He Said Facebook Knows Things About You That You Cannot Even Say About Yourself

This conversation was filmed two years ago in San Francisco. I am sharing it today because everything Raffi Krikorian told me then is more relevant now than it was when we recorded it.

Raffi Krikorian was employee number 50 at Twitter. He became the company's Chief Technology Officer. He then left to become CTO of Uber — overseeing self-driving car development at one of the most ambitious technology projects in history. Then he left that to become Chief Technology Officer of the Democratic National Committee — integrating AI into American politics at the highest level for the first time. Today he is Managing Director and CTO of Emerson Collective and has spent over twelve years as a board member of TUMO in Armenia.
He helped build the internet we have today.
And in this conversation — filmed two years ago, republished this week as X is back in every headline and AI deep fakes are reshaping elections around the world — he said it is broken.
He also said Armenia could be the AI powerhouse of the world. He said that before ChatGPT existed. He still believes it.
In this episode of INSIGHT, Emma Sargsyan sits down with Raffi Krikorian for one of the most wide-ranging and honest conversations about technology, power, privacy, children, politics and the future of the internet that this show has ever produced.
─────────────────────────────────────
A NOTE ON TIMING
This conversation was recorded two years ago when Raffi was Managing Director at Emerson Collective. Since then the world he described — the broken internet, the AI deep fakes in politics, the data privacy crisis, the question of children and screens — has not improved. If anything it has accelerated. I am sharing this episode now because the questions he raised then are the questions everyone is asking today. And his answers hold up completely.
─────────────────────────────────────
ABOUT RAFFI KRIKORIAN
Raffi Krikorian is Managing Director and Chief Technology Officer of Emerson Collective. He previously served as CTO of the Democratic National Committee, CTO of Uber's Advanced Technologies Group, and Vice President of Engineering and CTO at Twitter where he was one of the company's first fifty employees. He has been a board member of TUMO Center for Creative Technologies in Armenia for over twelve years.
─────────────────────────────────────
ABOUT INSIGHT
INSIGHT is the first Armenian-English language power and culture podcast in the United States, hosted by Emma Sargsyan. Each episode is a long-form conversation with an extraordinary guest at the intersection of business, identity, technology, leadership and culture. 
Subscribe for new episodes every week.

Transcript

SPEAKER_00 0:05

Thank you very much for Green for the interview. I want to make the world a better place. Say it's on your LinkedIn. What is your recipe to making the world a better place?

SPEAKER_02 0:34

Well, I mean, there's so many different things I think are wrong in the world and that can use people thinking about them. And so, like, for me, what I want to do is I want to just get more people thinking about the problems that actually matter. Like whether it be problems in education or healthcare or the environment, like these are the things that like are really important for the entire world, right? Like we need to figure out how our next generation kids are going to learn better, be better equipped for as the world is changing. Like health is incredibly important. We're all living longer, we're all having different diseases and things like that. So we need more people thinking about how to make our lives better in a lot of different ways. And the and the climate is changing. So we should be focusing on that. So for me, making the world better is like really trying to get as many people as I can to really focus deeply on innovations or whatever it takes to move those kinds of problems, like these big global systemic problems.

SPEAKER_00 1:29

So we are not like we are thinking about more shallow problems. We are worrying about our everyday lives and stuff and do not pay too much attention, enough attention to the global problems. Do you think so?

SPEAKER_02 1:41

Well, no. I mean, like, I think like everyone just operates at different levels, right? So like you can be doing climate work, for example, on your local level, right? That we can talk about recycling, we can talk about food waste, we can talk about all these things for regular every single day people. Or we could be thinking about big systemic ways to think about like we need like we need to rapidly scale up how much solar power is being produced in the world. We need to rapidly change how wind turbines are being deployed so we can get wind energy into the into the global mixture of energy. So I think that like everyone can be thinking about these types of problems. It's not that we're thinking about the wrong things, we're talking about thinking about at different levels. Like, I think people just need to be reminded every once in a while that there are these bigger things that we all are members of one world that we can all be working on together. We just need to be thinking about them.

