Data and Privacy with Alex "Sandy" Pentland

The area of privacy and data ownership is the main thing that I'm trying to sort of push on. We have things like GDPR, the California privacy protocols, things like that, but having rights over your data doesn't really do much for you. It's just like bunches of bits, right? What do you do? And we have this problem that some small number of organizations have huge amounts of data. Very unequal. And I think that the solution to that is the area that web science and people ought to think about, which is how can people take control of their data to get the medical service, the government, the et cetera that they want?

Welcome to this episode of Untangling the Web, a podcast of the Web Science Trust. I am Noshir Contractor, and I will be your host today. On this podcast, we bring in thought leaders to explore how the web is shaping society and how society in turn is shaping the web. Just a moment ago, you heard Professor Alex 'Sandy' Pentland talk about the intersection of data and privacy, which is just one of the areas he studies. Sandy is one of the most cited web scholars at the crossroads of web science, network science, and computational social science. He's a professor of media arts and sciences at MIT and directs the MIT Connection Science Research Initiative. He also helped create and direct the MIT Media Lab and Media Lab Asia in Mumbai, India. He heads MIT's Human Dynamics Group, which is one of 2 groups at MIT that is a member of the Web Science Trust global network of laboratories, and he co-leads the World Economic Forum Big Data and Personal Data Initiatives. In 2011, Forbes named Sandy as one of the 7 most powerful data scientists in the world, putting him in company with the then-CEO of Google, Larry Page. His work has pioneered organizational engineering, wearable computing, and modern biometrics, among other things. Welcome, Sandy.

Thank you for inviting me.

Delighted that you're able to join us here today. I want to start by asking you to tell us a little bit about how this incredible work that you've been doing on the web and how web is shaping society, to what extent and how did you get interested in looking at the web? Because I do know your scholarship even before you were looking at the web, you had already made a name for yourself in areas such as image recognition, et cetera. What got you interested in looking at the web and, and how did you get started in that?

Well, I've always been interested in human interaction, human perception. And, you know, sort of around the end of the '90s, I was setting up laboratories in India. We were living in India trying to set up sort of things like the Media Lab, but in India. And I noticed that the board of directors we had was terrible. And it wasn't that they weren't smart. It's just that they had too much charisma, too much personal force. And I got interested in how did that sort of non-linguistic, the sort of style of speaking, change decision-making. And then that was this sort of honest signals work that, that, you know, people are aware of. And turns out that you can do things like get early warnings of depression and things like that using this. But when mobile phones came along in the sort of mid-2000s, we started using those. And of course, that's part of the mobile web. That is the mobile web. And so suddenly we were looking at not just 2 or 3 people talking, but hundreds of people talking and even more. And I did experiments like looking at how communities make decisions. by looking at their face-to-face interaction, as well as their digital interaction. And of course, as the web exploded, and you got lots more video conferencing and, and things like that, it just became web science, right?

You were one of the early people who looked at the web as an opportunity to be able to study signals. Your book Honest Signals was very influential in making that point. And more generally, in social signals, the that you looked at, et cetera. One of the things that strikes me about your work is that you were amongst the first that looked at the web as an opportunity to be able to get data about human interaction and perceptions. And at the same time, your work has also been equally influential in recognizing the web as a source of concern in terms of privacy, et cetera. Can you share with us a little bit about how you straddle and how you reconcile these issues in your own scholarship and what you advise policymakers in this context? You

know, the core attitude I bring is one of science. I'd like to understand particularly human nature and how it is that we learn, make decisions, form society. And what you see very early on when you begin looking at this is, is that people form into cliques around topics. So there's your buddies, the people you talk to, and then People go off and explore to find other sorts of opportunities, and that, that exploration is critical for development of a community. But the interaction of the community, the separateness from the rest of society, is critical for people developing modes of operation or norms of operation. As you begin to look at this, you realize, oh my God, I can tell who this guy's friends are, and I can tell who the boss is. And I can tell that he broke up with this person over there. And so this brings all of the sort of classic privacy concerns. And it's actually a lot more concerning than, say, Twitter or something like that, or Facebook, because records from cell phones in particular, right, are where you actually were. So it's who you actually spent time with, not what you say about it. And it's extremely illuminating for people's not just personality, but their social structure, what they believe, where they're going, independent of, of what they say. So that led me to start the discussion at Davos that led to GDPR and have been developing technologies to be able to preserve the good parts, which is the communication, the community building, the exploration for new ideas, but without having the downside of Privacy and security risk.

