Latest / ContraMinds Podcast - Unlocking Personal Growth and Professional Excellence / Adrian Gropper on The Intersection of Data, Technology, and Individual Rights (#027)

Transcript

Adrian Gropper on The Intersection of Data, Technology, and Individual Rights (#027) · ContraMinds Podcast - Unlocking Personal Growth and Professional Excellence

• 0:00
  SPEAKER_01: And what is a professional doing?
• 0:01
  SPEAKER_01: A professional is basically selling their time.
• 0:04
  SPEAKER_01: Now, if you happened to then get into a situation where you're selling more than your time, uh you're lucky, or maybe you're smart in a different way and you become a hedge fund manager or whatever, um, or an angel investor or something in between.
• 0:24
  SPEAKER_01: Uh but if you basically can find a way of developing individual expertise in a as a professional, then you could be an artist.
• 0:49
  SPEAKER_03: Hey everyone! Welcome to a new episode of Contra Minds.
• 0:53
  SPEAKER_03: Today we are in conversation with Adrian Cropper.
• 0:57
  SPEAKER_03: Adrian is the CTO of Patient Privacy Rights, a national organization representing 10.3 million patients.
• 1:04
  SPEAKER_03: He is among the foremost open data advocates in the United States.
• 1:09
  SPEAKER_03: Adrian's work focuses on solving critical problems of patient matching, cybersecurity, transparency, access, and dedication to family caregivers.
• 1:19
  SPEAKER_03: He holds a bachelor's degree in mechanical engineering from MIT and he is also an MP earning his degree from Harvard Medical School.
• 1:27
  SPEAKER_03: We had a wide-ranging conversation with Interim where he spoke to us about patient privacy, data platforms, and the chasm between science and the law.
• 1:36
  SPEAKER_03: So without further ado, here is our conversation with interim property.
• 1:49
  SPEAKER_02: Hi Adrian.
• 1:57
  SPEAKER_02: And uh thanks for taking time and talking to me.
• 2:00
  SPEAKER_02: My pleasure.
• 2:01
  SPEAKER_01: My pleasure indeed, Ran.
• 2:03
  SPEAKER_02: Adrian, I think uh I've never seen a designation or a purpose that uh probably somebody put out saying I am a I am a privacy entrepreneur, right?
• 2:15
  SPEAKER_02: So uh so therefore, uh, where did this thing around uh privacy come in and uh what is really the goal of what you want to achieve as a uh privacy entrepreneur?
• 2:30
  SPEAKER_01: Uh my career has been as a entrepreneur, medical device serial entrepreneur.
• 2:38
  SPEAKER_01: And uh towards maybe 15 years ago, I started a company as I had many times before.
• 2:49
  SPEAKER_01: Um and it failed.
• 2:52
  SPEAKER_01: Uh it failed very badly in the sense that nothing was sold, and you know, there was no just money went in and disappeared.
• 3:02
  SPEAKER_01: Uh what was interesting about the company, though, uh, was called MedCommons, is that uh six months after I started that business with a very senior uh uh colleague, uh partner of mine, also a classmate, uh Google and Microsoft both got into exactly the same business.
• 3:24
  SPEAKER_01: Um Microsoft uh called it uh Health Vault and Google called it Google Health, and we were called MedCommons.
• 3:33
  SPEAKER_01: And uh Google got out of the business within a year or so.
• 3:38
  SPEAKER_01: Microsoft stayed in that business for four or five more years, spent a billion dollars.
• 3:44
  SPEAKER_01: None of the three of us, uh, me with the shrimpy little startup, uh, and uh two of the largest companies in in the in the world, none of us anything came of it whatsoever.
• 3:59
  SPEAKER_01: Um not not an experiment, not a thank you, not something sold to anybody else.
• 4:05
  SPEAKER_01: Um, you know, uh Microsoft Health fault abandoned even the trademark after spending, like I say, all this money on it.
• 4:16
  SPEAKER_01: And um so what happened at that point was I I realized that I just wasn't gonna do what I had been doing as an entrepreneur before.
• 4:29
  SPEAKER_01: I was gonna focus on why this failed and try to figure out another way of uh commercializing what uh you know what patient-centered health records would look like.
• 4:42
  SPEAKER_01: And uh that's um I decided that was more important than making money or having investors that I first because obviously money didn't help.
• 4:52
  SPEAKER_01: If Google's and Microsoft's money didn't make it work, uh there was no way I was going to raise money to do this.
• 4:59
  SPEAKER_01: Now, incidentally, and by coincidence, um Microsoft had hired a foundation called Patient Privacy Rights, which I became uh Chief Technology Officer, a volunteer position, uh, but I became their chief technology officer.
• 5:16
  SPEAKER_01: Microsoft hired and paid for a large, very large accounting firm under the direction of patient privacy rights to write the privacy terms for health vault, because nobody today until then had tried to make a patient-centered or a patient-controlled health record.
• 5:36
  SPEAKER_01: There was no uh, they needed expertise in how healthcare laws, healthcare regulations in the US, and privacy.
• 5:45
  SPEAKER_01: And so when I made this decision to work in um in the non-institutional context, uh the patient doctor context of medical privacy, and also made the decision to join uh with Deborah Peel, the founder of patient privacy rights, I then discovered I inherited this intellectual property uh that Microsoft had effectively paid for and donated.
• 6:14
  SPEAKER_01: Uh, they just wanted the result, not the not the rights to it.
• 6:18
  SPEAKER_01: As again, in an open source sense that you would expect for anything involving security or privacy nowadays.
• 6:25
  SPEAKER_01: And I'll stop there, but that's basically uh how, and so I'm partly an entrepreneur with my project HI of one, and and partly I'm a volunteer uh CTO of a nonprofit, uh basically a charity.
• 6:40
  SPEAKER_02: So uh being somebody who deeply believes in uh privacy and uh individual rights, uh, I saw you talk about the fact that uh an unprotected uh you know individual right and lack of privacy is not good for society, and that's not something that's good because it doesn't respect the individual.
• 7:05
  SPEAKER_02: Can you can you kind of uh you know explain why why and what was your thought behind such a point that you made?
• 7:15
  SPEAKER_01: Uh well there there's two kinds of respect and uh very important um situations.
• 7:26
  SPEAKER_01: Um one of them think of it as diversity.
• 7:30
  SPEAKER_01: Um you know, most societies, certainly not around the world, have learned to that it is good for the society itself uh to respect the diversity of the people in it.
• 7:45
  SPEAKER_01: And if you have too much institutional control, whether it's political or commercial, um you tend to reduce diversity in what is going on.
• 8:03
  SPEAKER_01: Um and so privacy uh basically allows for diversity.
• 8:10
  SPEAKER_01: Very simply put, it'll it avoids self-censorship, it avoid it it allows different points of view to evolve or different practices, you know, whether sexual or educational or whatever, work associations in general.
• 8:29
  SPEAKER_01: So um to have a resilient society, you need diversity.
• 8:34
  SPEAKER_01: Um and uh privacy is core to that ability.
• 8:40
  SPEAKER_01: Um the second um aspect of it is uh is very different, is very technical, um, having not to do with technology itself, which I'm an engineer and I very much do that, but having to do with uh regulatory practice or how you regulate technology, how you regulate medical practice.
• 9:05
  SPEAKER_01: And it turns out that again, on a very large scale, uh regulation follows these two models.
• 9:14
  SPEAKER_01: Uh one of them is centralized, and the government basically, through uh an agency that tells you this drug is approved, this device is approved, uh, this hospital is approved.
• 9:27
  SPEAKER_01: Um, these things are regulated centrally across the entire country and sometimes even more globally.
• 9:34
  SPEAKER_01: Uh but then in all of these countries and all of these societies, you also have the regulatory model of a physician, which I'm trained as, though I don't practice medicine.
• 9:44
  SPEAKER_01: And or you have it for things like uh defense lawyers, you know, lawyers.
• 9:51
  SPEAKER_01: Um you you have the practice of letting professions regulate themselves with licensing to some extent, but um in order to again, why do you do that?
• 10:06
  SPEAKER_01: You do that in order to give people a choice of who their advocates are, so that an individual who may not know about healthcare or may not know about the law that is facing the large institutions, even if they're regulated, you know, facing the state, facing a corporation, facing an employer, can go and hire an advocate that they choose, because the individual cannot choose the large entity.
