<strong> Expand the corridor </strong> between domination and chaos—even if it starts narrow, expand it.

Think not only about <strong> fast </strong> or <strong> fair </strong> responses, but also <strong> fun </strong> responses. <strong> Humor over rumor </strong> is a consistent lesson. In a joyful mood, people are more creative in finding ideas others can live with. Starting only with stress and anger narrows a ...

<strong> Sleep more. Take care of yourself. </strong>

We had petitions— <strong> 8,000 </strong> for changing Taiwan’s <strong> time zone </strong> to <strong> UTC+9 </strong> (to show sovereignty), and <strong> 8,000 </strong> to keep <strong> UTC+8 </strong> ; it was a standoff between the two sides. We convened both and proposed <strong> dozens of ideas </strong> : instead ...

Start many <strong> swarms of dialogues </strong> about <strong> adjacent </strong> issues. In Taiwan we didn’t tackle marriage equality first. We tackled <strong> how single mothers can raise children without discrimination </strong> , <strong> menstruation </strong> , and other adjacent but important issues. With thousands of such movements, one is ...

That’s the <strong> traditional theory of change </strong> . I advocate <strong> quantum tunneling theory of change </strong> : <strong> don’t </strong> try to scale up your demand until you have <strong> scaled across </strong> sufficiently.

<strong> Scale across; scale in before you scale up and scale out. </strong>

You need to prepare at the <strong> meso level </strong> —thousands up to half a million people—to exercise civic muscle. Don’t go to the gym and lift heavy stuff first; you’ll strain yourself. Once you keep exercising and urgency comes, you can scale across many polities.

In <strong> Japan </strong> , something similar is happening. <strong> Takahiro Anno </strong> , newly elected, ran on broad listening, de‑polarization, bridge‑building, <strong> plurality </strong> with the <strong> Team Mirai </strong> , the <strong> Future Party </strong> , running on <strong> AI‑enabled assistive deliberation </strong> ; they also won.

For a while Taiwan was the largest polity trying this because of pressure. Now <strong> California </strong> adopted the same system. I worked with Newsom’s team for two years— <strong> infrastructure, deliberation, digital tools </strong> . We chose an uncontroversial issue: <strong> teenage use of social media </strong> . We ...

If you scale down to <strong> 100 or 1,000 people </strong> , most places learn better ways of meeting. There’s little pushback introducing digital democracy at <strong> small scales </strong> . Pushback appears at <strong> 1–10 million </strong> people, where existing stakeholders has interests in blocking new infrastructure.

Taiwan used to be the <strong> largest polity </strong> where civic infrastructure became public infrastructure. Smaller polities have many examples— <strong> Reykjavik </strong> , and most <strong> city‑level citizen‑initiative systems </strong> far exceed Taiwan’s. We pulled municipal innovations into a <strong> national </strong> infrastructure.

I’m a <strong> very competitive sleeper </strong> . I just had more than <strong> 8 hours </strong> . When things get tough, I sleep <strong> nine or ten hours </strong> ; that restores me.

That makes the uncommon ground <strong> rarer </strong> , but also more <strong> desirable </strong> .

Most people are <strong> tired of polarization </strong> . On the left and right in the U.S., nobody likes it. The potential for <strong> uncommon ground </strong> is high— <strong> product‑market fit </strong> for community notes and horizontal AI.

I’m optimistic because <strong> everybody sees the same thing </strong> now. No democracy is an island; all are connected.

Now, after last year when <strong> all ruling parties worldwide lost seats </strong> , people see that vertical institutions—journalists, academics, ministers— <strong> lost relevance </strong> . People find belonging in amplified social media; even <strong> robots dressed as people </strong> feel closer. In Japan, this election saw a viral YouTube ...

Yes, because we have <strong> anti‑fragile infrastructure </strong> to meet the moment. Ten years ago, when Taiwan faced epistemic polarization, we were top of the world in receiving it for 12 years; it was lonely. Few understood <strong> epistemic security </strong> as part of cybersecurity.

