Thank you, Professor. Good local time, everyone. Very happy to be here. It is my immense honor and privilege to deliver this talk at LSE IDEAS. And today I'll share with you the development and strategy of moda, the Ministry of Digital Affairs.

Now, the moda is very new. It's like a startup. We were founded last August. And it is one of the very few ministries in the national government that combines democratic participation in the ministry proper, progress in the administration of digital industries, and safety in the administration of cybersecurity in the same ministry. So, participation, progress, and safety are merged into a single ministry. And this is partly because of our experience encountering the pandemic. Basically, the task force is the one that is responsible for the development of the moda. And we are working with the Ministry of Digital Affairs, which was four different ministries and departments merged into a rapid response ministry. And it has been going on quite well with digital resilience for all as our main call to action.

Now today, I will talk a little bit about our international efforts around digital democracy. I'll also discuss various topics related to the internet, that is tomorrow that is based on this idea of assistive intelligence or assistive AI, open-source AI. I'll also talk a little bit about the progress side with our Taiwan employment gold card in the digital field. And then, I'll then bring back the progress and safety and participation together in a summary.

Now, as I mentioned, moda, the logo can be read as ministry of zero-one-zero-one, is about connecting the digital democratic networks across all sectors, particularly in light of ever-evolving information technology landscape. And so, we don’t make compromises or trade-offs between safety and progress for each and every emerging technology. We aim, through democratic participation, to find what we call plurality or collaborative diversity, ways to interoperate for both the safety and the progress concerns, promoting co-creation from the tensions, from the conflicts. For example, during the pandemic, we have privacy concerns on one side, the public health on the other. And then we overcome these using zero-knowledge cryptography and many advanced methods. And so, we would continue to do so, turning conflicts into co-creation.

Now, as I mentioned, we're rather like a startup. We continuously adjust ourselves and serve as a sandbox, really, within the government. Any of the latest technologies, I mentioned zero-knowledge, zero-trust, FidO, and so on; we try it ourselves within our ministries and the two departments. And because we have those three different values, participation, safety, and progress, within our ministry proper, we can find in the internal regulations and code that turns these emerging technologies in a way that satisfies all three values. And then, we share it with our democratic partners and with all the sectors in our society.

Now, the idea of international digital democracy dialogue is based on the three different resilience. We have the emergency resilience, which is about responding effectively to various crises, such as cyber-attacks and natural disasters. So, to counter emerging threats, including the interactive deepfakes interfering with democratic processes now, we're consolidating the public and private cybersecurity capabilities. The N-CERT and the TWCERT/CC are merging into a very quick response team that implements zero-trust principles.

And in light of the subsea cable's fragility, two of them were cut earlier this year, connecting Matsu islands and Taiwan proper were also emerging satellite operators in our diverse network backup communications to ensure seamless operation of critical infrastructure. And we also engage the decentralized civic infrastructure community (web3), using IPFS or other emerging technologies to ensure that anyone, anywhere can donate connectivity to back us up, effectively defending against cyber-attacks.

So, as you can see, what we have done to link Taiwan to the world in the digital field is to take the kind of emergency response, societal resilience, and also the security contributions, and then foster cross-border collaboration. It includes participating in international interactions regarding digital development policy.

I was just attending a month ago the UN IGF in Kyoto, where we shared with the Declaration for the Future of the Internet Partners our ways to increase not just communication, but societal resilience, and also how to govern AI together using what we call alignment assemblies, which I will explore in some later slides.

And we've been actively engaging in various international events with democratic partners from 27 countries, participating in over 87 physical and virtual events. And so, the idea is to establish this value-based, long-term and close collaborative relationships based on the idea of public code.

Indeed, if you go to our website, moda.gov.tw, on either ipns:// or https://, you will see something that looks very much like gov.uk. And this is because we embrace public code, and we share the digital public infrastructures that is not just a design system, but also the cybersecurity audited forms and authentication and so on, the very basic layers ultimately strengthening digital resilience across the democratic world. And so, now, I'll also talk a little bit about our own international initiatives.

Right. So, here, what we have is a very quick timeline about our engagement to the .tw and .org, .net, .com, and some communities. In the early days of the internet, pioneers like John Postel and Vint Cerf were strong advocates for open-source software, or public code nowadays we call them, and a free, inclusive internet. So the TWNIC played a key role in allocating internet resources during this experimental project phase. And the internet's early development was shaped by those open and inclusive principles, which were crucial to the success of the internet.