• 10:32
  SPEAKER_01: You know, there's only one Apple and Google, you know, there's only one Microsoft, there's only one state where you live.
• 10:40
  SPEAKER_01: Um so uh because of again my background, which I can go into, but it doesn't matter, I I have lived uh technically and business-wise between these two worlds.
• 10:53
  SPEAKER_01: One of them as I was making these device companies, uh, some of them more successful than others, and the other one where I was watching all of my colleagues that went to school and were in medical practice, my children, etc., sort of in this other regulatory mode.
• 11:09
  SPEAKER_01: And uh that became the innovation, if you would, or the the goal, the mission, whatever you want to call it, to my work in the last 10 to 15 years, is to have this balance so that you don't have just AI replace the doctor, but in the process it's all owned by IBM.
• 11:33
  SPEAKER_01: Um instead of, right?
• 11:35
  SPEAKER_01: So, you know, how does the AI become more like the way you train an engineer or the way you send a uh a student to medical school so that they end up in control of their tools, not being and in control of their individual practices, and if they go astray, yeah, they get, you know, they lose their license or whatever.
• 11:56
  SPEAKER_01: Anyway, so uh this idea of both diversity on the one side, again, in a societal sense, and completely differently, this difference between institutional regulation and professional regulation.
• 12:13
  SPEAKER_02: Some people don't trust companies.
• 12:16
  SPEAKER_02: Some people don't trust governments, and therefore, whom do they uh trust to share their data with?
• 12:25
  SPEAKER_01: So, how does how does this work?
• 12:28
  SPEAKER_01: Well, do you how much data do you share with your wife?
• 12:34
  SPEAKER_01: Yeah, okay, most of it, presumably, but not all.
• 12:40
  SPEAKER_01: Um how much data do you share with your close friends?
• 12:45
  SPEAKER_01: How much data do you share with a club or a church that you join voluntarily that you choose?
• 12:56
  SPEAKER_01: So what what's the common denominator there?
• 13:00
  SPEAKER_01: It's the fact that you chose your spouse and they chose you.
• 13:07
  SPEAKER_01: Same thing with your fraternity brothers or the club or whatever.
• 13:13
  SPEAKER_01: Uh people uh learn to depend on each other uh to share their interests, and then the technical word is delegate, delegate certain capabilities to a spouse so that you don't have to make every single decision yourself.
• 13:37
  SPEAKER_01: Now, in the limit, you choose a doctor if you're sick.
• 13:41
  SPEAKER_01: You don't go and start reading, you know, Wikipedia.
• 13:45
  SPEAKER_01: Uh some people do, but um you and you don't necessarily just go to the hospital and assume that they'll take care of you.
• 13:57
  SPEAKER_01: Uh sometimes you have to, you're forced to, but it's not a pleasant situation.
• 14:01
  SPEAKER_01: What you do is you again, you choose a delegate.
• 14:04
  SPEAKER_01: So whether so the most uh important thing in terms of scaling the use and the application, whether it's clinical or research or public policy, which are three very different domains, of health-related data is that you be able to have an intermediary that you choose.
• 14:29
  SPEAKER_01: Or many intermediaries that you choose.
• 14:31
  SPEAKER_01: Um, you know, so in in in many families the the man leaves these decisions, at least in the US, to the woman.
• 14:38
  SPEAKER_01: Um you know, it's to a large extent, and uh and etc.
• 14:45
  SPEAKER_01: So that's um that's a principle around trust.
• 14:55
  SPEAKER_01: Don't force people to trust the government, don't force me to trust Apple, even though I do trust Apple.
• 15:02
  SPEAKER_01: I trust them more than Google, less than Facebook.
• 15:05
  SPEAKER_01: I mean, uh, you know, uh I trust Google more than Facebook.
• 15:09
  SPEAKER_01: I don't deal with Facebook.
• 15:10
  SPEAKER_01: But you see what I'm saying?
• 15:12
  SPEAKER_01: Don't force me to trust somebody that has built up a barrier to entry based on the network effect, right?
• 15:20
  SPEAKER_01: Whether it's a social platform or a government platform or a technology platform.
• 15:27
  SPEAKER_01: Um so basically, yeah, introduce choice, and I mean, you know, people who study ethics, which I don't, I've never studied ethics, basically that one definition, Heinzweig Forster actually used this of ethical behavior, and I'm quoting him more or less, is I will always act so as to maximize the number of meaningful choices.
• 15:55
  SPEAKER_01: Right?
• 15:56
  SPEAKER_01: If you if you're c trying to constrain somebody into your platform, your business, your walled garden, that inherently is not ethical behavior.
• 16:07
  SPEAKER_01: Now, nobody holds companies to that that strong and ethical thing.
• 16:14
  SPEAKER_01: But if you think about it, if you're trying to be ethical with your friends, with your colleagues, etc., you would always give them that choice.
• 16:26
  SPEAKER_01: It's like they say, if you love something, let it go.
• 16:29
  SPEAKER_01: If it comes back to you, you know you've under, you you know that it's love.
• 16:35
  SPEAKER_01: Anyway, so yeah, uh that's that's how I talk about trust.
• 16:41
  SPEAKER_02: So uh so when you really talk about uh you know an intermediary who will give you the choice uh of interoperability of your data, right?
• 16:51
  SPEAKER_01: Uh well, no, no, no.
• 16:53
  SPEAKER_01: Uh the intermediate you have a choice of intermediaries, just like you have a choice of doctors uh or a choice of lawyers or a choice of spouses, you you have a choice of intermediary, and then that intermediary, either by social habit, like in a club or a gang or whatever, or by contract, in the case of a doctor or a lawyer, is bound to learn your policies and execute them for your good.
• 17:25
  SPEAKER_01: So you um so you get to choose the intermediary.
• 17:30
  SPEAKER_01: In choosing the intermediary, they suggest policies that they come with.
• 17:36
  SPEAKER_01: You know, it's like you know, you find a spouse and they well, you know, I'm this particular, you know, I'm Jewish, okay, you know, so you you then have chosen a set of policies, whether you're Jewish or not, and you understand that they because they are your intermediary, they will reflect the fact that you're whatever, uh Buddhist.
• 18:01
  SPEAKER_01: Um but uh you agree that if you don't tell them a policy, they will use theirs by default.
• 18:12
  SPEAKER_01: Because again, you haven't.
• 18:14
  SPEAKER_01: So um so what I'm saying is, and and this is pretty much the discovery of or the innovation of my my work over these last decade plus, is that I separate out the policies that are either individually controlled and then delegated or delegated from.
• 18:38
  SPEAKER_01: It doesn't matter which direction, right?
• 18:40
  SPEAKER_01: As long as you are in control of the policies, in other words, your wife doesn't do something, you say, don't do that.
• 18:46
  SPEAKER_01: I don't care whether you're what your religion says, you know, that's not me.
• 18:51
  SPEAKER_01: And then and then this aspect of choosing the intermediaries because that's what keeps them honest.
• 18:57
  SPEAKER_03: Yeah, so I I just have a follow-up question because I think one of the there are multiple threads which are being woven together in this conversation.
• 19:05
  SPEAKER_03: We're talking about individual rights, we're talking about data, we're talking about privacy, we're talking about technology, we're talking about entrepreneurship.
• 19:13
  SPEAKER_03: So it's a multi-threaded conversation that's being uh discussed right here.
• 19:18
  SPEAKER_03: But uh on the other hand, I would also say that there must have been an inflection point for you to start questioning the status quo for you to take the path that you have taken, right?
• 19:28
  SPEAKER_01: So when you start thinking about it was the failure of my previous business, of my last sort of for-profit with investors and capital looking stuff, yeah.
• 19:39
  SPEAKER_01: Um in a in a business that in and of itself was again interesting enough to have brought in Microsoft and Google to do exactly the same thing.
• 19:50
  SPEAKER_01: That was the that's what caused it.
• 19:54
  SPEAKER_03: Yeah, but I'm I'm sort of looking at it at a little at a at a more at a more base, at a philosophical level, which is you have the way that things are being done.
• 20:03
  SPEAKER_03: So when you were at medical school, for example, there was a standard view on how the on how the practice is to be conducted, how the relationships are to be had with the pharma companies, with the research community, etc.