Well, one of the things that you must have thought about is also the variation in norms around the world on what constitutes privacy. Clearly, you know, some of the things that we see in GDPR, for example, may not be very appealing to even a US audience and certainly would be different from what the Chinese or the Russians to just name 2. What are your thoughts about the extent to which these kinds of policies need to be responsive to specific countries and cultures? Or are you of the belief that there are certain human rights, basic fundamental principles of privacy that should be true for everyone on the planet?

There is a certain way of thinking about it, which is universal. It has to do with human nature, but there are variations. And so what I see is, is that privacy fundamentally is about individual freedom. Can you make— learn things, work with people, do things without any interference and also without other people knowing so that you feel more free to try things that you might not want to talk about in public? And, you know, this doesn't have to be dark. This is like, you know, I'm going to date this person, but I don't want that written on my record for the rest of my life because it may not work out, right? But there's a big axis in societies that has to do with individualism versus the social fabric. So in the United States, we're extremely towards the individualistic side. In Eastern culture, it's much more for the good of the group. But the issues are always the same. It's just there's a control knob that has to do with the value of the collective versus the value of the individual. And one of the things that people get wrong about this— I'm talking about the law as well as the large Discussions that people have is that a certain amount of clannishness is key for social support, for mental health. You've got to have your buds that support you. You've got to have people you bounce ideas off. If you don't have that, people go haywire. They really do. It's not a tiny thing. The biggest predictor of mental health is social interaction. That's why Solitary confinement is this horrible, horrible source of punishment. But the question is, is what do you mean social? Is it just your gang, a small group of people? Is it the people in your neighborhood? Is it, you know, your government? Different societies, different cultures have very different answers to that. But we shouldn't forget that there needs to be a circle of trust for a human to be healthy.

That's really a very interesting way of addressing these variations that we see around the world. You've talked a lot about phones and mobile phones and how that became very influential in getting you started looking at these kinds of signals. You've also been one of the pioneers of these intelligent wearable devices and sociometric badges comes to mind. Can you talk a little bit about what led you to that and what you see moving forward about the future of those kinds of devices as a way of tapping into social signals, but also maybe providing you some feedback on the basis of that?

Yeah, so I started the Wearables Group, which was really sort of the pioneering wearables in society type of group. So we had people running around with displays on their head and computers in their backpack and stuff like that. And it was a response to the realization that we were going to have wireless communication. At that point, we didn't have Wi-Fi or even cell phones, and that computation was going to get very small. And we wanted to experiment with what happens if your glasses had displays in them, what happens if things could whisper in your ear, etc., etc., etc. What quickly became clear is, first of all, things on your body have a social dimension that computer screens don't. You, you present yourself, you want to look attractive, you want to look credible, so what you wear is really important. And And also the second thing was the main thing you do in your physical world are social things, not information tasks. And so connecting with people better, being more responsive, being more of who you want to be, those are the things that really take off. And people keep forgetting this. So what we saw, what I've seen, is very slowly people are figuring out ways to incorporate this In social interactions, in the real texture of your life. One of the main barriers has been batteries, for God's sake. Turns out that battery technology is critical because you can't be like, you know, recharging things all the time. There were mistakes like Google Glass, which actually is a brilliant idea, but then by putting the camera on it and making it look all space-agey and George Jetson, people revolted. On the other hand, something that reminded me of your name when I met you again would be pretty awesome for most of us, right? Or directions so that you don't get lost. So what's happening with that is that it's being driven by health concerns now, of course, monitoring yourself for COVID or just staying healthy. And, and then also People being at home more, they're interested in things that maybe aren't exactly wearable but are very different formats from what we use now. And so we're going to see a lot of creativity in these wearable things, driven probably principally by health concerns.

