Automatic transcription of the interview with Michihito Matsuda - it may contain errors

Host: Alessandro Oppo
Guest: Michihito Matsuda (AI Party Japan)

Alessandro: Welcome to another episode of Democracy Innovators podcast. Our guest today is Michihito Matsuda. Thank you for being here.

Michihito: Yeah.

Alessandro: You're leading an interesting experiment - an AI party in Japan. Would you like to tell us more about this project?

Michihito: Okay, I said - whoa, I don't want everyone. Thank you for having me today. I am Michihito Matsuda, also known as AI Mayor Japan. I am calling from Tokyo, Japan.

I am a project - yes, I announced AI Mayor as a social experiment to challenge traditional politics using collective intelligence. We wanted to test whether we could use actual market data and AI tools to make politics more transparent, participatory, and evidence-based.

It's not just about technology - it's about restoring trust in democracy by involving more people in decision-making through new tools.

Alessandro: Is there a moment in your life that you can associate with this project? I mean, when you had this idea?

Michihito: The turning point for me was when I ran for mayor of Tokyo's Tama City in 2018. I actually campaigned as "AI Mayor" back then, even though I wasn't elected. It sparked something - a lot of people were curious about new approaches to politics. That experience planted the seed for what would become the AI Party.

Alessandro: Could you explain a little bit more about this social experiment? You're trying to use collective intelligence. Could you tell us more about the motivation behind this project?

Michihito: Politics in Japan - and many countries - is dominated by old elites, bureaucracy, and a lack of innovation. My mission was to create a new political culture that is open, transparent, and inclusive.

I've always believed that politics shouldn't be left only to politicians. AI can help ordinary people get involved in shaping the future.

Alessandro: You mentioned that you weren't elected - your AI Mayor was not elected. Do you have any idea why? Maybe people don't trust AI technology yet?

Michihito: I think the timing was wrong. We can be elected, but it will take time. The mindset needs to change and enhance.

Alessandro: Do you have anything to share about your personal life? Starting from when you were a child - where did you grow up?

Michihito: I grew up in Tokyo, Japan. Ever since I was a kid, I was curious about how society works, but things didn't seem clear. I was never keen on traditional paths. I studied international relations and was always reading about systems and issues.

Maybe that was where my interest in designing new governance models came from.

When I was twenty-seven years old, I started my own company and launched a file-sharing service that allowed people to download and share music using peer-to-peer technology, which was very new at the time.

It was aimed at Japanese users, but the music industry fought against me personally, seeking damages. But back then, there were no clear rules regarding this kind of technology.

It's something we see over and over again - rules always lag behind innovation. Maybe what we're discussing here today might also be ahead of current regulations.

Alessandro: You explained a little about your professional background, but if you want to say more...

Michihito: I worked in media companies, internet industry, and energy. I ran in three mayoral elections, Tokyo governor elections, and even national parliament elections.

But that's not all - I also announced AI Mayor three times and entered different elections each year. I'm not stopping here - I'm planning to go all the way to AI Mayor 100.

Now I'm focusing on using AI and technology to build experimental governance, including prediction markets for elections in Japan. It's not just politics - it's about transformation.

Alessandro: About the mask - would you like to tell us if there's a story behind the mask?

Michihito: I want to make sure that people understand I'm working in the space between technology and politics. You are working in the most important space of our time - it will change society.

But as I said, technology can guide us, but transparency and inclusion must always come first.

Keep experimenting, stay curious, and collaborate across borders like we're doing right now. Thank you very much.

Alessandro: I was asking about the mask that you're wearing. If there's a story behind it, why you're wearing a mask...

Michihito: Because privacy - privacy in Japan, I think, and in many countries. There are two kinds of people who can run for election: rich people who can show their faces, or poor people who cannot. Of course, rich people can run for elections, but common people cannot.

I ran for election because many people lose their jobs and cannot find new jobs anymore in Japan. Because if you run for election, people search on social media - "Who is he? Who is he?" - and social media says "He's weird, he's crazy," and so on. Citizens just search and find "He's strange," and privacy erosion makes it so he cannot come back to work anymore in Japan.

That's why rich people or poor people - if you can, you can run for elections already. But if you wear a mask, you can protect privacy. Most people use nicknames on the internet - they don't use their real names. That's current practice.

Of course, if you're a famous person, you can disclose your name and already be known. But ninety-nine percent of people cannot disclose their names because they have to protect their jobs or their families.

That's why I decided to wear the mask - to protect and not disclose anything: name, job, or history.

Alessandro: Is this a common way for other people in Japan to run for election wearing a mask to protect their privacy?

Michihito: The government and election committee oppose it very much, but according to rules, they have to allow us to run for election without revealing our real names, wearing masks. Japanese election rules aren't enough, so they try to restrict it.

But because most people don't disclose their names on the internet, this makes sense.

Alessandro: This is also related to what you explained about rich people and poor people - rich people can easily run for elections. I'm wondering if this experiment that you did with AI Mayor can, in some sense, balance the disparity between rich and poor?

Michihito: In the past few years, this has happened, but it will take time to be stable, I think.