• 20:14
  SPEAKER_03: Right?
• 20:14
  SPEAKER_03: And then there must have been an inflection point where you must have asked yourself, look, something is deeply wrong here, right?
• 20:22
  SPEAKER_03: Like if I go and I'm not getting the care that I need, if somebody's not subsidizing the money that I don't have for the drugs I need.
• 20:30
  SPEAKER_03: So where did that philosophical sort of uh the inflection point actually come where you decided that look, I might be a formally trained physician, uh, but I'm not going to go into the practice, but I'm going to do what's right for my uh patients or the customer.
• 20:44
  SPEAKER_03: Because then you're playing with the market.
• 20:46
  SPEAKER_01: You're you're you're absolutely right.
• 20:49
  SPEAKER_01: I um my most successful uh business um sold things to hospitals, uh the the things were software that manages uh radiology images, digital software uh to manage.
• 21:06
  SPEAKER_01: It's called PAX, P A C S.
• 21:08
  SPEAKER_01: It doesn't stand for picture archiving and communication system.
• 21:12
  SPEAKER_01: And because um I had the opportunity to start this company at Massachusetts General Hospital, which is the largest, maybe the most famous or one of the three most famous hospitals in the in the world, it's also the largest radiology training program in the in the country.
• 21:34
  SPEAKER_01: Um you know, they train more residents, and um it's a big moneymaker uh for the hospital because they do a lot of procedures, etc.
• 21:46
  SPEAKER_01: And I started this business and it was very successful.
• 21:50
  SPEAKER_01: And I was selling this technology to other fancy academic hospitals um because it was early.
• 22:00
  SPEAKER_01: And it was again designed out of this luminary institution.
• 22:05
  SPEAKER_01: And as I was doing this, the costs of healthcare and the lack, the disparities, the lack of access in the U.S.
• 22:16
  SPEAKER_01: in particular to healthcare were just going through the roof.
• 22:21
  SPEAKER_01: You know, I was watching the institutionalization.
• 22:25
  SPEAKER_01: So what started out as being just digitizing medical practice, the practice of surgeons that consult radiologists in this hospital setting, a lot of telemedicine work around the world as well, and stuff like that, just became clearly more focused on administrative activity than on patient care, or for that matter, medical doctor practice.
• 22:56
  SPEAKER_01: And it was very unpleasant to me because that I didn't want to see that.
• 23:05
  SPEAKER_01: I did not want my work, you know, when I went to medical school, my first device, the first thing company that I tried to start, it never got out of my living room, was to try and make an ultrasound machine that primary care doctors could use themselves because it was safe.
• 23:22
  SPEAKER_01: And if they could understand the images, because I had a big enough computer to help the images become understandable to a primary care doctor, then they wouldn't have to have a radiologist.
• 23:34
  SPEAKER_01: So all my career, I was always focused on building tools for the individual practitioner.
• 23:41
  SPEAKER_01: My success was in building tools for the largest institution in the country by some measure, and the largest academic.
• 23:50
  SPEAKER_01: And to me, that was, even though I was successful, that was a very big shock.
• 23:58
  SPEAKER_01: And and that's kind of what got me started saying my next company, the one that failed, uh, because now I had money, I thought I was smart, I wasn't that smart.
• 24:10
  SPEAKER_01: Um you know, the one that failed, you know, I said, well, this one is going to be between patients and doctors.
• 24:18
  SPEAKER_01: Anyway, so that's probably what did it.
• 24:22
  SPEAKER_01: But it happened over time, Vignesh.
• 24:25
  SPEAKER_01: It it wasn't something that happened on one day.
• 24:28
  SPEAKER_03: Absolutely.
• 24:29
  SPEAKER_03: Um, and when you began exploring these concepts a little deeper as you move post uh medical school and you know, when you were really exploring, hey, whether these, you know, there's always, you know, for somebody living in India, I'm also aware that there is a constant debate on the patient charter of rights or a patient bill of rights constantly in the United States.
• 24:52
  SPEAKER_03: And I think nothing's come.
• 24:54
  SPEAKER_03: I mean, it's instead I I have heard of it, living a half-a-world way.
• 24:59
  SPEAKER_03: Uh, but my question to you is when you are exploring these relationships on how do you really protect data rights, how to actually protect patient rights, how do you allow for choice in the medical practice?
• 25:12
  SPEAKER_03: Uh, but these are fundamentally ethical questions if you think about it.
• 25:16
  SPEAKER_03: And uh the solutions may be technological, but the questions are ethical in my opinion.
• 25:21
  SPEAKER_03: So when you have these questions, knowing the fact that you know you come from esteemed institutions such as MIT and Harvard, what was the conversational responses like to these very basic questions that you were asking your peers and your professors and stuff like that?
• 25:34
  SPEAKER_03: What is that experience like?
• 25:36
  SPEAKER_01: Uh well, with respect to data, you know, and how data is used, which is um the um the point is that there are two um uses, two kinds of uses of personal data, um, or misuses of personal data, right?
• 26:01
  SPEAKER_01: Uh there's identity theft, in other words, uh misuse of personal data that damages you as an individual.
• 26:11
  SPEAKER_01: And then there is political manipulation or lack of transparency in practice.
• 26:20
  SPEAKER_01: That, for instance, in the United States, means that there's absolutely no way to tell who's a good hospital, who's a bad hospital, who's a good drug, who's a bad drug, almost by design, who's a good doctor.
• 26:32
  SPEAKER_01: You know, the whole system is designed to be opaque as much as possible.
• 26:40
  SPEAKER_01: And so you have stupid situations, you know, like US News and World Report, a magazine rates hospitals.
• 26:48
  SPEAKER_01: And um you have uh, you know, maybe independent international organizations, not US-based like the Cochrane, that might rate whether a drug is good or bad compared to another drug.
• 27:02
  SPEAKER_01: Um and you have almost nothing for doctors.
• 27:06
  SPEAKER_01: Doctors' reputation um nowadays, not like 30, 40 years ago, is entirely based on where they work.
• 27:13
  SPEAKER_01: Do they work at Harvard, they're a good doctor.
• 27:14
  SPEAKER_01: They all work at Harvard, they're not a good doctor.
• 27:16
  SPEAKER_01: It's like saying, you know.
• 27:18
  SPEAKER_01: Um, and so um what happens is that when you see the huge healthcare problems in the United States, by huge I mean that when you're spending 40% or so of a three or four trillion dollar budget on healthcare, that money isn't going to schools, isn't going to parks, isn't going to, you know, whatever.
• 27:48
  SPEAKER_01: Uh that money is coming from somewhere.
• 27:51
  SPEAKER_01: Uh and it is wasted.
• 27:54
  SPEAKER_01: That is a social harm of the first order.
• 27:59
  SPEAKER_01: Um you know, you you can be as rich as the US, but you see the problems that that there are with discrimination, with disparities, with poverty and so much other country and stuff.
• 28:11
  SPEAKER_01: When when you're taking that much money and and siphoning it into what's basically a mafia, because a mafia controls prices.
• 28:23
  SPEAKER_01: Well, they provide the services, right?
• 28:25
  SPEAKER_01: You know, they pick up the garbage or they replace the police for the most part.
• 28:30
  SPEAKER_01: But what do they do?
• 28:31
  SPEAKER_01: They they charge you double for that.
• 28:34
  SPEAKER_01: And how do they do that?
• 28:36
  SPEAKER_01: Well, they do that because everything they do is secret.
• 28:38
  SPEAKER_01: So when you let government do those services, then for the most part, in most places, there's transparency.
• 28:46
  SPEAKER_01: And whether you think they're charging too much or too little, at least you can see what their budget is, who's working for how much.
• 28:54
  SPEAKER_01: When you get the mafia to do it, the model is they still do the same job, but everything is secret.
• 29:03
  SPEAKER_01: And the result is that you're paying twice as much for as much police presence or as much garbage pickup or whatever it is that you're getting.
• 29:12
  SPEAKER_01: And so what I'm saying is that when it comes to actually doing the technical work, the engineering and the policy work that I talk about, that it you know, um, what I do is I tend to ignore the individual harms uh in my work.
• 29:32
  SPEAKER_01: In other words, I I don't I I'm sorry that your identity was stolen and you then had to spend a year unwinding and creating new bank accounts, or you know, um, or I'm sorry that you were discriminated against because somebody found out that you had a mental health problem and it was none of their business, right?