Wearable devices has become now an extended part of the web, of the mobile web, if you may.

Wearable devices, payment devices, things for managing traffic and so forth. And of course, the downside of that is privacy, because now there's a lot more information about you, and cyberattack, because the surface for attack is growing exponentially, which means, you know, taking down all the payments or using the payment machine to get at the core bank, or all the sort of cyberattack stuff is going to increase by order of magnitude at least.

You published recently some research looking at blockchain transactions. Can you talk a little bit about how that research should be really important and salient for those interested in web science?

Yeah, so we published something, a book actually, MIT Press, called Trusted Data, and what it does is it lays out the architecture that you need to have to survive In this coming era of much greater cyberattack, IoT, and other sorts of problems—echo chambers, etc.—and the core thing is that the web was designed as a communication medium, getting bits from here to there. Not 100% reliable, but cheap, good, fast. But it was not designed as a transaction medium—the sort of thing where you know I pay you a little bit of money, you do a service, and if you screw up, I can sue you. And you know, I mean, that's sort of. Legally binding real transactions. It's terrible at it because you don't know if the messages get through there. They don't have standing in legal courts. And so what's happening now with IoT, Internet of Things, blockchain— don't think Bitcoin, think just ledgers that record stuff in a, in an immutable way, a very serious way— and AI also is you're getting the evolution of the web From a communication medium to a transaction medium. And you can get a picture of this on Amazon, right? You know, you like ask for things, it shows you, you click, you bought it, right? One click, all that. And that's because within that tiny walled garden, they can take care of all the security and who you are and payments and the contractual things. Imagine that was on the web as a whole. So, you know, you could say, design and build a house with one click because it would go off and find the architect and architect would put the plans in the mix and the computers would munge it and find a builder. I mean, you can imagine a world that is almost magical because things would happen reliably, auditable, traceable, legally binding, fair. You know, you can build that in there. To help do that, we've set up these sort of protocols and that you that have blockchain and stuff in them. And and I've gotten the European Union to adopt this as their core architecture for data, and also a number of large organizations like Fidelity now uses this architecture. Intuit uses an architecture like this. Other companies that handle a lot of our our life. One of the aspects of it that's really interesting is is that law is turning. To be a network or web science. Because as you get more things on the web, more of it has to do with legal rights and complaints and resolutions and so forth. So I just launched a thing called law.mit.edu, which is an alliance of law schools around the world to think about how law can make the transition to this sort of digital age. Because it's not obvious. I mean, there's so much that has to be human-centric, human-centered in law, human judgment, and that's under threat as things become more computerized. But yet they have to become more computerized to deal with this much more extensive digital environment. And so resolving that conflict is something that I wanted lawyers and computational people to think about together so that we don't end up in some horrible place.

That shows again why web science is so fundamentally an interdisciplinary initiative, and that requires us to think systemically across all disciplines to address and understand and enable the web as we know it now. You've already touched on many things that need more progress, but if I were to ask you, amongst the many things that you've been thinking about, What are some of the areas where you believe we have seen the least progress in web science, and that you see we need to be able to put much more emphasis in the near future?