Now, with the growing importance of the internet, ICANN was established in 1998 as a global community to adopt that multi-stakeholder model for internet governance. And so, TWNIC became a legal entity around the turn of this century, and we have been governing internet-related things even before the founding of the Ministry of Digital Affairs through multi-stakeholder models. And so, as we know, the United Nations organized the WSIS, the World Summit on the Information Society, and the IGF in 2005, which was a significant milestone in the global promotion of internet governance development.

Now, the multi-stakeholder governance model is really quite crucial to the Taiwan model of governance. Specifically, we sincerely believe that it is the public sector's job to amplify the norms that is set by the civil society, not through any top-down means to impose the norms on civil society. And this is how Taiwan became, according to the Freedom House, the top place in Asia, I think, in Asia Pacific, when it comes to internet freedoms.

And so, the idea, very simply put, is that any and all internet-related governance issues in Taiwan, including the DNS rules, the DNS RPZ rules, when it comes to managing the emerging threats on the internet, all of this are done with public consultation, and as I mentioned, democratic, participation in mind, allowing anyone in Taiwan with an email address to participate in the governance in technology development and service innovation actively. And this also enables global internet users to see our policies more transparently, promoting the essence of an open and inclusive internet.

Now, the “Public Money? Public Code!” initiative, which was started, I believe, by the Free Software Foundation Europe since 2017, advocates that system progress established by the government paid by tax-payer money should be fully displayed with clear guidelines and specifications. So, the ‘Code’ here refers to not just the source code but also the code of internet regulations, the code of ethics, the code of safety and unbiased use of emerging technologies and so on. So, the idea is that with the spirit of the internet multi-stakeholder’s openness and the efforts of the open-source civic tech community over the past decade, Taiwan spearheaded this concept of public code as public infrastructure.

As of this year, we successfully argued to the cabinet that the code, including the standards and data infrastructure and so on, are just like the bridges and highways. They are public constructions eligible for public infrastructure budget and we now plan the institution of public code, soon to be online, at code.gov.tw and survey the potential system to overhaul our procurement system to manage the source code collectively as a shared pool across not just ministry and municipalities but also again like the Estonian X Road with our democratic partners.

We hope that the systems and websites developed this way will be regarded as public goods and indeed public infrastructures that is open to all democratic partners to further ignite innovation. And so, technologists and activists, instead of just forking the government — taking the government code and running a local copy — we can also merge back all the contributions and building everything with the real needs of the people.

Now, based on our recent achievements in adopting the CC0 or public domain for our own websites, for our own initiatives and working with municipalities, I sincerely believe that next year we will see a lot more digital public goods and infrastructure coming out of Taiwan, deploying a, as I mentioned decentralized civic infrastructure alongside a resilient public infrastructure and eventually merge those two different communities and code bases into a shared infrastructure, so that even if a certain part is attacked through cyber security attacks, phishing attacks, and so on, the security and availability of this cross-sectoral multi-stakeholder system can still be maintained through the joint defense across sectors, again, in a way that is multi-stakeholder.

Now, we're also working with our own ICT industry players, the system integrators, so that the public can directly provide modification suggestions to the relevant agencies, so that imperfect parts do not need to wait for the next procurement cycle, but we can jointly create inclusive services for all communities. Case in point is tuning the AI systems to fit the more than 20 national languages in Taiwan, including indigenous nations, so that we can shape this digital infrastructure with the people, not just for the people.

Now that I mentioned AI, I think the world is currently having a very vigorous discussion on how we can prevent falling into an abyss of abuse, exploitation, manipulation by bad actors, working for those who seek to extinguish freedoms, but still harness the power of AI for democracy, with democracy. And so, as one of the earlier signatories of the AI risk statement, I'm quite proud to stand with a large group of leading experts and public figures, who recognize the serious challenges, and so we share the same viewpoint, that is to say, we need to mitigate the risk, not by a handful of experts, but rather by working with the people to discover the ongoing societal impacts, with evaluations and mitigations, in a way that is truly democratic.

And so, it's essential to consider, of course, the potential short-term harms, like interactive deepfakes, which has the potential to unravel the trust that binds us together. But then, it’s equally important to apply AI so that we can coordinate with our cross-border partners to enhance information integrity, to build actual work examples of people working together so we can counter even greater risks that is in the horizon.

