I’m up for giving a bit more of the origin story of Reset in my own words rather than the polish that is the website. Maybe that will give you some more questions and then maybe end with some very practical stuff about the open call and any questions or overview of where it is now and where we see it going in the next year.

OK, yes.

That sounded like a heavy agenda, but actually, it’s not very much. [laughs]

It’s OK. It’s just surveillance capitalism and statism. Simple stuff.

No big deal.

I’m trying to ignore how heavy that problem is. That definitely is what drew me here. I’m Dan Blah. My real name is Dan Meredith. For the previous most of the last decade, I was working on Internet freedom at the Open Technology Fund.

One of the big limitations with that work was that we couldn’t really instigate more northern, western European actors to confront the issue that so much of what we were seeing happening in other countries, in repressive regimes, was being enabled by what was happening in surveillance capitalism for economic reasons with lack of policies in Europe and the US.

That was a major motivating factor for me wanting to work with Ben to create Reset and to get something that would try to rally that issue.

I’m a big believer of open calls and transparency, openness and free flow of knowledge exchange. A lot of what is baked into OTF is being brought over to Reset. You and I were talking briefly over email.

One of the differences is where OTF had a baked in process that everything had to go through an open call for any receipt of money, Reset is prioritizing that, but we can also work directly with people through who we know because it’s private funds. We get a little bit of best of those worlds.

That’s my role or where I came into it. My role at Reset initially is working with Ben to set up all the proto infrastructure and bootstrap the thing and get it going.

Then, as Alaphia and Ben get more into their work of advocacy and policy respectively, mine is to work behind the scenes with research and technology to both enable and augment their work but ultimately make sure that any of the technology and research we’re doing is in direct service of an advocacy or policy plan.

The goal for Reset is better laws and/or litigation proof of laws in places that are affecting the globe around this issue of surveillance capitalism. We’re investing in tools and technology and original research. If it can’t be directly drawn to advancing that cause, it’s not worth doing.

Different than at OTF, where we hoped a lot of that happened, but we didn’t fund that directly. OTF worked with technology and hoped that the advocacy and policy people would come along and make use of it and gain advantages of that.

I just wanted to frame that that’s a bit of a head cognitive dissonance wrestling that I’m constantly going through, is how to be making sure the tech is in service of rather than “Oh, that’s a cool, glittery piece of technology. Let’s support it and do it.” That’s me.

Question.

Yeah, please.

Because I’m not yet, anyway, affiliated with Reset, I get to ask questions that might sound like conflict of interest after I get associated…

That’s all right.

What about nonprofits like the Polis, now a charity. I’ve been working very closely with Polis, as you know, as a assistive intelligence technology that facilitates crowdsourced agenda setting, including the crowdlaw initiative and all that.

One thing I understand from the grant proposals that they’re working on is that there’s essentially two parts of it. One is the advocacy, the facilitation, and so on, that part, which sounds more like what Reset is doing and interested, but there’s the other, mundane part, which is just the software engineering and research and scaling. That’s the code level thing.

Without the code level thing, actually we can’t prove that this assistive intelligence is actually value aligning and accountable. The harder to fund part is actually a essential part, the reason of existence of the easier to fund part.

That’s not limited to Polis. We see this situation with the CONSUL Democracy Foundation and many other similar initiatives as well. What’s Reset’s position on this regard?

We don’t really want to choose any one…Before I talk about, I would say, the more mundane technology and research aspect of it, maybe this is a good spot for you, Alaphia, to jump in and describe your growing vision around advocacy and policy in the frame of that question.

Yeah, happy to. Really nice to meet you, Audrey. I did not know that Dan was already in touch with you. I happened to read about your work. I was like, “Well, we should definitely get in touch Audrey.” He was like, “I’ve got that covered.” Really nice to meet you.

I’ve been at Reset for a month and a half. I came from a global campaign group called Avaaz. In my last two years at Avaaz, we’d started working pretty heavily on the issue of disinformation.

The reason we did that was because Avaaz, it’s a generalist group, so it campaigns on a range of issues. We had started to see pretty systematically that every single issue we worked on — whether it was climate change, environmental rights, human rights — every single issue was impacted heavily by the crisis in our information ecosystem, disinformation, and hate speech.

It was a natural progression for us to then work on disinformation. We started to see how alarmingly it was starting to impact elections around the world. That’s what brought me to the work.

I’m originally from India, and I have seen in the last two elections how the rapid explosion of disinformation and hate speech has seriously impacted a society where we once knew how to live with each other, tolerate each other, love each other, even, I would venture to say.

It’s been painful to watch. That’s one of the motivations of this work for me. I am a point where I did spend the two years that I did that disinformation work talking to the big platforms directly. It just became really clear to me that, in some ways, that’s pointless.

