized in much the same period, alongside personal computing, which empowered each person with their own compute. For my generation, the internet and democracy were never two separate things. They arrived as one movement.
Still, I would not call this a Taiwanese philosophy or model of technology. Taiwan is a particular context that gave rise to this fusion, this blending. It is not a model to cut and paste. It is an inspiration — a demonstration, a demo — from which people in other cultures can draw. It shows that the alternatives are not limited to technological surveillance in either the capitalist or the statist sense. There can also be an alternative that does not concentrate power at either pole.
To use Acemoglu's metaphor, the narrow corridor can be widened by broad listening. Taiwan chose this particular direction, but we do not claim it is the only way to widen the corridor. There may be thousands or millions of roads that widen each society's own narrow corridor in ways that are not Taiwan-specific. We only want to share ours as a demonstration.
When computing and democracy arrived together
That moment when broadband and democracy arrived together was 2014?
No, in fact it was earlier in Taiwan.
Earlier — around 2000? It wasn't the student movement?
Right. My generation in Taiwan is the last to remember martial law, which was lifted in 1987, when I was six years old. That coincided with personal computing. When we first voted directly for our president in 1996, web browsers were already popular, and the government already regarded universal internet access as approaching a human right.
When the Sunflower Movement happened in 2014, broadband access was becoming generally available. By that year, Taiwan already had more active Facebook accounts per capita than any other country. Everyone here was exposed to the high-PPM information environment — radical polarization and the attention-capturing economy, what I call antisocial media — perhaps more than people in many other democracies, simply because our conversations and engagement on the internet were already so saturated.
Taiwan under geopolitical pressure
Perhaps we should turn to Taiwanese geopolitics and then to your experience as a Taiwanese state official. How do you view Taiwan's geopolitical position? As you say, it is caught between two colliding tectonic plates — not only metaphorically, but in reality, with China as the large tectonic force and the United States in the background. Where is Taiwan's future? How do you see it?
It is upward and skyward — toward the sky.
Yeah.
The pressure is, of course, real and constant. The question is how we translate that tension into lift. By many counts, including that of the National Security Bureau, Taiwan faces two million — almost three million — probing attempts against our cyber infrastructure every day. I call it free red-teaming, free penetration testing from our very friendly neighbour. Most places have to pay for this service; we get it for free.
As I mentioned, our information environment has also experienced information manipulation, malicious AI swarms, and so on before many other democracies. Our response has not been to crumble into authoritarianism or create a Ministry of Truth. Instead, we invest in radical resilience: civic infrastructure, rapid public correction, and humour over rumour.
The idea is that, instead of panicking over earthquakes — which does not work when you are in Taiwan — you always prepare for the next one. Every pressure should make our infrastructure antifragile, even more resilient after each earthquake. We recognize that we are simply a place where the pressures facing all open societies arrive early, certainly before they reach many other democracies. We are a kind of early-warning station. What works here under very high pressure is worth knowing about everywhere, because the earthquake may reach you in a few years.
Translating, demonstrating, and connecting
How involved are you still in these processes, and how is that related to your time as Taiwan's digital minister?
I am still doing the same translational work, demonstration work, and connection work. That is my whole job description now. When I was minister, I also translated between our civil service — our career bureaucrats — and the civic hackers, the creative technologists. I won the Creative Bureaucracy Award, so I suppose I am a bureaucrat too. Being a creative bureaucrat means that I am a good translator between those two cultures, between policy language and coding language.
Now I focus more of my work overseas. I have travelled to 28 countries in the past couple of years and change time zone every two weeks on average. I translate between technologists in frontier AI labs and the communities where their systems land — communities that urgently need civic muscle because it was not replenished, but instead decimated by all the slop we have been discussing.
I demonstrate, as Taiwan has demonstrated, that alignment assemblies and Civic AI can set fair rules that make technology giants more like assistive connective tissue among people. Instead of sidestepping accountability for online fraud or information manipulation, they can feel the pain citizens feel and contribute to a better information environment.