And so, instead of just a few board members of frontier AI labs governing this technology, we think that AI should be designed in such a way that it can take real-time feedback from all people and use that feedback to self-tune, to adjust and learn. So, we must establish an open, fair, and transparent mechanism for AI governance and usage to ensure that everybody has equal access and the opportunity to tune AI to its liking, to their liking. So, how do we achieve this?

So, in Taiwan, we're no strangers to thinking outside the box when it comes to building rough consensus and running code. And so, the public can now, in Taiwan, participate in policy making. Also, around our own AI models, our science ministry is training the TAIDE, the trustworthy AI dialogue engine with the input, constitutional input of the people's constitution of people's online and offline participation. And the idea here is that people can participate in the AI model, people can simply specify what they expect of the AI model, and then through an assistive deliberation, we can merge all these different viewpoints into bridge-making narratives that can then be used to tune AI in a way that fits the societal norms. It centers on equitable access as well as democratic participation.

And so, actually, we launched with the collective intelligence project, the Alignment Assemblies. And the reports from open AI with Alignment Assemblies and Anthropic has all been published, so I encourage you to take a look. Essentially, it's very encouraging. Anthropic showed that the people's constitution of AI resulted in a model that is equally capable, but even more fair than the default model that is designed by full-time researchers and experts on artificial intelligence.

So, this is how it works in action. You can see the people's ideas, just like constellations in the sky where people just speak in a deliberative workshop. And the system called Talk to the City can infuse all the transcripts, all the videos and so on, and automatically portray them as clusters, which then turns into avatars that you can then have a real-time conversation.

So, it's like an executive summary that talks with you. And so, you can present new situations, new scenarios and so on, and have a brainstorming session long after the workshop is over, and also enables automatic discovery of bridging narratives that connects the democratic participation, safety and progress arguments, so that people can quickly laser focus on the common understandings and the common values instead of spending a lot more calories on the one or two polarizing issues. So, this is like a social media that is actually prosocial, not antisocial.

And so, because of our investment on politics and talk to the city, as I mentioned, public code and open source, you're welcome to just work these on GitHub and also start your own collective intelligence project that can incrementally tune in AI.

So, upon participation, I think the data on the deliberation serves two purposes. First, it helps the local communities to tune, fine-tune, or distill their own models that is open-source AI, or decentralization. And also on the national level, it helps the deliberation to inform our policies on, for example, it’s after the Ideathon that we discovered that the people of Taiwan expect the public sector to lead a pioneering role to use generative AI in our own public sector everyday use, and then demonstrate, as I mentioned, through public code, how to use it in a way that can correct the biases and so on in real time and avoiding the cybersecurity and privacy concerns by leading with edge AI that doesn't have as much as concerns of uploading all the private info into one or two cloud operators. So, we expect a Cambrian explosion of experiments using this decentralized deliberative process, incorporating collective intelligence into technological development.

So, I think through this effort, we run into a lot of the teams that starts from a foreign country, but gets involved in the policy-making and democratic innovation in Taiwan. And so, Taiwan has always been an important base for the digital world. And in order to invite more partners from the open-source world to continue to moving forward with us, we made a very special residency program. Basically, there is no need for a proof of salary or a proof of diploma. All we need is a proof of eight years or more of opensource contributions.

And anyone who contributed to Wikipedia, to GitHub, to any of those Web3 publicly available ledger and so on are automatically then eligible for this all-in-one visa, which is not just resident visa that covers healthcare, but also covers for the spouse and children and so on. And it's good for three years. And if you renew once on the fifth year, we also invite you to hold a Taiwanese passport without giving up your original passport. So, this is a strategy of inviting open-source contributors to become also Taiwanese.

Yes. So, here we see actually the eight years of professional experience, including open-source contributions, is one of the eligibilities. But there's also, of course, eligibilities for academic, for salary, and other relative expertise.

So, in closing, I would like to share a quote from the Declaration for the Future of the Internet that holds a special place in my heart. And I quote, we believe we should meet these challenges by working towards a shared vision for the future of the Internet that recommits governments and relevant authorities to defending human rights and fostering equitable economic prosperity. And so, our unwavering dedication towards this goal is based on the principles of not making tradeoffs between participation, progress, and safety. This approach will help us collaborate even closer with our democratic partners to build and strengthen alliances in the future. No democracy is an isle, not even Taiwan, and we must work together to defend our cherished values. Thank you for your kind attention, and I look forward to the conversation. Thank you. Live long and prosper.

