David Dao: Can We Govern the Commons at Planetary Scale? | Intelligence at the Frontier 2026

Thank you so much. Some need to connect like the first video of every technology company is connecting its computer to the slides. Um and I think it will be the same here. If you can help me. Yeah.

Also, I really love that Einstein's birthday has been mentioned. Trisha, we have to talk about this because like when I was little, I I'm German. Um you will hear it through my accent while I'm talking, but I'll try to try to make it not that obvious, but I will I will I'm not sure I'm not your former governor. Um I'm not Schwarzenegger, so like I try to be like but like when I was little, I was reading every single letter from Albert Einstein because he wrote it in German. It was so fascinating like getting in this mind.

That's how I know he's not a good person. He's a good scientist. But So, how do you do this? I think that's probably the best way to So, this talk has been in my ads for a decade. Um but unfortunately it only made it to this last 15 minutes But what I'm going to try to tell you about is a little bit of a thinking of how I believe we can govern the commons in the intelligent age.

That's a mouthful, but I think that you'll start with okay. I think as I said, I really like to start with a poem. And I really like to start um with with a poem actually from Richard Brautigan that also inspired um Dory Young Day to the very messy poem um watched by machines of loving loving grace. And so the poem goes like this. I like to think and the sooner the better of a cybernetic meadow where mammals and the herbivores live together in mutually programming harmony.

That pure water touching gaze come. I like to think rain or please of a cybernetic forest filled with pines and electronics where deer stroll peacefully past computers as if as if they were flowers with spinning blossoms. I like to think it has to be of a cybernetic ecology where we are free of our labors, and joined back to nature, returned to our mammal brothers and sisters and all watched over by machine machines of loving grace. I've spent my last 15 years in the rainforest, literally. So, I founded a non-profit called Rainforest.

And my background is computer science and a PhD in AI systems back in 2012. Moving on to kind of create like AI systems that can monitor and support local communities around the world with forest protection and preservation. One of my more known works is I created one of the first prediction algorithms to predict deforestation through satellite imagery. I created a field um um co-created a field with amazing intellectual collaborators called data valuation that puts a price on the data point. What if you would basically sell a data point to like get a reward in a data marketplace?

And all of this culminated in a non-profit called Rainforest where I wanted to go to local communities where there is no there's economic infrastructure, uh but there's a lot of opportunities, but it's beautiful, rich, and important life outside of all the the natural Amazonian forests, the mangrove forests in the Philippines, and that are trying to kind of put the eco to planet Earth, but are not funded really well. And so, we created multiple kind of um works with communities around the world that help them use AI to digitize and monitor um to kind of try to figure out how can we create these guardrails to fund them, and that ended up in winning an Xprize last year, Xprize rainforest, which is really cool. And so, my question and the talk that I want to give here is like, how can we scale human cooperation through trust with machines in the intelligent age? And I think we can start with like this guy, which I'm not a big fan of, Hardin. He is he he asked the question, or we all asked the question, can we have a nice day today?

We are like in the college dorm and as the the case he's saying is really he out there cuz no one else is eating it. Can we have that nice space? No, because he says if there's a commons and that is like a society believes in a field of the commons then ruin is a destination towards all men are kind of like pushed towards because they're all that selfish as I said. He didn't say it in example and he turned that terminology that many of you might have heard called tragedy of the commons, right? Um 1968.

And then this wonderful woman came in, Elinor Ostrom, and said no. Humans are neither all angels nor all devils. And what she found out is she looked into my beautiful home in Switzerland, but also in like small villages in Nepal, in fishing villages in Mexico, and saw that people getting along, people are trusting each other, and there are various societies that are able to self-govern. Right? So, like that tragedy of the commons holds that we all humans are kind of like devilish humans.

Sorry to interrupt, that is just just a mistake. I just I struck as she actually with that work won the Nobel Prize in the economies as the first woman to do so in 2009. And basically what she figured out is like all these communities they have eight really cool principles um that they kind of get together and these are now termed Elinor's robust design principles for governing commons. Number one, clear boundaries. Number two, local rules.

Number three, collective choice. Number four, monitoring. Number five, graduated sanctions. Number six, conflict resolution. Number seven, external recognition.

And number eight, nested systems. Right. So, if these eight rules are somehow all part of a great governance, it can have self-sovereignty. But we also know from enormous amounts of data and our belief system and what we have observed that this is not the case in larger scales. You can see this is a beautiful and horrific image of satellite imagery over like decades in Brazil, Mato Grosso.

And you see the scale of destruction you're like cancer of eating forest. We have lost more than half of the world's forest that ever existed since the industrial revolution. We are losing the size of barrel of the UK forest every year. And that is just enormous if you bought this. So, it's this deforestation is one of the most destructive activities humanity ever do cuz it takes away the livelihood of the people like the long term stewardship of the fresh water.

Nature itself provides so much more value. So, obviously there's a failure of coordination failure issues. And I believe the reason is um cognitive load overload. So, Elinor Ostrom rules really work well in small trusted communities, but they break down what I think you can some people call Dunbar's number, the cognitive limit of every human to have stable relationships. Which is estimated at 150 people.

You as a person, if you're not a hyper networker like some of us are, um but even if you just truly value friendship, you will not care about 150 people in your personal life that you can't have realistically you can't have communicate trust there. And so that is a a system and nice. Like if you have 150 people and they all have relations together as I did then thousand relations each other may take. This year is my issue is a really relations as you really is just here in our possible right. Because the patients got to be the pollination cost is really specifically all of this leads to the breakdown of all is Ostrom's rules.