And I connect people who would not otherwise be in the same room. In Oxford, for example, the Civic AI Conference on the 6-Pack of Care connects care ethicists on one side with AI safety and security researchers on the other. It also connects postliberal conservatives with anarchist thinkers and crypto-libertarians. They have more common ground than they realize.
Working around the world allows these demonstrations to travel much farther than ministerial connections — so-called track-one diplomacy — ever could. This is how I continue the work of a conservative anarchist, or a Taoist, as I prefer to call it, across time zones and spiritual and ideological traditions.
A messenger from tomorrow
So, as a translator, are you also a messenger from the future?
Well, it is almost 3 p.m. here, so in some small way I am in the future compared with where you are. But it is only slightly in the future, not the far future.
When I describe Taiwan's experiences in Washington or Berlin, I am not saying, “This is the far future — adopt it.” I am saying that, on the same day — or, all right, the next day, but no further away than tomorrow — there is proof from a place under maximum pressure that chose broad listening over surveillance, and it helped. What you build from that proof remains yours.
I carry these demonstrations to Germany's federal government through SPRIND, the Federal Agency for Disruptive Innovation, and Deutschland, was nervt? Deutschland, was geht? But this does not mean Germany adopts Taiwan's presidential hackathon or anything like that. Inspired by what we have done in Taiwan, Germany grows something for its own public sphere — but it is German, not a Taiwanese import. The organizers, planners, and people tilling the soil together are German.
We are answering questions that thinkers such as Jürgen Habermas have always asked, using new infrastructure. They remain the same discursive questions about democracy that generations have worked on together.
Optimism as a practice
One last question: where do you find your optimism and hope?
First of all, I always sleep for eight hours every day. That is where I find my optimism. If I face a particularly difficult challenge, I work overtime — meaning I sleep for perhaps ten hours that day — and I always end up with more optimism.
Optimism is not really a mood. It is an active practice. During the day, I take in conflicts, polarization, and disagreement without passing judgment on them. During sleep, I digest and metabolize them.
I also turn all my screens 80 percent grayscale, keeping only a little colour, so my dreams are more vivid than my screens. It helps my interpersonal relationships too: when I meet you, your face is more vivid than the face on my screen, so I am not distracted.
When I talk to AI systems, I instruct them to give me a brochure or an interactive page, without pretending to be my best friend by saying “I think,” “I feel,” or otherwise using first-person pronouns. Then I can share the result with my actual friends instead of cultivating synthetic intimacy.
I practise these very small civic muscles so that I can have good-quality, dedicated sleep. I always wake up feeling better. Conflicts, gaps, or ideological differences that seemed impossibly large go through some process in my dreams. I wake up with a fusion of those different horizons and think, “Actually, it is the same thing underneath.” Then we can turn polarization and conflict into co-creation again.
Webinar and essay logistics
Fully noted. Thank you. That was also a very meaningful statement, because my son, who is 16, always insists on both things. With sleep, he is always saying, “Dad, you don't sleep enough.” And gray — what is it called? Grayscale. He also says that you shouldn't read on your phone before going to sleep; it is destructive. After talking to you, I believe him more. But this adds another perspective. Thank you. That was a very meaningful conversation.
I'll send you the transcript. I suppose I could do it in ten seconds, but I'll look at it first. Have a good time there with your family. And I suppose we will see each other — I don't know whether you can make it on the 17th, whether that is also too early, or whether you would rather join on the 31st. It depends.
Grayscale. Yes, the 17th runs past midnight for me. So if I can still get eight hours of sleep, I may join the first part. Otherwise, I am happy to join when I am in a more humane time zone.
Yeah, yeah. And do you think you can send the first essay by the end of next week? Would that work?
Of course.
Good, good. Have a lovely afternoon, and thanks for your time. Talk to you soon. It was a very good conversation. Thank you.
Looking forward to working on the essay together. Live long and prosper. Ciao, ciao. Bye-bye.
Yes. You too. Bye-bye.