• 29:51
  SPEAKER_01: Uh that's too bad.
• 29:54
  SPEAKER_01: And I I I hope you solve it quickly, but it is not the focus of my work.
• 30:00
  SPEAKER_01: The focus of my work is the way data is used to either train doctors or nowadays train AI and regulate AI, which is a big problem, and how it is used to make policy, and how it is used to do research.
• 30:16
  SPEAKER_01: Is the research going to result in a for-profit thing that then competes with another for-profit thing, and neither of them necessarily is transparent about which one is better than the other?
• 30:28
  SPEAKER_01: That kind of stuff.
• 30:29
  SPEAKER_01: So, anyway, I'll stop there.
• 30:31
  SPEAKER_01: But um that uh that's a distinction that's at the core of my work, both with patient privacy rights, and again, you know, it's not that we ignore the individual harms, but my work is primarily focused on making sure that society, either in the public health or the research or the clinical sense, benefits from the data.
• 30:55
  SPEAKER_02: So uh, Adrian, I have a uh, you know, one of the uh things that I read is you were one of the co-authors, and uh, you know, you contributed to a white paper on the Indian government doing a health ID, right?
• 31:12
  SPEAKER_02: And uh so there was a framework that was presented, and uh, you said the mistakes that have happened in the developed countries around the health records need not be uh you know mirrored in a country which is trying to build its health ID for the first time.
• 31:30
  SPEAKER_02: So, what are the mistakes that were done, and how do you really ensure that the new health ID that's being built in a country like India, what are the uh you know do's and don'ts that you believe are more fundamental in getting this right?
• 31:46
  SPEAKER_01: So, the mistake that was done in both the European model as well as the US model for health data management, is to build the software, the thing that people interact with, the doctors, the patients to some extent, but mostly the doctors and the nurses, of course, and uh to build that software first and then uh try to add interoperability after the software was dominant.
• 32:21
  SPEAKER_01: Right now, the model by which it was done, you know, in many European models and around the world, you know, Singapore, whatever, uh Hong Kong, I guess, et cetera, you know, people would build national software systems.
• 32:37
  SPEAKER_01: Even in the US, you know, they had like the VA system was a public system.
• 32:44
  SPEAKER_01: Um, and in other countries, people would have private enterprise, so you might have five, six, ten large software vendors.
• 32:54
  SPEAKER_01: Um the problem in both cases is that the system becomes very rigid.
• 33:03
  SPEAKER_01: You cannot introduce innovation in any really effective way.
• 33:10
  SPEAKER_01: Um because whether they're well-meaning or not, and in many cases they're not well-meaning, you know, in the commercial sense, they just want to lock the system in and make sure that it's the barrier to exit to going to a different vendor is very high.
• 33:28
  SPEAKER_01: And in the case of a single supplier, like the government provides the software for the entire country, in England, for example, I in different countries, then in effect the whole process of innovation about how you practice medicine, how you do healthcare and health insurance as well, um, becomes basically a political or or centrally planned affair.
• 33:57
  SPEAKER_01: And uh India was a huge opportunity because it was what we call a green field, right?
• 34:05
  SPEAKER_01: There there was no, there was neither policy, of course, yet, there was, but there also wasn't any software in place or any standards for how the software would be hooked up.
• 34:20
  SPEAKER_01: And so my contribution uh was to basically say the principle should be that the government specifies the interface to the software, but not what the software does.
• 34:37
  SPEAKER_01: So, in other words, you you you regulate only how and what is being exchanged from one institution or one software package to another and let people do whatever they want.
• 34:56
  SPEAKER_01: Um and you can only do that if you're talking about a green field situation with the kind of technical capabilities and government capabilities that India has.
• 35:07
  SPEAKER_01: You can't do that going into some random, you know, African country where they have neither the technical nor the government.
• 35:16
  SPEAKER_01: Uh, you know, what do those people do?
• 35:18
  SPEAKER_01: Well, what those people do is they hire one contractor, you know, they pick one cell phone company and say, okay, you get the contract for the country, we'll regulate you, right?
• 35:28
  SPEAKER_01: Because they really don't have a choice.
• 35:30
  SPEAKER_01: But India has a choice because India is governed, you know, uh, you know, more or less effectively, and uh and they have the technical capability to do that.
• 35:42
  SPEAKER_01: And that was basically the foundation of my suggestion.
• 35:45
  SPEAKER_01: Now, whether what's happened, you know, eight years later, seven years later reflects that, I think it does to a large extent, uh, but we still don't know, right?
• 35:57
  SPEAKER_01: We it the that game is still being played.
• 36:00
  SPEAKER_01: Uh we don't know the final score.
• 36:03
  SPEAKER_02: So, so the real question, therefore, is uh, you know, you have today, uh, you know, uh, I would call it uh, you know, we are a wearable society, right?
• 36:14
  SPEAKER_02: I have my health, uh I have my health band, I have my you know, cell phone which tracks my uh you know fitness, so there are so many fitness apps.
• 36:25
  SPEAKER_02: So uh in this conflict of context of a health ID, and uh when all of this get connected to a single health ID which then has my health records, okay.
• 36:37
  SPEAKER_02: Who owns this data?
• 36:39
  SPEAKER_02: Because the device that I buy also says that I own the data, and the wearable company also says that's my data, while it's actually my data that defines the health uh and the medication that I need.
• 36:57
  SPEAKER_02: Therefore, how do you really govern that?
• 37:01
  SPEAKER_01: Well, first of all, let's agree on a definition, uh, which is own, you know, when when uh there's only one working definition of owning data and it has nothing to do with you as an individual, as an identity.
• 37:17
  SPEAKER_01: Owning data simply means that I have a piece of data and I can either delete it or take it offline, right?
• 37:26
  SPEAKER_01: So if you try to use any other definition of ownership of personal data, you you just introduce confusion.
• 37:38
  SPEAKER_01: Because the doctor says I own the data because you know I created it.
• 37:45
  SPEAKER_01: Uh it's maybe your health record, but it's my work.
• 37:49
  SPEAKER_01: The device manufacturer wants to own the data if they can, uh, unless, again, they're Apple and they say we don't see the data, therefore we can't own it, right?
• 38:01
  SPEAKER_01: Um, and uh and so on and so forth.
• 38:06
  SPEAKER_01: So ownership is a you and you know, there's also the issue of property rights.
• 38:11
  SPEAKER_01: You know, uh governments, uh societies have this idea of ownership of real estate or cars, right?
• 38:18
  SPEAKER_01: Because you can own and transfer property rights.
• 38:21
  SPEAKER_01: Um, you can't do that for personal data, because personal data is not a uh rivalrous good, it's not a limited resource.
• 38:31
  SPEAKER_01: It's like educate personal data is like education, right?
• 38:35
  SPEAKER_01: If you share your mathematical formula or your medical um protocol, you're better off.
• 38:44
  SPEAKER_01: People improve it, people teach what you want them to learn.
• 38:49
  SPEAKER_01: And that's not ownership of anything.
• 38:52
  SPEAKER_01: You know, you don't own your education, you own your degree, uh, but you don't own your education.
• 38:58
  SPEAKER_01: And it makes no sense to own education.
• 39:00
  SPEAKER_01: Okay, so now back to wearables or basically anything else that's personal data, whether it's generated by your insurance company, by your doctor, by a lab, by a wearable.
• 39:10
  SPEAKER_01: Uh the principle that I would suggest we have to build on is not around the data at all, but rather around the inferences, the learning, the education, the science that that data enables.
• 39:31
  SPEAKER_01: Now, some of those inferences are personal to you.
• 39:34
  SPEAKER_01: So, for example, if you go on an airplane booking site and they decide, they infer from the kind of web browser and the neighborhood where you z where your IP address is in that you have more or less money because their AI has figured out how to guess accurately about your demographics, then you're going to be shown a very different set of airline choices and fares than the person next to you, unless there's a law against that, which there generally isn't.
• 40:12
  SPEAKER_01: Now that that's an extreme case, you know, differential pricing.
• 40:17
  SPEAKER_01: Um, but so in general, um when you end up accessing personal data, where again, wherever it comes from, it doesn't matter.
• 40:28
  SPEAKER_01: Wearables are just a very good clear example.