And I believe together with James here from Protocol Labs, we are actually figuring out a field definition of how we can improve this system and we came up with the term of programmable economic infrastructure. Ways where we can scale Elinor Ostrom's eight principles using modern frontier technology. And so programmable economic infrastructure talks about the theory and practice of designing social economic intelligence systems that preserve or enhance social capital, trust, and agency through scaling the governance capabilities with the technologies that we have right now at our hands, right? So and how this would look like exactly is you take like if you want to be very academic, you can take Elinor's eight rules, you can look at them, and you can look at how it can be enhanced through technology. Rule number one, clear boundaries.

Your reference borders are only multiple universes. That's Elinor's. But but what if we have digital identity systems, access, wallet addresses that allow us to see who are part of which type administration? In local rules, they are looking at traditional data systems that that constrain. What if AI systems can form and create these local tables and refine them at large scale so communicate them, right?

The easiest thing that Elinor talked about the second because that's where we probably expressed and showed that that's possible is monitoring. Like Elinor has collective choice monitoring. These are only possible with face-to-face meetings. And I monitor with basically increase walking food forest and greater controls. Now you get doing this we can leverage artificial intelligence to trace the really top each other and we can narrow this autonomous sensing and AI to large-scale monitor.

For example, in the case of the forest all of the Amazon rainforest. So how this looks like? This is some of the work that I did in past. As you can leverage robotics and AI to attach sensors on trees autonomously. On the left you can see a really really cheap drone a DJI Mavic Mini 2 that is like $300 now.

So they can be scaled at car and no cost around people and so on and they carry like a small biopsy sensor that can listen to the sounds of this forest and that drone can autonomously attach to other branch. Um And then just this is to the bird's sounds and variables. And you want to monitor. Also if some customer change so as well. On the right side you see a drone that has a tissue uh tissue paper attached to it on a long rope and it puts this towel tissue paper through the forest and swipe the forest leaves because you can catch the DNA of the leaves, bring it back to the lab, sequence it and then figure out what species that forest has, what monkey touched this leaf, what water was drank by like let's say like a small insect and that is a very rich sensor as well.

And the best part like local communities are natural in flying these drones and I've been there tested out a few of these drones at large scales where I first meet them on the road and the fact is that they know the forest the best. You know exactly how to do navigate these these drones. They're also very playful. All they do at like most of the time in forest is like they're running around navigating and so that is these activities are very natural to them. They know the biodiversity hotspots.

So we scale of technology their national ability mitigate with the union arm. And you can see all these things that we capture through that just working with the communities. Like some of the rare species that for example this jaguar on the left, this is something super rare that like usually in the past, like a decade ago, if you don't wouldn't have like these sensors and eDNA traces, you would take like it would be like your lifetime achievement to take this picture. Now they're able to if that's really rich, we're able to get this in a regular basis. And you can see here the community engagement as well.

This is like this is really lovely. So, digital identity and wallets. Like you can connect now this data infrastructure, the monitoring to crypto native payment rails. The second big problem of Eleanor Ostrom's thing is that the group is really hard to define because the group in a nation-state is a passport, right? And so, here you can use like identity systems to basically pay them out and like a wallet address itself already allows you all the access to banking infrastructure that would be traditionally like prevented by not being able not being part of a group because many of these communities in the Amazon rainforest, they don't even have a passport or paperwork, right?

So, how do they even have a bank? So, that is really really amazing. And then like there's a lot of cool things that um this community has worked off. One is for example data make a fairly with impact infrastructure. One of my most bullish of episodes, um that's Eleanor Ostrom's now own version foundation that allows us to create this early monitoring here as I talked about in a format that every machine to read.

You have to imagine like how the World Bank right now this used by us is someone this a picture of a picture of a PDF document. Like how are we going to be feeling still to evaluate ID notices? Like having new formats like habit sets allows us to instantly focus on video and AI data set like data set up and allows to carry all the data in a research. And one of the most exciting case that's happening right now in the frontier tower is phase two of the funding experiment. I mentioned briefly because you base one was like in all four leads money and see what they do.

And obviously we know exactly what's happening. So we will we will report on this later. But the second experiment is we have $5,000 that we're going to distribute by the last 72 hours. We went through every single floor and we created a digital copy of them by having them interview for 30 minutes to an AI. Now these digital copies, they're now walking around here.

I can show it to you in the platform called Simocracy. And you can visit this as well in simocracy.work. You can for example, if you really like and I'm a big fan of the 14th floor, the human flourishing floor, Judy created a avatar called Judy Cat which you can click in here and you can chat with her. You can ask her like how um do you value and she copied the way Judy speaks as well as the way Judy thinks through a 30-minute interview.

Um and now everyone of these floor leads, they have so different persona. And Judy already uses this platform to kind of because they usually don't meet each other across the floors to talk to floor leads they never talked to and instead talk to their AI twin to figure out okay what are they what do they care about. And so the goal is like we have these $5,000 every floor and every member here in the event can propose projects to fund and these AI agents will then get together and talk to each other in a discussion later on to decide through a cognitive quite like important process called the S process to decide how things are should be funded or not. And yeah, this is an early spirited really societal thought. So this also is because I see that I'm getting pushed out just to prove it.

We at PL we're really excited about this field. We care a lot about this. So together with James Canning also all these here if you has up here. We are figuring out exactly the biggest opportunities of this space to push forward and get involved. And we come up with four opportunity areas, sovereign digital public infrastructure, AI crypto native public goods funding, crypto native democracy, and decentralized climate infrastructure.

Can learn more about this in plresearch.org. So if you have really cool ideas you want to participate you will come tell us like David, James, what you're thinking is totally We also accept this. Just visit the website. You can apply for a grant if you're also interested in this.

So really excited, really really really stoked to shape this future of public goods funding with all of you. Thank you so much.