Even when there are good people there — and there are many good people there — this thing is just too big for them to rein in and to bring under control. I am very passionate about the three pillars of what I think hopefully will be Reset’s work, which is regulation, legislation, and potentially litigation, but we shall see.

I’m not a technologist, so I think that, like Dan, I feel like you might be better placed to answer that question. I just wanted to lay out my motivations and Reset’s priorities, which is to work in countries where there’s legislative activity.

Right, but disinformation is a really good example. Just setting aside Polis as a vaccine against divisiveness through informed conversation and so on, just in the disinformation space itself, there are technologies that, if properly funded, will actually make real time response systems much more agile.

This is like in the early days of the spam fighting algorithms that, when advanced, it makes the spam issue not really go away, but it equips the multinationals, the mail providers, with this sense of how even the naïve Bayesian, the early 20th century machine learning, can actually put a check to the spam issue.

That enabled then the norm around the junk mail folders and the flagging email as spam. Then led to the targeting and the other technological measures, after which it became possible to bring, as you said, the regulations or the litigations.

There’s a clear way out, and that’s the norm we want to shape. Then we sue the mail providers that are negligent to the existing technologies that we can show convincingly that it’s a net win for everybody but the spammers involved.

Funding of those foundational technologies, like the security layers of the Internet, has long been in neglect, of course. I understand if Reset is mostly about looking at the landscape and advocating for the things that has been proven to work, instead of Dan just mentioned, which is more like Open Tech Fund, like the early stage incubation.

Like just funding a bunch of things and hope that one will work. My question was more around how mature in the readiness scale is the cutoff when you decide to start the regulation, advocacy, and litigation work to start saying, “OK, if a certain videoconference platform doesn’t do end to end encryption, we’ll just litigate against them.” That’s another example right there.

Dan, before you jump in, maybe I can take that from another angle, which is that ultimately, I think that we will want to campaign on these issues as well. We want to shift public opinion on the big question of surveillance capitalism.

I think our case becomes much stronger if we can show alternative technologies that can do that. I think that’s where the technology piece is actually vital in shifting consumer demands.

OK.

I don’t have a whole lot more to add to that. I would just say that I think one of the goals and the thing that excites me about working with you, Audrey, is that you can help us define what is worth investing in that’s mature. Then what isn’t worth investing in that might be more nascent and new?

Where you see a gap in policies that don’t exist but technology that address that social problem, and you see Reset as squarely fit within that, yes. Then vice versa, where you see us looking to support something that clearly has a regulatory answer.

From a value proposition, in your expertise, it obviously makes more sense to go down a policy advocacy litigation. Like, “Let’s fight this with law, not just inventing a new piece of technology.” Then the point of the Strategy Network Council, advisory council, or councilmembers writ large, is to make sure that we’re following the right currents.

We’re getting pulled there, because we couldn’t do it on our own. What I’m really excited about is you just asked the fundamental question that’s going to be driving Reset for as long as we’re around. How do we get that balance right?

Where you’re asking it and have opinions about it, I feel like we’re going to be more in a good direction.

I see.

I’m putting it back on you, basically. It’s like, the answer to that question is going to be moving, and it’s really going to be dependent on where you pull us.

That’s good. It’s a really good answer. Like sometimes we press reset, and sometimes we press turbo. Nowadays, the computer doesn’t have the turbo button anymore, but we don’t shut down. I get the answers that I want. Thank you. Please continue.

I want to just, where we’re going back from Alaphia framed it up, where she’s a month and a half in. That was where Reset, Ben and I’s eyes, we didn’t really want to start going until we had the three of our positions representing those three underlying types of work for those pillars.

Where Reset has been actively making grants and supporting folks in many different countries, which I’ll describe just quickly running through those, and provide you a list of all those things…It’s all coming up on the website, but it’s a work in progress.

There will be a place that you can point to publicly that shows what kind of work you are a part of. I would say 2020 was supposed to be, we had a real finite plan in November and December of 2019.

Then with COVID, with the turning political tides in the United States, and our focus and our plan for Europe being very Euro centered in our initial approach, got a bit scattered. We started to have to look, because misinformation became such a health emergency.

All of a sudden became something that people could talk about because of the health related issue, rather than just politics, meant that it was much more approachable. Which meant that we, just to be really frank…

Alaphia, you may have a different interpretation, and if Ben were on the call, he would not clinch, because it’s the truth, but he would be like, “I don’t know if we should say that.” Just speaking really frankly, we spread ourselves really thin.

We’re in Australia. We’re in Canada. We’re in the UK. We’re in Europe. We’re in the United States. We’re trying to figure out, as Alaphia said, what are the countries that are not within those typical global north countries that we should also anchor ourselves in?

I would say 2021 is where we’re actually going to really get started with a stronger vision that Alaphia has been championing, and start following our councilmembers’ guides, and having them all in place.