Thank you, very, very good questions. Indeed, turning conflicts into co-creation makes a lot of sense in Taiwan because we're literally in the place where the Eurasian plate on one side and the Philippine Sea plate on the other, bumps into each other all the time. There's on average three felt, mostly small earthquakes like every day in Taiwan. And this is how the tip of Taiwan, Yushan, or Mount Jade, grows by half a centimeter or so every year. So, this earthquake makes us grow.

And we see similar dynamics playing out, well, in the place pioneering public code, Estonia. When I visited Lithuania… I'm a Lithuanian e-resident, by the way, and Lithuania has a very similar motivation to do so. I’m also a digital resident of Palau, among other things. And so, there are ways for us to turn this conflict into co-creation by those emerging pressing needs, but I would argue that there are societal scale risks that are truly global. For example, the Taiwan model of the mask and washing hands and zero-knowledge, no contact tracing and so on gained prominence in the pandemic times because the virus knows no borders, and so is the AI deepfakes information manipulation risks and so on.

So, and I see on the Q&A, somebody asked, why did we not put climate crisis at an extinction risk? It's because it's firmly on the top of the ladder. We're just comparing the second tier in that statement, at least to me. And so, climate definitely is another one.

So, I would say that although that Taiwan is particularly in a place where research of this plurality or collaborative diversity happens, the application of this research makes total sense when we have common threats, societal scale threats across the globe. And at this time, it is, of course, AI risks, climate, information manipulation and many other things.

And the second question about scale. So, our ministry proper is around 600 people. And indeed, I run the ministry in a thoroughly co-creative way. We not just use collaborative documents, but actually anyone can sign a petition pseudonymously in our ministry. As long as they get 100 votes out of 600, I respond in my monthly live stream to whatever ideas that people bring. And we have very systemic ways to get those democratic inputs.

Now, the size of around 600 is at the kind of very edge of what this kind of fitting everybody's positions in the single room, in a single facilitator's head. Any more amounts than that, then we're left with just structured representation of these, which is why I think technologies like Talk to the City or Polis is so important because it let us retain the same fidelity of information to keep the same sort of nuance without losing the ability to scale.

So, this kind of nuanced conversation became really possible now, with assistive intelligence and language models because the human brain simply cannot hold 1000 different positions, which is why I think the idea of plurality, which is increasing collaboration while maintaining diversity, or increasing diversity while maintaining the degree of collaboration, is so useful. I'm actually co-writing a book on this very topic on plurality.net. So, check it out.

And finally, we were able to get the political buy-in for the gold card. I think because if people are going to cause trouble on the cyber domain, they're going to cause it anyway, whether they have a residency visa or not. So, having them physically be in Taiwan actually makes it easier to track them down if they're going to be cyber criminals or troublemakers, right? So, I think in the digital world, there's really only proximity of values. There is no natural geographic distance.

And so, by saying that we offer this kind of digital residency or as in other countries, e-residency and things like that, it doesn't automatically guarantee the voting rights or citizenship rights guaranteed. Also, Taiwanese passport goes very far. But otherwise, I think the other e-residency programs can coexist by showing that actually they don't enjoy the same physical rights. They just enjoy the same digital rights.

Yeah, a great question.

Well, everybody thinks of Ukraine when people think of low-earth orbit satellites as an alternative infrastructure. But Ukraine also showed that placing too much reliance on a single provider has its repercussions. And so, in Taiwan, we believe in a plurality of providers. For public cloud, we already work with the local resilience computing centers of Google and Microsoft, and hopefully some Amazon as well. That means that when the subsea cables are cut, as in when not if, around Taiwan, then those local resilience routings can keep most of the video conferencing and so on work if both sides are domestic. And in terms of satellite connectivity, we work with both geosynchronous providers, but also mid-earth providers like SES, and also low-earth providers like OneWeb.

And again, this provides full flexibility. And by not putting all the eggs into one basket, we ensure that there is some bandwidth left for international respondents to communicate to the world, whatever happens. And also, because we are very strong in domestic 5G and O-RAN development and deployment, we're now successfully demonstrated prototypes of a portable, just on a truck. just on a truck and now on a car and soon on a motorcycle, 5G, portable network, so that not only can we receive the satellite signals, it can be relayed to just ordinary people's phones that still associates with what we call disaster roaming, so that no matter which telco you use, when time comes that you really need access to international bandwidths; for example, if you're international correspondents from a major journalism outlet, you're able to communicate through wireless teleconferencing. Then, whatever your original telco was, you now have the ability to livecast to the world.

