>(Audrey Tang) “Started this idea called presidential hackathon and we use quadratic voting to let people have a say in what future projects are the most likely to harness this conflict into creation. Then we will make sure that those local projects get the national funding to become infrastructure the next year. But at the time we used the science fund funding for it. The problem of science funding is that it usually runs for three or four years, and then if you do not publish a paper, you don't have more maintenance anymore.

>So, when we encountered this issue in 2022, when we started the additional (Digital) ministry, we talked to our legislators about which committee we should be in. Because, you see, we are in charge of the AI platform, of buying some NVIDIA chips — so that sounds like the economy committee. Or we’re in charge of cybersecurity — that sounds like interior or defense. Or we are in charge of scientific research that can be used by the public web — that sounds like science or education.

>But we didn't go to those committees. We ended up going to the transportation committee. The reason is that by going to the transportation committee you have access to a different kind of budget — it’s called infrastructure budget in Mandarin. That kind of budget is like really building an airport, really building a train station. The argument is that, for example, when we built air quality sensors or water sensors, they have a property that, unless everybody interoperates, it has no effect. If each region runs their own measurement system and doesn't share a ledger, there's no common good. It’s something that only works supermodularly. It’s very much like a train station or an airport: if one airport only serves a single airline and you have to go to a different airport for another airline, there would be no real air industry.

>So, this interoperability qualifies for the infrastructure budget — just like building a highway or a train station. That is how we successfully got the infrastructure for privacy-preserving computing, for different government agencies to share statistics, or how we got the shared computation and digital signature infrastructure for small or medium enterprises.

>Long story short, it is three steps. First, it always starts from conflict — from civil society pointing out a problem. Second, the civic prototypes that prove supermodular synergy can then be promoted to a national scale (using something like quadratic voting to pick them). And third, you make it ‘infrastructure,’ i.e. permanently funded. That’s how local initiatives become part of the national system.”