You’re coming in at the right time. All this to say what I’m trying to dodge is to get too specific about exactly what Reset’s strategies firmly are right now. We know what kind of work. We know what kind of focus. We know what our key outcomes are.

We have these countries that we’re working on. We’re working with technologists, lawyers, advocacy folks, but where we’re drawing firm lines on how surveillance capitalism, grounded initially within disinformation and other issues, is what our attack vectors are going to be, where we’re going to try to take the bricks out of their wall.

That’s still a very open question. I want to leave that to where you feel strongly about things that we should be doing, there’s a lot of space for you to push us, pull us, yell at us, grab us, and make us go in that direction.

That’s very clear.

I would be inclined to go towards some of the mechanics of the open call, because I’m really excited about some of the underlying technologies and things there. Do you have any questions at all of high level strategy, vision, remit of Reset, before I get into that?

No. It sounds quite clear. Actually, that with the infodemic, as known by people, instead of doing the mis/mal/dis-information dance, now, it’s just known as the twin crisis of pandemic and infodemic.

Which as a good ring to it, actually, because we’re essentially taking then a vaccination approach toward the infodemic, and not necessarily this administrative shutdown vision. The people understand how bad it feels to be under lockdown.

An extension of that could be made to the infodemic scene. I totally see that the tide has changed and the strategy that used to be “raise people’s awareness” can now actually ride the wave, so to speak, of the popular awareness about infodemic.

I’m totally with you, and I want to hear more about the open call before doing any interpretational work.

Great. Then Alaphia, I would say if you want to embody your best bent on the harder policy and advocacy stuff and then the work that you’re moving into, I think that would be useful, just to know some of the policy realms that we’re engaging in.

As an overview, what’s happening with the UK, the EU, the US, Australia, for example? Before that, the open call is, in a lot of ways, very easy to describe and then hard to articulate its impact and its effect, because we don’t really know yet. It’s all dependent on who applies.

I’ll skip right to some of, I’ll tell you what’s the important thing that I see in your role and then come back and describe that a little bit more. The network effect, and you as a supernode within a region and a particular kind of interest of technologists that are societally and politically, is going to be…

Whether you see and/or are able to convince people to apply to Reset through open calls is going to be a direct factor of whether or not our open calls are going to be impactful. It’s not just about you reviewing what it is.

It’s about you feeling empowered to speak on behalf of Reset and say, through that position, “You should apply to this thing. This is a thing that I’m a part of. I believe in this process, and that’s why I’m engaged in it.”

Then ultimately, if we support really good things because of that, we can all be excited about the open call process. Stepping into that, how do we make sure that the mechanics of the open call align with those principles and vision?

It’s an open source piece of software. It’s hosted on GitHub. We have all these plans to get it to GitLab. It’s the same thing that OTF is making use of. In that way, it’s baked in from the start to be something where there is a clear set of policies, guides, principles, and questions that can change and iterate very quickly based on feedback.

Feedback is a crucial part of every major problem that someone has with the application process that isn’t me actually just being delayed, underwater, and slow to act, is something that we change very quickly. We have a dev team that manages that.

It’s built from a process that starts with the team and then ultimately an external set of advisers, and then broader community feedback. These are the ways that a proposal can loop in. Reset’s going to be the new one that’s going to start with the broader community feedback.

If people apply and are considered within remit, we’re going to end up saying no to more than 99.9 percent of people that apply. That’s unfortunate. We just don’t have enough money yet. Even if we do have more money, it’s still going to be one percent or less of people that apply, because of just this issue and the excitement that gets around it.

That’s the first part of it. How do we go about creating a process that gets a good decision, and we make sure that those one percent of folks are as good as we possibly can choose to help our cause?

Money isn’t the only thing. One of the things that we’ve been trying to do on the other side is how do we create a network of those folks who are dismissed and approved, but ultimately, they’re dismissed because there are still a lot of really good ideas that are there.

We can call them the 99 percent.

What’s that?

We can call them the 99 percent.

That’s right, the 99 percent. How do we help the 99 plus percent collaborate and commiserate, both on their dismissal, but also still be a part of the ecosystem building? We’re all within the Reset.

People that apply consent to sharing information basically about their proposal. They’ll be able to see everyone else’s basic information about their proposals. We’ll have an opportunity to communicate and contact with them.

Reset’s offering not a huge amount of money, but a very low bar for two or more of those applicants to come together and do something collaboratively. We’ll just give each of them between $5,000 and $10,000.

Very low question ask, a much smaller process, just to see the collaboration, start cross stitching the network, and building the network up. We’re calling that our flash grant. In those that we award, because what we totally recognize initially is I’m a six foot, white American guy who told everyone that I know.