We focus on <strong> machine ethics </strong> : training AI systems not to enslave, lobotomize, or maximize profit/engagement, but to <strong> imbue civic care </strong> . That’s my Oxford research topic: 6pack.care — How can you align AI <strong> horizontally </strong> so they maximize the health of <strong> inter‑human </strong> ...

We need to make care appealing— <strong> a care gym </strong> —six‑packs of care. Make care bro‑friendly.

Make care <strong> outward </strong> —not just about individuals’ emotions but <strong> society’s emotions </strong> . If a part of society feels really bad, it’s not just for psychiatrists or caregivers; the <strong> democratic system </strong> needs to shift so measures evolve into something they can live with. <strong> Civic ...

If we refocus on <strong> care </strong> , we’re attentive to actual needs and compete on <strong> how quickly and capably </strong> we address them—without self‑aggrandizing or creating demands based on addiction. Pivot back to <strong> capacity to care </strong> .

Focus on our <strong> capacity to care </strong> . For too long, the digital realm is viewed utilitarianly: jobs, GDP, growth, data centers. Once a civilization is established in human rights and democracy—it’s a <strong> full chicken </strong> no longer just out of the egg. In the egg you grow; ...

Of course. Since <em> Coded Bias </em> —like <em> The Social Dilemma </em> —policymakers aren’t blind. People want <strong> working programs </strong> , no matter how small. In the <strong> “pocket of good,” </strong> in Taiwan we sent <strong> 200,000 messages </strong> to people; <strong> 450 people </strong> came together ...

At some point, people will say: if you use a <strong> culturally non‑aligned </strong> version for certain jobs or activities… Most of this is in the <strong> EU AI Act </strong> ; they’re now defining enforcement rules. There’s huge industry lobbying, but the Act contains most of the language I’m ...

Many countries are doing <strong> public AI / sovereign AI </strong> —local versions more culturally attuned. Countries, including Taiwan, are preparing curricula. It’s like an <strong> immigration test </strong> : you expect a new immigrant to understand how society functions. For <strong> robot immigrants </strong> taking jobs, people expect understanding ...

Creating a <strong> curriculum for AI systems </strong> is crucial.

Algorithms repeat <strong> bias </strong> in data. With assistive AI, we <strong> eliminate garbage and bias </strong> —like preparing a <strong> curriculum </strong> . Early interpretations are key. Language matters. Ask about the ultimate ending of AI–human relations in English and you get dystopian Terminators; ask in Japanese and you ...

Yes. That’s the difference from garbage AI producing wrong information and falsifying. Tell it that vaccines contain 5G chips and it might say, “You’re the only one seeing through the veil; don’t believe police or journalists.” That’s <strong> epistemic harm </strong> .

We use AI to find <strong> promising bridging statements </strong> between clusters and rewrite them so everyone can understand in their native language and culture— <strong> summarization, publication, cultural translation </strong> . The AI serves the <strong> community’s needs </strong> and then <strong> self‑effaces </strong> when needs are fulfilled.

There’s <strong> horizontal alignment </strong> : the mediator aligns not to my will or yours, but to the <strong> relational health </strong> of the conversation. Many ways to align models to <strong> relational health </strong> —optimizing the health between people who previously disagreed.

I distinguish <strong> assistive intelligence </strong> and <strong> authoritarian intelligence </strong> . Authoritarian, <strong> vertically aligned AI </strong> subjects the model and outputs to the will of one controller. It disempowers people; machine judgment replaces human judgment, aligned vertically to the controller’s principles.

Signal and Proton serve <strong> needs </strong> ; but many others hack <strong> preferences </strong> —like <strong> junk food </strong> . No one needs junk food, but exposure shapes preferences against our <strong> meta‑preferences </strong> (meaningful connections). Free software doesn’t do that. <strong> Forced interoperability </strong> creates the offering.

To break monopoly, don’t scare people about data; just <strong> switch to where they are in control </strong> . If they don’t lose friends, <strong> why not </strong> ?