• 40:33
  SPEAKER_01: Um the question we should be asking as a modern society is not who owns that data, but who benefits from the learning, from the machine learning, the inference learning, the artificial intelligence, because when you look forward 10 years, not more, maybe much less, you see that there's basically almost nothing that doctors do today that won't be done by the AI.
• 41:07
  SPEAKER_01: And I'm not saying the doctors are going to go out of business.
• 41:10
  SPEAKER_01: The reason I'm saying doctors aren't going to go out of business is because of the regulatory point I was making before.
• 41:16
  SPEAKER_01: You're you, you know, I don't know if you've all watched Blade Runner, the movie Blade Runner.
• 41:20
  SPEAKER_01: You know, you have Tyrell Corporation, right?
• 41:23
  SPEAKER_01: Tyrell Corporation is what happens when you allow centralized regulation of technology, right?
• 41:34
  SPEAKER_01: Um you and again, you can think that Apple is great in many respects, particularly privacy, but if you allow Apple to make all the rules in India, in the Netherlands these days, they're talking about in the US, again, you end up with Tyrell Corporation in the long run.
• 41:53
  SPEAKER_01: So we have to think in terms of the AI, and then in terms of the professions, your delegates, the people you choose as being the regulatory model.
• 42:08
  SPEAKER_01: And for the most part, then you know how to regulate the AI.
• 42:14
  SPEAKER_01: Because you're if the AI is open source and managed as if it was education rather than as if it was a medical device, which has been my business my entire career, now we have a path forward.
• 42:27
  SPEAKER_03: So so, how much more complicated do you expect things to become, both from a technology standpoint, from a regulation standpoint, uh, when you have products like uh Neuralink coming into the market?
• 42:41
  SPEAKER_03: Because it's one thing for your Apple Watch to collect data and to protect that data.
• 42:45
  SPEAKER_03: It's another thing to actually have uh, you know, a brain machine interface literally sitting right on top of your head collecting data.
• 42:54
  SPEAKER_03: So, how much worse or how much better or how much wiser can we become?
• 43:00
  SPEAKER_01: Uh, I don't know how much time we we are into this.
• 43:04
  SPEAKER_01: Uh that is a uh that's a very interesting and it gets technical uh question, um, because the the management of complexity um in the face of Moore's law and other technological, you know, uh not just Moore's law, but uh the power of networks, right?
• 43:30
  SPEAKER_01: The fact that networks become more valuable, the fact that your data from all those things that Swami was mentioning becomes more valuable as it's aggregated.
• 43:40
  SPEAKER_01: You know, I, for instance, signed up for 23andMe uh many years ago, again, because it was an experiment for me and my family to see what they did.
• 43:48
  SPEAKER_01: But what am I doing with 23andMe five years later?
• 43:52
  SPEAKER_01: They're begging me for more data about my phenotype, right?
• 43:56
  SPEAKER_01: About my relatives, about my lifestyle.
• 44:00
  SPEAKER_01: And I don't give them any.
• 44:02
  SPEAKER_01: In other words, I'm not allowing 23andMe that I basically paid for as a lab and to do what a lab is supposed to do, which is to tell you this is a normal value, this is a not normal value.
• 44:15
  SPEAKER_01: I didn't pay them to aggregate data about me.
• 44:18
  SPEAKER_01: I don't care what they're going to do with it.
• 44:20
  SPEAKER_01: So it's easy.
• 44:21
  SPEAKER_01: I just don't give them the data.
• 44:23
  SPEAKER_01: And if they don't have an obvious way of buying it from some data broker or getting it some other way by spying on me, then that's the end of the story.
• 44:32
  SPEAKER_01: So what I'm saying is one aspect of the complexity question that you just asked going forward is what does it look like to aggregate data?
• 44:45
  SPEAKER_01: Like, for instance, aggregating data into my iPhone because it's free and they do a good job, and Apple will collect the stuff.
• 44:53
  SPEAKER_01: And Apple says we don't see the data, so you don't have to worry.
• 44:56
  SPEAKER_01: We just will never see the data.
• 44:58
  SPEAKER_01: If it's in the cloud, it's encrypted, and we don't have to keep.
• 45:01
  SPEAKER_01: And then uh the other question I mentioned a few minutes ago is how is how are you aggregating the learning from the data?
• 45:11
  SPEAKER_01: So that's a separate question in terms of complexity.
• 45:15
  SPEAKER_01: Is the learning being aggregated in an open source way, the way we aggregate learning in math or physics or medicine, traditional medicine?
• 45:23
  SPEAKER_01: Or is it being aggregated in a proprietary way that we really don't know and will not be able to regulate?
• 45:30
  SPEAKER_01: So the complexity that you're talking about will be limited if the learning, if the machine learning is not open source, if it's black boxes including these deep networks, deep neural networks.
• 45:44
  SPEAKER_01: No, you people are gonna have to do the extra work in AI to do model-based learning, to do federated learning, to do, even though it's gonna cost, take another five years and it's gonna cost five times as much, it's gonna have to be explainable.
• 46:00
  SPEAKER_01: And then the question is if it's explainable or if the model is now understood.
• 46:05
  SPEAKER_01: Is that an open source model like it would be if it was physics or math?
• 46:10
  SPEAKER_01: Or is somebody gonna copyright or patent that model?
• 46:14
  SPEAKER_01: And I I'm telling you that complexity will never work if it's secret.
• 46:19
  SPEAKER_01: Now, there's another aspect of this, which I can go into detail, but when you introduce complexity and the output is a calibrated output.
• 46:31
  SPEAKER_01: So for instance, you provide an image and the image accurately reflects the size of your pancreas or the size of a ping pong ball that's embedded in a phantom and it looks clear and it's the right size.
• 46:46
  SPEAKER_01: Or if you build a blood chemistry analyzer and the output is you have 100 milligrams per deciliter of glucose.
• 46:54
  SPEAKER_01: Whether now that calibrated output could be the result of a black box that's not open source, that's secret.
• 47:03
  SPEAKER_01: It may be patented or it may be secret, because nobody cares as long as the output is standardized, right?
• 47:14
  SPEAKER_01: Because the measure of quality is in the quality of the result.
• 47:18
  SPEAKER_01: Is the result accurate?
• 47:19
  SPEAKER_01: Is the result uh precise, etc.?
• 47:23
  SPEAKER_01: And so it doesn't, you can have a lot of proprietary complexity and investment in things that result in calibrated output, with and again scale that way because now you're scaling on things where everybody is being measured fairly on the same basis.
• 47:45
  SPEAKER_01: Um and that brings me back to the question Swami asked about you know the work on the Indian basis of the Indian system.
• 47:55
  SPEAKER_01: If the API is what's being standardized, then you don't care whether the people are building open source health records or closed source health records, as long as relative to the aggregation, the learning, the policy making, the research, everybody can benefit up to the limits of the privacy laws or the ownership, what you were calling ownership, which again I don't call ownership, right?
• 48:25
  SPEAKER_01: So you you can allow for private enterprise and private investment to create the complexity, but you also have to then regulate the either the networking or the calibration of the results, and then everything else becomes open source, open science, open medicine.
• 48:51
  SPEAKER_03: Right.
• 48:51
  SPEAKER_03: So, which is what my follow-up was going to be, which is it's one thing that science and engineering move in a particular pace, and it's another thing for regulation to actually catch up.
• 49:03
  SPEAKER_03: And I think one of the challenges that I'm actually seeing is if you look at the United States Congress, for example, I think more than 99, mostly 99% or 99.9% of them are traditionally trained lawyers.
• 49:16
  SPEAKER_01: So there's a word for what you're asking, which is called regulatory capture.
• 49:22
  SPEAKER_01: Yeah, uh the lawyers use the term regulatory capture, keep going.
• 49:28
  SPEAKER_01: There's a technical legal term.
• 49:30
  SPEAKER_03: That's the term regulatory capture, that's right.
• 49:32
  SPEAKER_03: It didn't strike me at that moment, but yeah.
• 49:34
  SPEAKER_03: So what ultimately happens is you have people who are really trying to push uh what's really possible on the edges of science, and you have people who are legislating uh at a pace where the people in between, which is the citizenry, are being deeply affected by these uh uh decisions.