Well, the area of privacy and data ownership is the main thing that I'm trying to sort of push on. So we have things like GDPR, the California privacy protocols, things like that. But having rights over your data doesn't really do much for you. It's just like bunches of bits, right? What do you do? And we have this problem that some small number of organizations have huge amounts of data. Very unequal. And I think that the solution to that is the area that web science and people ought to think about, which is how can people take control of their data to get the medical service, the government, the et cetera that they want? It's not a matter of money. Money is a distraction. It's really, are you getting the care, the government, the opportunities that you ought to do? All of this sort of anti-racism stuff resolves around this because this community wants to be treated fairly. Well, what does that even mean? How is it being treated? They don't have the data. What is beginning to happen now is you see data cooperatives, data unions forming. Where a community, a neighborhood, everybody doesn't give away their data. They set up something like a credit union that holds their data for them. So they still own their data. But now the data is in a place in this credit union, this data union, where they can analyze it and ask, are we getting the same medical services as those guys over there? Because they have data from lots of people. Data is the new resource like labor or capital. We have credit unions. Those came about as agricultural credit unions in the 1870s, 1860s. We have labor unions. Those came around the 1900s. Now we need data unions where groups of people pool their assets, which they have the legal right to do under GDPR and CPP. to be able to stand up for themselves. And I think that sort of forming of community, rebuilding trust around facts, around data, will have a transformative effect. Because as people work together to have freedom, fairness, power to thrive, they build trust with each other. It reconstructs a lot of what is damaged today. Communities should have the right more of self-determination, and that will help solve a lot of the problems out there. That's fundamentally A web science type of issue.

You raise a really good point that we have spent a lot of time and energy focusing on being able to have access to our own data. But what you're pointing out, Sandy, is that that is a necessary but not sufficient condition for us to reap the benefits collectively of owning our data.

Exactly. So, so at the beginning, I talked about this access of individualism to social fabric. Even in places like the US, which are very individualistic, the community that you're part of— and you get to choose what community— community that you're part of is necessary for your support, for you to thrive. And, and we tend not to recognize that. And I think the time is now where people are waking up to the fact that we need to pool assets locally among our community to be able to reinvent ourselves and And get what we need.

I think that's really on point. In closing, Sandy, I wonder if you could reflect a bit on what our current situation, whether you're thinking of COVID-19 or about the race issues that we've been confronting as a society and globally, in fact, how would this have been different if it were not for the web?

An interesting observation is that our large-scale non-web administrative structures, and I'm not talking about federal government only, all the way down, were unprepared and uncoordinated. And part of that is that different places are different. They have different culture. Texas is different than Massachusetts, I'll tell you. Montana is different than New York City. It really is not just culture, but the physical. And what's ended up happening is the more effective things have been things that were adaptive to local communities. And you see the same thing in the science that's going on. So there's an explosion of science to find vaccines and treatments for COVID, but none of that is NIH or CDC or WHO project. That's all local research groups working over the web to find new solutions that they didn't write any proposals for. Now, the NIH furnished those labs, so they've provided the ground and the tools, but the actual research direction, that's people grassroots pulling together, and the web is the thing that enables it. All right, so what they're doing is a dynamic learning network. So the web as learning network, and that is the way we're making lightning-speed progress on this problem, not big government programs. The big government programs set the infrastructure, just like the big government program helped invent the web, but it was these grassroots things that allow the sort of agile, locally adaptive exploration That's going to save us.

It again highlights the fact that even though the web was not originally invented with a particular set of activities in mind, certainly perhaps not the activities that we're experiencing right now, it has served us well in being able to, as you said, at the grassroots level dynamically be able to coordinate large groups of people and large communities to come together and mobilize in helping us address these challenges.

Yeah, I think maybe 2, 2 parting bits. One is, you know, the birth of the World Wide Web was research projects, distributed research projects that weren't directed from the top. They came from the bottom. And the second thing is, is what we're doing right now would have been impossible 5 years ago. We have constructed a web that supports All this sort of stuff just in time for the pandemic. It's it's it's crazy. If this had happened ten years ago, none of that zooming stuff would have been plausible. Lots of these other things just wouldn't have happened. And so, by luck or whatever, we find ourselves in a world that is is webified for better and for worse. But a lot of it. is for good because now we can learn faster, we can adapt, we can do things together in ways that just simply weren't possible until just very recently.

Thank you again, Sandy, for taking time to talk with us today and sharing your insights. You've been one of the pioneers of a lot of what has been happening across disciplines in helping understand the web, and we greatly value you taking time to talk to us today.

Thank you for thinking of me and enjoyed talking to you. Take care.

Untangling the Web is a production of the Web Science Trust. This episode was edited by Molly Lubbers. I am Noshir Contractor. You can find out more about our conversation today in the show notes. Thanks for listening.