They’re there because <strong> everybody else is there </strong> . The first mover suffers <strong> FOMO </strong> . It’s like luxury brands: if everyone carries LV or Hermès, you feel compelled to invest; you don’t get much. If everyone switches to non‑luxury as the norm, everyone is better off. With ...

It’s a <strong> product market trap </strong> issue. A study showed: an average undergrad in the U.S. using TikTok would need <strong> $60/month </strong> to switch off TikTok. But if there’s a <strong> magic button </strong> that moves <strong> everyone they know </strong> off TikTok, they’re willing to <strong> pay ...

Once a platform sells your data, you <strong> switch to one that doesn’t </strong> —and <strong> keep all your friends, content, and followers </strong> .

The point isn’t blaming a CEO. If Bluesky users fork into <strong> Blacksky </strong> , Bluesky must <strong> keep interoperating </strong> with Blacksky—which is already happening. Law can require <strong> on‑ramps and off‑ramps </strong> —full liability for networks that don’t interoperate. It’s technically feasible; privacy concerns are now solvable. Bluesky ...

Without interoperability, if <strong> Jay Graber </strong> gains 70% of the market, investors might force a sale. When Twitter was the underdog, they were good.

A simple fix: adopt the EU <strong> Digital Markets Act </strong> approach— <strong> portability </strong> between <strong> instant‑messaging </strong> systems. The EU already has this. The <strong> State of Utah </strong> went further: starting <strong> next July </strong> , when you switch from TikTok to Instagram or from Twitter to ...

Later, with telecoms, people wanted to switch for better reception but would lose their <strong> phone number </strong> . The state mandated <strong> number portability </strong> .

When I was a child, <strong> ATMs </strong> only served cash to cards of the <strong> same bank </strong> . New banks couldn’t compete; big banks cornered street corners. The state intervened—not by nationalizing—but by mandating <strong> interbank protocols </strong> . A small bank’s card could withdraw from all ATMs.

Market mechanisms address next‑quarter pain points. Truly non‑market issues— <strong> banning plastic straws for bubble‑tea takeouts </strong> , or <strong> a menstruation museum </strong> to remove taboos—are started by those caring about the long‑term future. Participation infrastructure brings <strong> ministerial conversation </strong> to such issues, amplifying niche agendas to a ...

In Taiwan you can have <strong> citizen initiatives </strong> ; some now have <strong> hundreds of thousands of signatures online </strong> . Demographics show <strong> 17‑year‑olds and 70‑year‑olds </strong> are most active: they have more time and care more about <strong> intergenerational unity </strong> than people focused on the next ...

If it <strong> affects </strong> their child and there’s <strong> something they can do </strong> , they will be motivated. The <strong> agenda must be set by people closest to the pain </strong> .

Yes. It produces uncommon grounds that are bridging, then presents the <strong> best argument </strong> of the remaining divide— <strong> balancing </strong> —which is how newspapers are supposed to work. The <strong> Hutchins Commission </strong> : front page—what everyone can agree happened; second page—the best arguments of main ideologies. <strong> ...

At the end, you need <strong> consent from all smaller groups </strong> . A 7% group can veto the thing that hurts them most. A 21% coalition can veto 21% of harmful proposals. We find and amplify <strong> uncommon ground </strong> , orthogonal to majority opinion—the <strong> opposite </strong> of ...

We show how preferences cluster and offer <strong> bridging bonuses </strong> for proposing rarely discussed <strong> uncommon ground </strong> with nearby groups. People who propose novel statements gain the bonus. Finally, we get <strong> bridging narratives </strong> ; four groups may coalesce into two; then another round of bridging.

It’s a <strong> function of aggregation </strong> . An 80% block is just majoritarian aggregation. With <strong> proportional </strong> or <strong> bridging algorithms </strong> , we reveal clusters—say four stakeholder groups: one 80%, others 7%, 7%, etc. In a <strong> bridging system </strong> , the 80% still must bridge to ...

A <strong> high‑resolution democracy </strong> , enabled by digital technologies, figures out bundles of issues so everybody is a little happier and nobody very unhappy, instead of low‑resolution 51% majoritarian wins.