• 49:54
  SPEAKER_03: So the question is what can be an accepted model where the people building the technology understand what are the basic ethical requirements from a data perspective that need to be within the design of the product itself, and that standardization is accepted by the regulators to know that yes, this bare minimum is being met by the technology community.
• 50:15
  SPEAKER_03: So let's progressively legislate, but uh let's progressively legislate with respect to that.
• 50:21
  SPEAKER_03: Because, for example, if you look at the 60s, you had Veneva Bush really running TARPA and really ensuring that the government was working hand in hand with the scientific department.
• 50:32
  SPEAKER_03: Are we losing, or is the United States not having that sort of a figure to aid the development of scientific process?
• 50:39
  SPEAKER_01: Yeah, yeah.
• 50:40
  SPEAKER_01: Um I'm gonna disappoint you.
• 50:44
  SPEAKER_01: Uh I don't have uh as clear a vision as I've had so far.
• 50:52
  SPEAKER_01: You know, I've been sort of answering your question saying, oh, yeah, this is obvious if you look at it this way, etc.
• 50:58
  SPEAKER_01: Um the solution to regulatory capture is not obvious to anybody.
• 51:07
  SPEAKER_01: Uh now I can give you uh an example uh of why this is such a hard question.
• 51:15
  SPEAKER_01: I don't pretend to have.
• 51:17
  SPEAKER_01: If you look at the way people are dealing with not what Vannevar Bush was doing in the you know in the 60s, but today in 2020, if you look at it in Europe, if you look at it in China, if you look at it in the United States, if you look at it in India, if you look at it in Africa, they all face the same question that you're asking very broadly.
• 51:50
  SPEAKER_01: You know, what how do we regulate technology?
• 51:55
  SPEAKER_01: And there is no it it ba you s you start to see it in cultural terms, not in terms of science or um even politics.
• 52:13
  SPEAKER_01: You you can't say that democracy is going to be better at doing this than uh you know than whatever people end up doing in Africa uh or in China.
• 52:27
  SPEAKER_01: You just can't.
• 52:29
  SPEAKER_01: Uh you know, the jury is out.
• 52:31
  SPEAKER_01: I can tell you, for instance, that in this country, in the US, the government, the federal government is, I won't use the the obscene term, but it's it it's constipated in the extreme, um over the fact that China is in a much better position to develop AI than the US, because China can enforce this flow of data to train the AI and to regulate the outcome.
• 53:06
  SPEAKER_01: But neither of them is accessible to a situation like the US government, right?
• 53:15
  SPEAKER_01: They they have limited ability to collect the data by design, you know, it's part of the maintaining freedom and whatever, and they have very limited way of making regulations on how the technology is distributed.
• 53:29
  SPEAKER_01: And again, that's partly because people want to rip people off to the tune of a trillion plus dollars a year, but it doesn't matter.
• 53:37
  SPEAKER_01: And Europe doesn't have that solution either.
• 53:40
  SPEAKER_01: You know, Europe is very upset with the fact that, you know, they're uh sending out all this money to Apple and to Facebook and to software companies that have almost nothing to do with European, you know, profit or what are they gonna do about it?
• 54:01
  SPEAKER_01: They don't know.
• 54:02
  SPEAKER_01: They really don't know.
• 54:06
  SPEAKER_01: Where's India gonna be?
• 54:08
  SPEAKER_01: I don't I'm not close enough to India to to place it on any particular cultural spectrum.
• 54:15
  SPEAKER_01: But it's gonna be different from Europe, and it's gonna, from what I can tell, it's not gonna be European, it's not gonna be American, it's not gonna be Chinese, it's not gonna be African.
• 54:25
  SPEAKER_02: So Adrian, uh my follow-up question uh is uh you've been on the W3C working groups, and uh uh one of the things that I certainly believe uh is uh I don't think uh the interoperability standards uh have been successful uh with the W3C group, right?
• 54:52
  SPEAKER_02: Uh I think there is a they've been working on it for years.
• 54:57
  SPEAKER_02: Uh I think the big tech firms sit there uh and therefore uh uh therefore these questions around centralization, aggregation uh have been uh uh you know have been questionable.
• 55:14
  SPEAKER_02: So uh so therefore what would you think are the alternatives if somebody is looking at uh you know breaking the deadlock?
• 55:24
  SPEAKER_02: And what do you think are the challenges and solutions?
• 55:29
  SPEAKER_01: Well you have uh you you have very high profile uh two different alternatives.
• 55:42
  SPEAKER_01: Uh one of them is very obvious, which is uh have W3C defer to IETF every place it possibly can, um rather than develop uh privacy, what they consider to be privacy solutions, um within just the governance of the so in other words, I'm not saying you shut down the people that make web browsers and the people that do the web protocols, but what I am saying is that you force them, like I've tried and failed, I'm almost certain I failed, uh, to embed uh as many IETF protocols, which are more open, less dominated in the regulatory capture sense, than the W3C by design, and even by the way W3C was created.
• 56:50
  SPEAKER_01: If you listen to the history of where it was, it is basically spun out of IETF for regulatory capture reasons.
• 56:58
  SPEAKER_01: Um you know, I mean, basically, uh I'm not a study student of tech history, but that's the way I oversimplify it.
• 57:07
  SPEAKER_01: Um so um you have that challenge.
• 57:12
  SPEAKER_01: And then separately from that, you have what people are calling Web3.
• 57:17
  SPEAKER_01: And whether you like the Web3 term, I don't care.
• 57:20
  SPEAKER_01: You know, I'm not I'm not in that debate, it's fine or not fine.
• 57:25
  SPEAKER_01: It's like ownership.
• 57:28
  SPEAKER_01: Ownership of data, it's whatever.
• 57:30
  SPEAKER_01: Um, but fundamentally, you look at what's being called Web3, and what do you see?
• 57:39
  SPEAKER_01: You see the fact that they are not dependent on any standards development organization, it is purely a market-driven, decentralized, finance-driven, and centralized finance driven.
• 57:54
  SPEAKER_01: It's what can we get investors and individuals to invest in?
• 58:00
  SPEAKER_01: And don't confuse me with regulations, don't confuse me with the cultural difference between El Salvador and Russia and the United States and India.
• 58:12
  SPEAKER_01: You see what I'm saying?
• 58:13
  SPEAKER_01: Uh, and don't confuse me with any standards.
• 58:16
  SPEAKER_01: Uh, we're gonna make everything open source, all of the governance mechanisms will be as decentralized or centralized as people choose to do that particular thing, and people will choose what becomes successful.
• 58:31
  SPEAKER_01: And so what I'm saying is you can go from W3C, which was your question, you can go in these two directions.
• 58:39
  SPEAKER_01: And W3C hasn't made up its mind.
• 58:41
  SPEAKER_01: Like I say, I I think right now uh they are very much stuck.
• 58:46
  SPEAKER_01: Uh and it's not good.
• 58:49
  SPEAKER_01: They they haven't decided to interweave the GANAP protocol, for example, into the work related to digital identity, so-called self-sovereign identity, which is not self-sovereign at all.
• 59:02
  SPEAKER_01: It's a total misdirection because all of the work was funded by the Department of Homeland Security, which is the ultimate, not all, but 80% of it, which is the ultimate sovereign.
• 59:14
  SPEAKER_01: And uh, you know, and they're running into problems because the governments using the ISO model, standards model, is basically ignoring what's going on in W3C and saying, here we have a government-to-government model for the way we use for telephony or you know, passports or or whatever, and we'll regulate identity standards in a way that governments do.
• 59:37
  SPEAKER_01: And I say, fine, you know, your governments, you're gonna do what the sovereign governments do.
• 59:43
  SPEAKER_01: Um, anyway, so yeah, W3C is stuck because they're they're trying to say, well, you can't use Web3 because we we are web.
• 59:54
  SPEAKER_01: And so it's in idiotic uh position.
• 1:00:01
  SPEAKER_01: And they're not really doing the the other thing with IETF.
• 1:00:07
  SPEAKER_02: So, which brings me to the question of uh, you know, what is self-sovereign technology?
• 1:00:13
  SPEAKER_02: So, because that's really what I think in your definition, what is self-sovereign technology?
• 1:00:20
  SPEAKER_01: Self-sovereign technology is technology that you do not have a privacy policy with, because there's no counterparty.
• 1:00:31
  SPEAKER_01: So let me give you a very clear example of self-sovereign technology.