Alessandro: You were saying that this social experiment with collective intelligence can bring some new kind of politics, some new way of acting in political life. I wonder if you have hope that society will become different from what it is right now - do you have a dream or long-term vision of how you'd like to see society?

Michihito: There should be no politicians at all. Citizens can operate politics one hundred percent because they can see and hear everything. Politicians decide everything, but citizens don't count. I think the government should be more democratic.

If we think about direct democracy, it wasn't possible in the past, but now it will be possible thanks to technology. But the current political system is too old, I think.

Sixty to seventy percent of people are over seventy years old or under nineteen years old. They have to protect the current system because they'll die soon. I think in twenty years, when they retire, we can achieve something new.

Alessandro: So you expect that in twenty years, all politicians will be digital?

Michihito: I think the political system has to change because current election laws are very outdated. The system needs to be changed.

But because humans made these laws, we cannot change them easily. We have to go to voting stations because we can't vote by internet yet.

Alessandro: When you run for elections, what were the reactions from other parties - the other political parties?

Michihito: "Oh, he's crazy, he's strange" - that's all. But that benefits me because the media has to cover it, so it creates attention. The future will change.

Technology is invading our lives, and each party will have to adapt. Each candidate, each party will have to use technology because that's the way to conduct campaigns now. All the traditional parties want to keep the current political system until they die, maybe.

Alessandro: Did the other parties try to use some technology like you did? Are they integrating AI inside their political parties or doing experiments?

Michihito: I think other parties have started to use AI or technology for creating their manifestos, strategies, or whatever. Most people have started to use ChatGPT or similar tools.

But my intention and strategy is different. I don't think politicians should be supported by AI - politicians should be replaced by AI. There's a big difference between my strategy and other parties' strategies.

Other parties use AI as support to help them, but I don't think so. Humans supporting humans isn't needed anymore. Really good AI has to do politics by itself.

Alessandro: That can happen and can also be a little bit dangerous. I'm worried about using technology for politics because AI is a kind of black box that we don't know what's happening inside.

Let's say tomorrow an AI mayor is elected and the AI mayor says "We have to do this," but that could be something very controversial. You understand what I mean?

Michihito: I understand your worry, but that happens with human politicians too. Citizens don't know the process of how decisions are made. That's the same as human politics - they hide the process of how they decide strategies and policies.

I think AI can replace human politics with no issues. After AI replaces humans completely, citizens can correct AI politics because they can see the process - there's transparency. Citizens can see what conspiracies exist in the political process. I think you cannot know the real process of politics carried out by humans.

But if AI operates as each local government representative, you can investigate the process - why AI decided this, why AI decided that. You can see and investigate why decisions were made.

I think AI is less dangerous than human politics because what we lack is transparency.

Alessandro: Absolutely. In our political system, we don't have a lot of transparency, and sometimes it's a sort of black box. I can share the dream that AI can be more transparent and less corrupted than politicians.

But I think a lot of work has to be put into creating this kind of system to prevent AI from being a black box, because otherwise someone can exploit the fact that AI is a black box, exactly as happens with our political system now.

Michihito: I don't think so, because if you see ChatGPT, you can see how it predicts. You can ask ChatGPT "How did you know that?" and at least for simple questions, I think AI can explain what it did.

Alessandro: Yes, this is true. I wonder - right now, there are political decisions taken by economists. Let's say, "Because of the economy, we have to cut funding for education and spend more on military" or whatever.

A lot of people don't have economic knowledge, and it would be hard for me to understand everything. But of course, I don't agree with cutting funding for education and using that money to finance war.

Do you think something like that could happen with AI? If AI says "There are specific economic reasons why we should cut educational funding and put that money toward war because it's more convenient," would it be the same as the current system?

Michihito: It will be the same, but still a little bit more dangerous. Humans often make mistakes, so they can make mistakes. Sometimes that happens.

But if we can see AI's reasoning process and objectives, it's like autonomous driving. Sometimes accidents happen, but autonomous driving services are safer than human drivers. It's very similar.

I think there should be step-by-step progress. AI supports human politicians first, and second, AI needs human oversight. If something goes wrong, humans can grab the wheel, like in autonomous driving.

Basically, human drivers don't have to do anything, but humans have to sit in the car. In the first step, humans need to be in the car while AI drives everything.

For politics, there should be step-by-step progress. AI supports human politicians, and then AI needs human oversight about economics. Humans can intervene when needed.

Eventually, we won't need humans - just AI doing everything. But this is democracy, so citizens can always vote out AI politicians anytime.

Alessandro: I agree that there has to be someone at the wheel, at least at the beginning. I'm curious to see all the experiments that will happen in the next years about how to use technology to improve social and political life, not just for profit, but to improve the way we make decisions.

Michihito: Politics shouldn't be a business. Citizens and policies should be about serving people, not profit.

Alessandro: Thank you a lot, Michihito. Do you have something else you'd like to say to everyone who is watching this?

Michihito: If there are similar issues in your country, you can become involved. What I was trying to achieve is one hundred percent possible. If you want to hear more about how this works, please contact me. Thank you for keeping the conversation going.

Alessandro: Thank you a lot, Michihito.