Most of those people, unfortunately, look and feel and think more like I do than a lot of the people that are being harmed by surveillance capitalism. Another aspect of our flash grant is allowing a referral program.

Ideally, we don’t select, over time, as many people that fit that box. Then ideally, they have a bunch of awesome people that they wish were a part of this network as well. By just the fact of having someone who’s trusted refer someone else, we give them, again, a low bar.

Here’s $5,000 to $10,000 to just get in and do something that’s interesting. In that way, we look at the open call as a real knowledge exchange, not just a decision process for us to make best use of our money.

That, I guess, also, while there’s…It’s as easy as it is to change the code in the system. How you want this policy, or how you think we could change to make this more of an interesting network effect and spread this out, for both Reset, but then also, as a model for other philanthropies and other civil society oriented folks to make use to.

That’s what I’m really excited about. I want to be super, super frank right now, that I had a baby, moved in COVID in the middle of all of this, got really excited, and had…I’m a very big time optimist is my problem. It’s my curse.

We are really behind right now. We unfortunately hired and are bringing in folks to help me manage the open call. We got over 200 applications and statements of interest from folks that I did not expect on our first unadvertised open call.

Where I was intending to have all those decisions made before October, it’s now looking like we won’t get all that done until November. We’re going to open it back up in November. The goal is, once I get my head above water, there will always be an open call, and there will be a rolling process of making decisions every two months.

That means that every couple months, very practically, we would be sending you a set of finalist applications that we would…It might be two to four, tops, that I would expect take you between 30 and 45 minutes to take a look at, review, give your up or down and some basic feedback.

Then also, you would have access to all of those that we would consider have been in scope to Reset’s cause. From your own personal research perspective, if you hear or someone comes to you and says, “Oh, I applied to Reset. It didn’t really happen,” and you wanted to have some insight to, what was our rationale for that?

And/or you were just really curious about what is the world that Reset is seeing from this open call process? Hypothetically, if we succeed, we may very well end up in a situation where some of the best ideas globally are all coming through Reset.

It will be really useful for, as a seed for other initiatives globally, to do some of Reset’s work to be able to make use of the proposal data and any information that Reset has. We really want to give that to you with the understanding that this is for you to do what you will with under best practices of data sharing and all these other things to use it and advance the cause in any way that you think can.

OK.

That was a lot.

You mentioned the qualitative part, which is writing reviews. I’m fine with that. You mentioned also a quantitative part, which is like or unlike. That part, I am less sure about. Do I have to do a thumbs up or thumbs down? How does that even work? Is this like dot voting…?

I love it. There used to be maybes, and I think there still is a maybe, where you don’t have to say a yes or a no. Then the issue that arose with maybes…Here’s, my dream is to replace the whole quantitative voting with a more modern voting scheme.

There’s a couple that I would really love to do. I think there’s my very logical, analytic side needs something quantitative. The value is definitely in the qualitative review. I would say, once you see this, and you know a better way…

The dream of mine is having someone like you in this role, because you’re going to know exactly how it should be, if you don’t already know already. I’m going to have one simple job. That’s just to implement it.

You tell me how you think we should do this in a more quantitative fashion that feels a bit more fair, that would align with a value system that you think Reset should represent. I promise you, within a month, the developers will have it so.

I know a few folks in the Gitcoin project have already implemented quadratic voting.

That’s what I want.

It doesn’t even need to be implemented. Of course, I think there’s two part of it. One is this vested power for just a handful of people to quadratically vote. That’s one model. The other model is, of course, more crowdsourced, but with maybe less voting powers for the community, which has a bootstrapping problem. We can solve that.

Anyway, there’s actually the third model, which is called quadratic funding, which makes it more crowdfunding. I am not sure that your financial instrument is set up to do crowdfunding. The Gitcoin chose the third model, just for the record.

The funny thing about this is, here’s…I love it. You really got right to the nerve. At OTF, because we were government funded, the idea of expanding the underlying software called Hypha that runs this to be something more than just a decision process for compliance purposes, we couldn’t ever justify investing funds into those features.

That wasn’t what the government funding source that we worked with was ever going to really support. We don’t have those limitations at Reset. We just have a time limitation, and we ultimately have a money limitation.

From a remit or an overarching governance, our donors don’t have those sort of concerns. The stepping into how do we go about better leveraging the network question, all the way down to this particular feature of how are votes tallied, is totally within scope and a priority.

Where one may seem…We’ve been talking with the Open Collective folks, who Pia is one of our council members.

I know. I am part of the donor network there.

Awesome, right. There’s so many platforms. I’m not saying Open Collective is one that we should use, but there are so many platforms that would offload the work of maintaining the crowdfunding aspect. We just have to leverage the API.