• 1:00:35
  SPEAKER_01: Uh, artificial pancreases.
• 1:00:37
  SPEAKER_01: Now, what's an artificial pancreas?
• 1:00:39
  SPEAKER_01: You have two pieces of technology, one of them is a continuous glucose monitors, they work very well.
• 1:00:45
  SPEAKER_01: You know, you put it on, it lasts two weeks or sometimes more, and it continuously measures your glucose.
• 1:00:50
  SPEAKER_01: And you have an insulin pump once a day.
• 1:00:53
  SPEAKER_01: You put some insulin in it, and it has an API, it has an interface, and it doses the insulin safely.
• 1:01:02
  SPEAKER_01: Okay.
• 1:01:03
  SPEAKER_01: What links these two things together?
• 1:01:05
  SPEAKER_01: What links these two things together is a piece of basically AI that right now is kind of very stupid AI.
• 1:01:11
  SPEAKER_01: Why does it have to be AI?
• 1:01:13
  SPEAKER_01: Well, because if it's just an algorithm that comes from somewhere else, uh, like your doctor, but you only see your doctor once every three months as a diabetic, it makes no sense.
• 1:01:24
  SPEAKER_01: 99% of the decisions that a diabetic makes between the continuous glucose monitor and the insulin pump in their belly are made by the patient, not by the doctor.
• 1:01:35
  SPEAKER_01: What's the patient doing?
• 1:01:36
  SPEAKER_01: Well, the patient is taking input from what's in front of them, how much exercise it is, you know, what's their current blood glucose, are they anxious or not, and trying to figure out within their physiology how much of a dose to give the next minute, right?
• 1:01:57
  SPEAKER_01: Um that piece of um that piece of technology can either be proprietary, Medtronic will be happy to sell you an algorithm that's proprietary or controlled, even if it's open source, it's controlled by them.
• 1:02:17
  SPEAKER_01: Or you can go to OpenAPS, a group of many thousands of patients, that basically created the open source alternative to this.
• 1:02:27
  SPEAKER_01: Now, when you use OpenAPS to do this job, you are using self-sovereign technology.
• 1:02:37
  SPEAKER_01: You do not have a privacy policy with the Git repository or the website that's giving you the instructions on how to install this thing for your 10-year-old kid that has to play sports and go on class trips, like in the other 10.
• 1:02:56
  SPEAKER_01: You see what I'm saying?
• 1:02:57
  SPEAKER_01: Yeah, that is self-sovereign technology.
• 1:03:00
  SPEAKER_02: So so then the challenge uh uh you know that we have uh Adrian is that uh many of our digital lives uh are actually connected with non-self-sovereign platforms.
• 1:03:15
  SPEAKER_02: Okay, and uh you suddenly want self-sovereign identity, right?
• 1:03:19
  SPEAKER_02: So therefore uh into interoperability becomes an issue, right?
• 1:03:24
  SPEAKER_02: Because then the challenge is that you know, what's I'm actually uh you know in a uh non-self-sovereign platform, and I can't do interoperability of my data, and I'm therefore stuck for uh you know uh for many things uh in the middle of nowhere.
• 1:03:39
  SPEAKER_02: So uh so therefore uh you know there is a lack of trust in digital platforms and uh institutions.
• 1:03:47
  SPEAKER_02: And uh, how do you get this trust back?
• 1:03:50
  SPEAKER_02: Because uh that's really the challenge of uh you know uh uh you know the fact that you talk about data trusts and things like that.
• 1:03:58
  SPEAKER_02: So, therefore, how do you bring this back?
• 1:04:02
  SPEAKER_01: Well, some of it is necessarily regulatory, and we're seeing it even in the newspaper today.
• 1:04:09
  SPEAKER_01: Uh, we're seeing it with the Apple App Store.
• 1:04:12
  SPEAKER_01: There's a there's a big debate going on in the Netherlands uh where uh they're starting, they want to fine Apple for not allowing the dating sites, the social uh the dating sites to use their own payment.
• 1:04:30
  SPEAKER_01: Um, you know, they have to use Apple, Apple's payment thing and pay Apple, whatever.
• 1:04:36
  SPEAKER_01: And Apple went back and said, well, you can have a dating app, but um, and you can have your own payment system, but you have to pay us 27%, I think they said, because you're using our App Store.
• 1:04:50
  SPEAKER_01: So this question that you're asking about a platform, and Apple is a beautiful example of a platform because it's not only the largest company in the world, but it's also, in my opinion, very successful as a platform, uh, much more preferable to alternatives.
• 1:05:07
  SPEAKER_01: And so what people are doing is asking the question in the antitrust context, in the regulatory context, but it's not tech regulation, it's antitrust regulation, it's the same thing.
• 1:05:21
  SPEAKER_01: Um how what is the public interest in terms of Apple's App Store being different from Google's app store?
• 1:05:31
  SPEAKER_01: Right?
• 1:05:32
  SPEAKER_01: Google's app store would allow you a lot more freedom in how you interact with the patient and with the other apps or the capabilities of that wearable device.
• 1:05:46
  SPEAKER_01: Um it this this is being worked on right now.
• 1:05:51
  SPEAKER_01: Um now in my project in HIU of one, I'm actually uh if I had more volunteer resources.
• 1:06:00
  SPEAKER_01: Is we would do this faster.
• 1:06:01
  SPEAKER_01: We'll do it slowly.
• 1:06:03
  SPEAKER_01: We're trying to call Apple's bluff with respect to health data.
• 1:06:07
  SPEAKER_01: In other words, we're saying, can we introduce an app into your App Store that has access to the aggregated data in Apple Health that Apple has collected?
• 1:06:19
  SPEAKER_01: Very nice job.
• 1:06:20
  SPEAKER_01: Thank you very much.
• 1:06:21
  SPEAKER_01: The standards are more or less industry standards, but if not, they're Apple standards.
• 1:06:26
  SPEAKER_01: And it's Apple's problem to convert them to whatever the industry standards say.
• 1:06:31
  SPEAKER_01: Good.
• 1:06:31
  SPEAKER_01: You're doing a perfect job.
• 1:06:32
  SPEAKER_01: Love your privacy policies.
• 1:06:34
  SPEAKER_01: But now you are going to restrict what kind of app has access to the data that you've put on my phone.
• 1:06:43
  SPEAKER_01: And uh our goal, HIE of one, the trustee that goes in there, basically says we don't control the privacy policies of somebody that puts in the app.
• 1:06:57
  SPEAKER_01: We just give them the app software, and then the privacy policies belong to that particular group of patients.
• 1:07:04
  SPEAKER_01: Just like a blockchain.
• 1:07:06
  SPEAKER_01: You know, you make the Bitcoin blockchain and it works, uh, but you don't know whether people are using it for cybercrime or for decentralized finance of great projects or remittances to their uh you know family in Africa, right?
• 1:07:25
  SPEAKER_01: So um so it's going to be an interesting situation uh when my project actually introduces the app connector, the thing that basically puts the iPhone data into the web under the control of the patient.
• 1:07:56
  SPEAKER_01: Well, you know, what representations can that be?
• 1:08:03
  SPEAKER_01: So, in other words, to and go back to your question is can I install a self-sovereign app that's self-sovereign to me?
• 1:08:11
  SPEAKER_01: I compile it from source, just like that artificial pancreas.
• 1:08:14
  SPEAKER_01: I go to my website, you know, uh my GitHub, right?
• 1:08:20
  SPEAKER_01: Trustee community, you look it up if you want.
• 1:08:23
  SPEAKER_01: It's my my project's website, right?
• 1:08:25
  SPEAKER_01: I take this code, I compile it into an app, and I go to Apple and I say, I want to install this on my phone, the way I could on my Google phone.
• 1:08:38
  SPEAKER_01: Will they say yes?
• 1:08:42
  SPEAKER_01: Tell me.
• 1:08:43
  SPEAKER_01: That's self-sovereign technology in the platform context.
• 1:08:48
  SPEAKER_02: So uh so here are my last questions to you, uh uh Adrian.
• 1:08:55
  SPEAKER_02: Uh so my first, and these are like rapid fire questions, and uh uh so my first question to you is uh uh if if you're really advising an 18-year-old in college, uh you know uh the skills that they need to build for the next decade, uh what would you tell them they need to have to be successful in the future?