Before end of next year, we're going to have 700 of those points around Taiwan, and many of them will be portable.

This is a great question.

In Taiwan, we believe in, as I mentioned, local resilience, which is quite different from data localization. Data localization is when the data that's collected only stays here. Local resilience means that at least a copy is here, so that when the subsea cables are cut, the compute and the routing and so on should still work. And so, including our backups of our important core information systems.

I think the idea of data embassy was pioneered by Estonia. They started with an embassy in Luxembourg, if I remember correctly, which keeps the data backed up, even if something happens to Estonia. So similarly, we're doing the same for the core data in public clouds. But that creates massive concerns about how do we ensure that it is zero-knowledge in nature, that no cloud providers can look at the personal data and so on.

So, which is why we consider privacy enhancing technologies a public infrastructure. Only by investing in the latest of encryption technologies, including homomorphic encryption and so on, can we ensure that the data storage is decoupled from computation. That is to say, computation, when we work with cloud provider, we should adopt this theory of trust principle.

So, think of like a chat bot. You send the model an encrypted version of the question you're going to ask it. The model operator knows nothing about your question. But through homomorphic encryption, the model can send back its response to you anyway. And then, only you can decrypt its response. So, it is advanced encryption algorithms like this that we're investing in accelerating that we believe will very clearly delineate between non-personal data on one side and PII, Personally Identifiable Information or data, on the other side. Whereas currently it's very much a gray zone with like k-anonymity and so on. It's partially identifiable, somewhat re-identifiable and so on. But we think we need to look at the web3 world where there is like, you know, zero trust all around. And they have developed ways to cryptographically separate these two kinds of personal and non-personal information.

And so, I think this will be a worthwhile investment. And the PEPs that we invest in, again, we will publish the guidelines and so on to the world, so that people can also think of this delineation in a way that retains the utilization of the aggregate data, especially indicator data. For example, if we want to measure whether people are overly reliant or dependent on AI systems. Again, this is deeply personal, but we can aggregate them in a zero-knowledge way, so that we can still do societal input and risk mitigation without any compromise on privacy.

I think this is the position that we take.

Yeah, maybe we can call it an existential opportunity, not just existential risk.

This is a great question.

So, before joining the cabinet, I worked on the basic education curriculum and also with the Oxford University Press to empower the majority world to co-create Siri-compatible inputs. And indeed, that was because for many of these languages, there is simply no comparable economic incentive for the largest AI labs at that moment to crowdsource data in a way that will inform the development of those frontier models.

And as I mentioned, in Taiwan, we have more than 20 national languages, 16 indigenous nations with 42 language variations. So, we have a total of over 20 languages. Many of these languages, if you try to ask a frontier language model about it, it outputs complete nonsense.

And so, we are happy to learn that nowadays distillation of those pre-trained models, fine tuning a quantized version of it, recently Stable Diffusion Turbo, all showed us that it is now quite possible to tune even the most cutting-edge models in a way that fits even a laptop. Indeed, I’ve been tuning my email replying model purely on my laptop since March. And so, my emails are drafted by AI models, and of course, I read all of them before hitting send.

But the point I’m making is that this is privacy conserving. It never leaks my laptop and if we extend that slightly with the technology like Talk to the City and Polis and so on, it means that the indigenous nations and the national language communities do not need to rely on the goodwill of a frontier lab, but instead they can just fine-tune their models in a way that's like Lego blocks, those low-rank adapter LoRAX, that can then compose together.

And this is exactly how we are doing our trustworthy and dialogue engine in the science ministry. We're working with not just Mandarin, but also Taiyi and so on, so that people in different languages in Taiwan have a real say in how to model their language instead of forcing people to speak a majority language. And I think this methodology is quite applicable to the majority world as well.

Yes, g0v. G-zero-V. So, this is a very interesting domain name. The idea is that for each of the government services that the people don’t like, people can fork the government. So, something that g0v.tw, you can change your o to a 0 and you get to the people's version of that website. And the idea is that the people's version is always open -source. And so, people can then merge back.

Indeed, our data.gov.tw, our open data portal, is going to be public code, pretty soon, next year. But its current version is already in form by the g0v re-imagination of the open data platform and many other websites as well. And the reason why is that we have a systemic competition called the Presidential Hackathon. The idea is that every year for the past five years now, more than 200 projects from the civil society that demonstrate some success in the local level, maybe just district level, can then enter this competition, which uses this quadratic voting and new voting methods to get the interest across sectors.