I am so 110 percent focused on, there is money that Reset will offer these projects that are considered the best, but there’s no reason why we couldn’t facilitate that into a crowd offering and a crowd vote as well and have that be some sort of metric for impacting success later. Thank you for just bringing all of those things up and being councilmember who would support moving into that direction.

I say that, because that would actually be a go/no go thing for me to really spread it to my network. Most of those people are into collective intelligence. Really, the only way to get them interested is to include some sort of social choice in this mechanism.

Otherwise, frankly speaking, there are many other places where they can post their ideas. Of course, you know all about it, that’s the thing about open governance. If you join early on, you get meta. If you get meta, that’s much more fun to be had.

I think I’m happy to hear that that is the direction it was taking. Of course, I understand that there is a time constraint on the bootstrapping round, but that doesn’t matter. All that matters is the pull request toward the constitutional kernel of this decision making process is open. That’s all I need to hear.

I was just looking through some of the issues. There is, actually, I have a “replace the current voting with the quadratic voting,” and then point to some implementations. There are much better implementations now.

What you just, as a to do I’m taking is to flesh out and modern that issue up. Then look to you for who do you know within your community that would be interested in either, not the development, just thinking about the implementation and being more of a scope and protocol architect on that side, if they want to take a look. That could be you as well.

Sure. I know a few people. Well, the Gitcoin people, and also the game designers of Civilization VI. They also have quadratic voting there… but maybe we don’t go there? [laughs]

No, actually, I think that they’re a great… To Alaphia’s point of alternate technologies, where the surveillance capitalism phrase can often provoke people to think that, if you’re a supporter of antagonizing surveillance capitalist, you’re a communist, a socialist, or you don’t believe in markets.

Gitcoin is an example of, “No, people are making money using much more fair and distributed decision making process.” As an alternate technology, an alternate ad network, an alternate… There are so many different things that they’re playing with that are great examples that I think align with things that we should stand up.

Where we could potentially bake that into our own process, that would be awesome.

Maybe we call it, I don’t know, “sousveillance capitalism” or something like that.

Yeah. Any other questions, thoughts, or things that you would want me…? I think I sent a list of the guide. It’s ever changing. One thing that I want to just really flag is where we’re asking people when they apply, when they review…

We’re not asking you to sign a contract for your honorarium or any of your work. What we are asking you to do is to abide by a code of conduct and some best practices on data privacy when you log into the system. That is a contract. I don’t want to not acknowledge that…

…in EULA style. [laughs]

That’s right.

That’s exactly what it is? Well I will click without reading it then, I’m sure. Anyway…

Here on the record, I can very much affirm that we have no intention to rally some lawyers around people that violate this. We’ll just remove access. Anyway, what I did want to, on the code of conduct, there’s a set of values and principles.

This came up with a previous councilmember, very rightly so, which is for them there are movements that are specific to their communities and their people that they don’t see as Reset identifying with or representing in strong ways.

My commitment to them was that’s just because we haven’t formally engaged you to know that those things should be there yet. It’s not that we don’t, it’s just that I’m the editor for this page. They should make a judgment and hold me to it.

I wanted to put that back to you, when you do see, and if you do hear any feedback from your community about the values and principles that Reset is abiding by, the code of conduct that they’re using is limiting, and/or hurtful, or affecting in any sort of negative way, we really want to know that.

At the top line, before you even get to how you’re voting, what you’re saying in your vote, and what it actually means, these things are defining us. They’re the ones that we want our advisory councilmembers writ large to have a say in.

When you go to log in for the first time, you see those things, and you either don’t click them or you click them and go see them, and you see things that are missing, or you hear people within your network who are applying, and they’ve got serious feedback, open door.

Please bring it to me. Tell me what needs to be changed. I will ideally, if it’s just an oversight, a done, I should have known better, I will be able to send you right back 30 seconds later the same page changed, and we can go from there.

I did want to just, before we get to the code, there’s the guiding lights. Those lights can be changed, shaded, or risen up.

I remember looking at the OTF code of conduct and seeing the unsolicited typo corrections and grammar corrections as microaggressions. I really like that, by the way. The first time I’ve seen anywhere, and I’ve seen my plenty, my share of code of conducts. It’s brilliant. It’s brilliant.

Yeah. I haven’t actually told Ben or Alaphia this yet, but I did publish an offline online code of conduct for Reset to abide, because I wanted it to be true before we got into reviewers, applications, people posting feedback, and all these other things. I think it’s based on Django’s, with a little bit of modification. They have a pretty solid one.

I checked the bylaws of the public servants here, and it seems that, if this is not a post, then, of course, I can do reviews. That’s entirely fine. I can accept honorariums up to 5,000 NT dollars per hour — that’s the ceiling.

That’s fine. Is that something that we can do? Can we do a single disbursement for a projected amount of hours annually? Or do you need to do an invoice?

We can do this in a single payment, but not annually, because that would mean it’s a post.