• 1:09:23
  SPEAKER_01: Um nobody's ever asked me that.
• 1:09:31
  SPEAKER_01: Um I what can I say?
• 1:09:40
  SPEAKER_01: Um I can tell you that I encourage both of my children to become doctors.
• 1:09:48
  SPEAKER_01: One of them is a very accomplished doctor, the other one didn't get far enough for various reasons uh to do that, though he um um so I guess what I'm saying is uh develop a profession, develop something that you are going to be expert at, um and hopefully that you also like.
• 1:10:18
  SPEAKER_01: Um, you know, it's not a yeah, basically uh figure out what a profession is.
• 1:10:28
  SPEAKER_01: You know, you could be a salesman.
• 1:10:30
  SPEAKER_01: I don't, you know, that's perfect, you know, don't have to be a doctor, but figure out what is the meaning of being a professional and what is a professional doing.
• 1:10:40
  SPEAKER_01: A professional is basically selling their time.
• 1:10:44
  SPEAKER_01: Now, if you happened to then get into a situation where you're selling more than your time, uh you're lucky, or maybe you're smart in a different way and you become a hedge fund manager or whatever, um, or an angel investor or something in between.
• 1:11:05
  SPEAKER_01: Uh but if you basically can find a way of developing individual expertise in a as a professional, then you could be an artist and you will be successful.
• 1:11:24
  SPEAKER_01: You don't have to do it, etc.
• 1:11:26
  SPEAKER_03: Anyway, that's my follow-up question is Adrian, what are some of the books that have influenced the most uh that have influenced you the most over the course of your life?
• 1:11:37
  SPEAKER_01: Catch 22 is probably the it it it's it I you know I I might have been a hippie if I hadn't spent so much time in school, but um it's basically question authority.
• 1:11:57
  SPEAKER_01: Um and it's captured very nicely in in Catch 22.
• 1:12:03
  SPEAKER_01: Um I have a hard time thinking of the next step, but yeah, basically, um and and part of it is that I was nine years old when I we emigrated, my parents and I from Romania to uh you know, first to Paris and then to the US.
• 1:12:20
  SPEAKER_01: And so I was brought up to question authority.
• 1:12:25
  SPEAKER_01: It was even though I wasn't old enough to be influenced by the system, I was influenced even more by my parents.
• 1:12:34
  SPEAKER_01: And uh so yeah, read any kind of uh stuff that teaches you to question the institutions around you.
• 1:12:47
  SPEAKER_02: Uh my next question, uh Adrian, is uh what does the word successful mean to you?
• 1:12:55
  SPEAKER_01: Oh, having an impact on as many people, uh an ethical or uh you know, a positive impact on as many people as possible.
• 1:13:07
  SPEAKER_01: That's for me, that's what's important, you know.
• 1:13:10
  SPEAKER_03: That's brilliant.
• 1:13:13
  SPEAKER_03: Adrian, my next question is if you could have a dream dinner, who would be the four guests that you would invite and why?
• 1:13:24
  SPEAKER_01: Um, I I can't answer that.
• 1:13:37
  SPEAKER_01: It's no, I I feel like I I don't have a ready answer, and I feel like no matter what I say, uh, you know, it would polarize uh my work.
• 1:13:52
  SPEAKER_01: Uh, you know, so for instance, you know, in my role both in the software developer and the patient privacy rights sense, I try very hard not to um be partisan politically or culturally for that matter.
• 1:14:09
  SPEAKER_01: Um and as soon as I start answering a question like that, people will read into it whatever prejudice they happen to have, and it doesn't do anybody any good.
• 1:14:26
  SPEAKER_02: Uh, what is one great piece of advice that you got, uh, which you've never forgotten till date?
• 1:14:36
  SPEAKER_01: No.
• 1:14:38
  SPEAKER_01: Um, where there is no conflict, there is no interest.
• 1:14:41
  SPEAKER_01: There was the comment of a mentor of mine, um uh very important mentor of mine, uh, with uh uh extremely senior in both business and medicine.
• 1:14:58
  SPEAKER_01: Um, and obviously for obvious reasons, I'm not going to say who it was or where it was along the way, but um he was incredibly generous.
• 1:15:09
  SPEAKER_01: Uh I've had the luck of having many mentors, and uh who were that that's how I got here.
• 1:15:16
  SPEAKER_01: But uh yeah, and uh he uh he was incredibly generous with his time and with his advice and with his help to my various things, and that was the saying that stuck with me.
• 1:15:29
  SPEAKER_01: Um the uh the corollary uh is from comes from some TV show uh or somewhere where somebody said uh uh that uh we knew we were gonna make money when the lawyers started coming out.
• 1:15:48
  SPEAKER_01: Which it is exactly the same thing when you think about it.
• 1:15:52
  SPEAKER_01: Is uh, you know, anyway, and certainly, you know, privacy in general, to the extent I'm a privacy professional, uh, it's all about managing conflict of interest.
• 1:16:04
  SPEAKER_01: You know, so much of what goes on under the guise of either regulatory capture or ethical behavior or avoiding dark patterns or user-centered design, all this bullshit, GDPR.
• 1:16:17
  SPEAKER_01: Um, it's all basically uh trying to engineer conflict of interest.
• 1:16:23
  SPEAKER_03: Got it.
• 1:16:24
  SPEAKER_03: So, Adrian, what is something that you believe in that nobody else agrees with you on?
• 1:16:29
  SPEAKER_01: Oh, that uh AI has to be open source.
• 1:16:35
  SPEAKER_01: Uh that, you know, uh both for pres self-preservation of the species in the sense of, again, Blade Runner, you know, Tyrell Corporation, but also uh just you know, humanity.
• 1:16:48
  SPEAKER_01: Uh you we can't compete with the machine, you know, uh creating uh competition between the US and China on the basis of you know how the AI manages the banking system or the credit system, uh you see what I'm saying?
• 1:17:03
  SPEAKER_01: It's just idiotic.
• 1:17:06
  SPEAKER_01: It it makes no sense.
• 1:17:08
  SPEAKER_01: And so um, yeah, I I can't find other than you know my Deborah Peel and the people that I work with in patient privacy rights, nobody is willing to go that far.
• 1:17:22
  SPEAKER_03: Yeah, so Aiden, my final question is if there is a guest whom you would like to see featured on a future episode of Contramines, who would it be and why?
• 1:17:31
  SPEAKER_01: Hmm.
• 1:17:33
  SPEAKER_01: Uh Tim Cook.
• 1:17:36
  SPEAKER_01: Yeah.
• 1:17:37
  SPEAKER_01: Serious.
• 1:17:38
  SPEAKER_01: Or as close as you can get to a proxy for somebody that is managing technology and platforms for the next generation.
• 1:17:51
  SPEAKER_01: Um obviously understands the politics much better than me, understands finance much better than me.
• 1:18:01
  SPEAKER_01: I'm an amateur of both.
• 1:18:04
  SPEAKER_01: But there are people out there.
• 1:18:07
  SPEAKER_01: Um anyway, yeah.
• 1:18:12
  SPEAKER_02: Thanks, Adrian.
• 1:18:13
  SPEAKER_02: Uh, it was a fantastic conversation.
• 1:18:16
  SPEAKER_02: Uh we thoroughly enjoyed the length and breadth of what we spoke, right from privacy to individual rights to impact of uh you know the health uh data, the empowerment of patients and patient rights.
• 1:18:35
  SPEAKER_02: Uh it was thoroughly enjoyable and thanks a lot for your time.
• 1:18:58
  SPEAKER_02: Thanks for listening to this episode.
• 1:19:00
  SPEAKER_02: For selected links and details or not, www.contramines.com slash blog.
• 1:19:08
  SPEAKER_02: Follow ContraMind on social media and let us know who you would like to see next on the podcast.
• 1:19:14
  SPEAKER_02: If you want to listen to ContraMind on Apple Podcast, look at your comments and give us a link.
• 1:19:20
  SPEAKER_02: We're keen to know what you're thinking.
• 1:19:22
  SPEAKER_02: ContraMind is also on YouTube.
• 1:19:25
  SPEAKER_02: If you want to listen to the podcast on YouTube, click the subscribe button and stay up to date on all our releases.
• 1:19:32
  SPEAKER_02: Thanks for listening and tell me.