And five teams every year domestically, the president gives her trophy. The trophy used to be a micro projector that if you turn it on, it shows the president giving you the trophy. So, very meta. And carrying with it the presidential commitment that whatever you did in the district level or local or municipal level will, in the next fiscal year, become public infrastructure on the national level.

So, there is no prize money, zero prize money. It doesn't go to the team. It goes to the idea. This sector of collaboration, electronic signature, telecare, many, many of those infrastructure level ideas that prove on the municipality eventually found its way into national infrastructure because of Presidential Hackathon. So, this pre-commitment to take the co-creation of more and skip to the national level, to me, it is one of the most important political motivations and incentives for the civil data community, the open culture community to work with the government and with the people and not just for the people.

Yeah. The Ministry of Digital Affairs, which as I mentioned, didn’t exist until last August, is itself a combination of the most successful ministries when engaging with open culture. As I mentioned during the pandemic, the electronic services and e-government task force in the national development council responsible for open data, e-services, and the part of National Communication Commission, the NCC, that works on universal access and infrastructure level bandwidth allocation, the part of Ministry of Economic Affairs that works on platform economy, and the part of the Department of Cybersecurity that work on, as I mentioned, the privacy enhancing technologies, including the zero-knowledge contact tracing system.

So, all these were the champions within their respective ministry departments. And we work so closely together, literally having weekly iterations where we just get people's input and roll out new versions every Thursday, every Thursday, eventually became the new ministry. So, I would say that these were the champions, and now they're all in the moda. So, we're now spreading out again to the other ministries and municipalities by saying, this time, you can just provide your request, and we will just plan a public construction project that address all the needs, for example, instead of each and every ministry having their own backup system or electronic document handling system, signature system, and so on. Let's just work on these as infrastructure. So that's our current position.

This is great. Even philosophical question. I think in Taiwan, we're quite fortunate that ever since 2014, where we first started with this cross-source democracy stuff, that we've been able to work with career public service directly, instead of associating with any political parties.

We are in a position where we are known for credible neutrality. I’m known for not belonging to any political party, being non-binary not just in gender, but also in political affiliations.— in everything, really, and taking all the sides.

It is essential for any democratic deliberation to gain legitimacy through resolving smaller conflicts at first. For example, we first, in 2015, helped resolve the Uber-Ride-Sharing regulation issue, working with the Ministry of Transportation, and Labor, of course, Finance, Economic Affairs, you name it. And if we started saying, let's resolve this extractive economy versus gig versus platform economy, we will get nowhere. But if we start with, let's figure out what's fair for someone who is driving to work and picking up strangers and charging them for it, suddenly we can get a lot of overlapping consensuses.

And building upon that credible neutrality, the civil society then learns that, actually, you don't need the government, especially the national government, to run this process. The civil society can run this process as long as they have partners that are equally credibly neutral. The g0v bimonthly hackathons has been happening for about a decade now, takes place in our national academy, which is also known to be above all the ministries and parties, and also credibly neutral.

So, long story short, I think there's two main pillars. One is to build credible neutrality in a multi-stakeholder governance model, and also work directly with senior career public service so that they become the most important stakeholders because they are around, regardless of the ruling party.

Yeah, the official, like, e-collecting for referendums and so on, of course, requires that you are an 18-year-old citizen with all what that entails. On the other hand, the platform I mentioned, Presidential Hackathon, the join platform, e-petition collecting for referendums, participatory budget, and so on, that is relaxed, such that you only need to have a, like, FidO authentication, which is extended to residents, not just citizens.

And the residency, as I mentioned, can be earned through open-source contributions. And it also extends to people younger than 18. We believe that, like, those almost citizens, people younger than 18, are the most important. Because they have the most incentive and motivation and, frankly speaking, time. In the National Preservation Platform, join.gov.tw, some of the most active petitions are started by people who are 18 and people who are around 80 years old. Because they also, you know, have more time and care more about next generations. And so, these are natural allies.

And if we forbid people younger than 18 and residents, including expats, to participate. Then we lose the important voices that are not going to be reflected equally well in the traditional representative democracy, simply because they cannot vote for local city councilors.

So, a lot of our experiments in presidential hackathon and so on is about increasing the bandwidth of democracy and reducing the latency of democracy in a way that augments and complements, but do not replace referendums in the national level and the parliament itself. It basically informs them better, creates agendas better, setting priority for budget better; but at the end the decision-making paired is still in the parliament and the national level referendums.

Thank you. Live long and prosper.