Got it, right. Of course.

On the next Reset, we may have another talk like this, but it can’t be guaranteed.

That means you will keep track of your hours spent doing the work of Reset and then bill us for it, and we’ll pay you.

Definitely. I will just bill it as a single transaction for this round, of course, of open call. That’s kosher, and as long as it’s transparent — which it is, we’re making a transcript right now — that is fine.

I don’t usually do downvotes. It’s a personal thing. Maybe I will just give qualitative reviews on things that I don’t like. I can leave reviews on how to make it better, but I don’t really like doing downvotes without this whole quadratic voting thing.

If we have QV, I can give it one point to show that it has potential, and basically treating it as, “I’ve read this.” Then with just upvote and downvotes, there is no such spectrum to be made, so which I will just skip… I will do the qualitative part.

I will do two things. I will make sure they’re not required, because they might be required right now. I haven’t actually used them yet, but I think that’s the default. Then if it is a setting that for right now is limited by needing to be required, I’ll just make sure there’s a maybe. You can put maybe or pass. You don’t have to do a downvote.

“Maybe” is just fine. Assuming that there is an ongoing dialog going on, then maybe it turns into a yes at some point. If there’s only one round, then the maybe is good for me.

I’m so 110 percent committed to… When we have this conversation this time, say, next year for another year’s commitment from you, there will be quadratic voting, and we will be using that. It won’t be an up, down, or maybe. I 110 commit to that.

OK, yeah. That’s all I need to know from my end. Is there any input from Alaphia?

I actually had a question. Dan, if you’re done with open call, I can jump in with my question, if that sounds good?

Absolutely. Take it away.

Great. It sounds like you do know Pia, right, in Argentina? Excellent. She’s on our advisory council, and we were having a conversation with her and another member of our council, also from Latin America.

The thing we were discussing is we have a good sense of what’s going on in Europe, North America, Australia. Obviously, technology is being used very differently in different parts of the world and impacting people very differently in different parts of the world.

I’m throwing a very open question to you about what are some of trends in Asia that we really should be paying attention to in the context of surveillance capitalism? That’s one. It’s a big, broad question. I’m aware of that.

The second is I am thinking about an open call, not for technology, but in some ways to invite communities that are impacted by the harms of surveillance capitalism, to apply and to be able to document and build the evidence base of the harms in their particular context.

The reason I’m excited by that idea is because, in the past, a couple of years ago, we launched a whole hate speech project off the back of just one stray comment from a community member who just said, “You know, it’s really weird. As the government does this citizenship census count, we’re seeing a massive spike in hate speech against particular minorities.”

That led to a whole workstream and proper documentation, and we unearthed quite a bit. I want to find those people, but I don’t know what I don’t know. In some ways, I don’t know who to approach. That’s my second question, which is what are some…Where and who are some of these communities that we should also be paying attention to?

I don’t have a good answer to the second question. That’s really not my forte. I will just be frank and say that. In Taiwan, the Doublethink Lab would be the people that you should talk to who are actually in that trade and do some pretty good, along with Graphika and the usual suspects, on hate speech/disinformation/election interference studies.

My role — as digital minister in charge of social innovation and open government — is to stay above the partisan stuff and just work on the general resilience strategy, usually around lifelong education and K 12 media competence training.

That is more like the immune system toward the hate speech and so on, but I don’t do particulars. I don’t track down the people who do informed disinformation campaigns, hate speech campaigns, and things like that.

If I do that, then I move toward law enforcement and investigation. That will actually nullify my position as the nonpartisan architect for digital resilience, but I do know the people that knows more about me — sorry, about this than me — actually, maybe about me, too. [laughs]

For the first question, I think really, in Asia, the main thing is surveillance state capitalism. There really is no escaping that.

It’s the data norm that’s conveniently called, I think, the rule by law, instead of the rule of law, norm that’s established by the People’s Republic of China regime. That has a capitalist shell, but then it’s statism underneath at its core.

Then through Belt and Road, through state subsidized telecommunication system, through all sorts of various different things, they gain traction much as surveillance capitalism would, but with subsidized state backing. That makes it even worse.

That’s really the thing in Asia. I’m sure you’re already familiar with that. Actually, at this very moment, my digital double is giving a talk with Oslo Freedom Forum folks and Stanford on this particular regard, on how to react as democracies, and not overreact to the PRC’s position as an authoritarian intelligence — “AI” — superpower.

I know a bit more about that particular regard than hate speech. We can have a more in depth talk, but this is on everybody’s mind. Many Asian jurisdictions don’t choose the PRC vendors for 4G or 5G — or other tech stack — because they like authoritarianism, but rather because there’s a path dependency. It’s state subsidized by the PRC, so it’s inexpensive. The data norm? Well through the “rule by law”, it looks like a regulation behind it.

What, of course, they often forget is that there is no appeal process, and no accountability, which makes it as good as the great leader in the closest party branch. That’s the reality around here.

That is, I think, in everybody’s mind. When we do our usual end to end encryption code and things like that, we all have that in mind. Like when Zoom conveniently suspended people’s accounts because they said something that may offend people who don’t like Tiananmen being mentioned in the communication. Or that they can with a misconfiguration, according to them, reroute the traffic.

However, we’re not simply saying “Zoom is bad, ban them” — we’re saying we need to keep putting pressure on them on true end-to-end encryption, which they are now working on with the Keybase folks. That their web client also needs to go through independent audit — so we won’t have to install their desktop software.

Anyway, back to surveillance capitalism. There are ways to advocate, to regulate, and to litigate, but in this part of the world, we always assume that there is a state behind the surveillance capitalism if it looks like it goes through the physical route to any of the PRC territories, either de facto or du jour.

That’s a great reminder. Assume that the state is behind it. You mentioned audit, and actually, that’s something that I think would, we’d greatly benefit from your expertise, which is, one of the policy objectives that we’re advancing, trying to emphasize, to regulate as rulemakers, is that they have to have powers to audit and inspect the algorithms.

What does that actually look like? How do you actually do it? What are the questions you ask? What do you look for? I don’t think anyone has a great sense of that yet.

I do. That’s kind of my area.

I think that would be something that we’d greatly benefit from, which is how do you actually do that kind of audit and inspection?

A couple of things. Why I like Polis so much is that it’s machine learning, but it’s 20th century machine learning. It’s not deep. It use only very easy to explain, even to middle schooler, algorithms, principal component analysis, k means clustering.

Anyone with middle school mathematics can verify how it works using open data, versus a deep learning transformer infrastructure. That may be opaque, even to the people who code them. I think there’s a great difference between with AI with 20th century technologies that we can audit, and AI with 21st century technologies that the technologists themselves cannot audit.

I think there’s a lot to be said of using tried and true 20th century technologies. That’s the first part. Then the second part is about mathematical properties. If you design with the privacy in mind, with auditability in mind, there is, with an expense in computational cost, there is new materials in mathematics.

Zero knowledge proofs, fully homomorphic encryptions, split learning, if you have to use deep learning, and things like that that are provably private or provably secrecy keeping. Differential privacy is a good example.

You can specify a privacy budget, and it’s proven not to exceed that privacy budget. What I am trying to say is that, unlike the ad hoc audits that’s more like looking at a complex machinery and try to calculate its carbon emissions, do instead starting from the primitives, and design things to be emission free, or emission up to a certain epsilon, or even carbon capturing.

Anyway, and then, using only those primitives, then you can be sure that you have a zero waste or zero emission plants. I mean the industrial kind. I think this is the idea of privacy first design. If we don’t have that as a primitive in our regulations and able to enforce it through an independent data protection authority, then all hope is lost.

Ad hoc auditing is just like reverse engineering. You can’t really do it to any scale. Of course, very good white hat hackers can penetration test anything through reverse engineering, but we don’t want to make that the entry barrier toward making any publicly useful software.

Old, tried and true primitives, or deep learning and stuff, but with a good mathematical primitive, and build only using those primitives. These are the two pathways toward auditability by design. That’s why the end to end encryption thing is so important, because that’s one of the things that the public actually understand now.

It’s a good first test toward auditability of end to end and forward secrecy in communication, because that’s what people care about. It’s that they will not be eavesdropped in the future. That sounds wrong, like “Tenet.” Anyway, you know what I mean.

I was just going to add a couple weird antidotes. One is like where functional programming, functional machine learning, functional algorithms built, I think that that’s like…

Then I was just extend that, Alaphia, to one of the things that is a nugget that’s been in the back of my head with the Lando project, which you got a brief introduction to is how do you do functional legislating in the same way?

I’m totally bridging very bad analogies, but how do you ensure that there’s a clear primitive for the policy that you’re putting in place as well that’s grounded in some sort of research or evidence based?

That’s how it’s supposed to be. It doesn’t often work that way, but Reset is trying to come up with the policy component to both auditing and other, more responsible algorithmic machine learning models, how do you make sure that you can stand by both of them being built on real evidence based and primitives?

I think that that’s one aha. Then I was just going to say, Audrey, your position about the tried and true, I have been astounded in the last two to three years about how many folks who, like machine learning, data scientists, experts, have all come to that same conclusion.

Where they’re like, “We got into grad school, or we went to a company, and we were so motivated by coming up with the cool, new model that we didn’t think existed that was going to do something.” Then after they became more informed, they realized that actually, that was dangerous.

That there was so much tried and true within existing models, algorithms, and mathematical functions that did exist that were being underleveraged that we knew were safe. Now, they’re moving towards more of that.

It’s like, what’s your decision trees, for which things do you want to use that are tried and true? Rather than go straight to the cool, new — a bad analogy, but — sexy thing that would get everyone really excited, because it’s glittery and new. That could be very harmful.

That’s right.

I just really appreciate those comments.

Instead of tried and true, deep learning models could be untried and fake. [laughs] It’s remarkably easy to fake, actually, with all due respect. Sometimes it’s called overfitting.

I think the coprocessor thing, that code is law and law is code, is one of those easy to understand but hard to legislate ideas. The most success I’ve seen is all about almost like contract law, but also, there’s a part of the criminal code, like unsafe driving, that are very mechanical in nature. You can build a good dual of it in its digital twin.

For the kind of legislation that we’re looking at for Reset, which is more about algorithmic safety, I don’t know, fire inspections or whatever, things like that, I think there really need to be a clear delineation between the coprocessable, like embarrassingly parallel, embarrassingly called auditable part and the part that requires human judgment.

Then the human judgment part, just like any environment impact assessment, is, of course, still science and evidence based, but code does not substitute for human judgment there, because each environment protection agency may work on the same scientific principle — well I’m sure they work on the same scientific principle — but the norm around what’s pollution and what’s not, what’s tolerable and what’s not, what’s long term thinking… Some jurisdiction may think four years is long term thinking, and some jurisdiction thinks seven generations is long term thinking. That affects their decision on the environmental impact assessment, too.

Currently, I don’t think we can — there’s no app for that.

I really want to get you and Pia together to talk about, when do we start passing legislation that allows for more fluidity in those decisions? Where it’s like, it’s so hard for the subjective stuff to pass a law that results in a contract that the government has to hold.

With quadratic voting and more real time assessment of where is the populous at right now but still avoiding populism swings, you could end up in a situation where a change in temperament from society, enshrined by a law requiring this vote, changes the individual contracts that agencies, companies, and society, ultimately, abides by.

Now, we’re getting into a liquid democracy and all these other things, so I’m just going to step away from that radical. [laughs]

Maybe not liquid democracy — maybe natural personhood? It sounds more like natural personhood, like giving a river or a mountain a legal seat at a table, through re-presentational algorithms, essentially, to prove that this poor decision harms the river, so the river can sue for damage as a legal person.

That’s much easier to understand intuitively than liquid democracy — which is essentially about shifting allegiances, for lack of better term — albeit without a formal formula for earning trustworthiness.

Now the river is there, and the mountain is there. The digital commons are there. If they become, magically, avatars…

[laughs]

…that’s not a lot more fictional than how corporates become avatars.

Sorry. I lost you.

You can bring us back to how we can…

[laughs] I confess I understood about 30 percent of that.

You sowed the seeds that got us there. I think that that’s, where do we start? What are the organizations now? How do we make sure that…? I think, Audrey, you said it, that there’s not…The idea that a model for auditing civil society dependent algorithms, it doesn’t exist, is just wrong. It totally exists. There’s lots of people that are already involved with this.

There’s a lot of primitive models that exist. We just need to tap them. I think the interesting thing of, how do we… Where Reset could be potentially different is it’s not that we aren’t political and that politics and the softer human nature won’t decide what policies we pass.

What can we learn also being tech centered as well to make things that are a bit more grounded in some sort of equivalent, formal policy setting primitives that we’re talking about when we’re actually looking at the technology?

We look at the technology, and we fault it for not having known safe primitives that are causing harm. Can we do the same thing for some of the policies that we’re making? That’s just an open question.

Yeah. Maybe we can call them Presets.

[laughs] Presets. Reset’s presets.

Reset’s… [laughs] I like that.

I want to be mindful that we’re at time. We’re actually over time. I have a list of things that I just want to send you to make sure we’re completely on the same page as far as access, commitments, and affirmations, Audrey. I’ll send that to you.

Then if you have any questions at all for Alaphia or I, please feel free to just email us directly. Then we’ll go from there. Where you’re presumed to be all good, one of the first orders, I’ll be sending an email with an introduction to all the other councilmembers, so that you can know who your peers are and engage with them directly.

Then ultimately, we’ve got a first tranche of applications for your thoughts and comments on that are ready for you. We’ll introduce those as well.

Sure. I think our legal counsel starts working Tuesday. I will just run this through our legal counsel, who usually responds not in 30 seconds, but 30 minutes. [laughs] We’ll see how it goes.

OK, great, cool. I will send all of the agreements that we’re asking people to affirm to you immediately, so you can pass through and make sure that your access is OK to leverage. Thank you, Audrey. Thank you, Alaphia.

Thank you. Thank you both. Bye bye.

Yeah, thank you. It was a blast. Bye. Cheers. Bye.

Bye, y’all.