SPEAKER_00 2:31

There's no planet B.

SPEAKER_02 2:32

There's no there's no backup plan. Like this is this is all we got. I mean, like some people are working on a backup plan, but right now we have no backup plan. So this is what we gotta do.

SPEAKER_00 2:40

That's the truth. I mean, uh, when we are speaking about these kind of big issues, uh psychologically, people would say, okay, I'm just a drop in the ocean. If I do something, I mean, is it going even to make difference? How to make sure that people actually feel that even the tiniest thing is going to make a difference? What can we do about that?

SPEAKER_02 2:59

I mean, these are these are really great questions. And I think like figuring out ways to have connectivity with your community, community at different scales, I think is like the most important thing. So if you think about it from the educational standpoint, like we all care about kids. If you have a kid, you care about your son, your daughter. Education matters a lot in that context. So like spending time there makes a lot of sense. If you spend time with your kids, maybe then you worry about the classroom that they're in. If you worry about the classroom, you'll worry about the school that you're in. If you think about the school, you worry about all the cor all their pathways to your educational system. I think it's just a matter of, I don't think people are not able to think about these types of global issues. I don't think it's people don't want to think about these issues. I think it's what you just said, like we just need to give them pathways into it. So it's easy to feel hopeless about what's called the environment or climate right now. The world is changing. What am I supposed to do? But like I said before, like even just thinking about how much waste, how much food waste happens in your house. Or like, are you using recycled goods? Are you buying clothes, fashionable clothes and just throwing them away next season versus the recycling them or figuring out other ways to do it? I think these small interventions matter and add up. And so I think the question for people like us is more than how do you just inspire people that you can think about these issues and not have a worse-off life for it. You can contribute to these issues and in small ways just make the world better for yourself, which actually can make it better for everyone else at the same time. We just need to give them concrete examples in order to think those through.

SPEAKER_00 4:30

Like save the energy, don't waste food.

SPEAKER_02 4:33

Yeah, don't waste food. Turn the lights off when you're done. Look at your electricity bill. Are you using renewable power versus not? Are you buying like the next time you buy a car, maybe buy an electric car, not a gasoline car. I think all these choices, yeah, for every single person, they're like, does it really matter?

SPEAKER_00 4:49

What's in it for me?

SPEAKER_02 4:50

On the scale of billions of people, it matters. So I think that we can, and I think that like this matters for entrepreneurs and innovators and et cetera. I think, I think a lot of people, especially when they think about interventions at this type of scale, always talk about it as trade-offs of just like, well, you're not gonna do like stop eating meat and instead be vegetarian. And I think like a lot of people will be like, but I really like meat. And what we need to do instead is tell like better have to innovation in what vegetarian meals could look like. We just need to make the world like you can have these things and get to a better place. Like an example I used to have is like, you know, I have I wear an Apple Watch, I buy a new watch every year. I don't need to buy a watch every year. Like, in fact, like before Apple watches came out, I used to literally wear the watch my father wore when he came to this country for the first time. Like, electronics have become disposable in a lot of ways. How do we inspire a movement to have heirloom electronics? Like, maybe it is the same watch that my father was wearing. Maybe it's the same phone that my my sons, you know. How do we make, how do we build these like circular cycles in our economy? I contend a lot of this is just storytelling. We just need to tell good stories that show people like our lives could actually be better if we just think about it in a different way.

SPEAKER_00 6:05

What's in it for me, what I'm getting of not, I mean, if I can uh waste some food, it doesn't change anything in my life. And I think it's not going to happen. What role do you think influencers, people of great influence, of a great like who have a great follower account, uh, do they have a say? Like we all know, like there are people who with one story can make or break something. Yes. Uh are they uh in the game? Are they yeah?

SPEAKER_02 6:31

I mean, so like I think the influencer strategy and the way that they tell stories obviously works. Like I think we see this in the US elections right now. Like the number of people who are doing influencing work on trying to get voter registration. Look, Taylor Swift just told people to go register and vote. She's like one of the ultimate influencers. 300,000 people within hours registered after her Instagram post within hours registered to vote. So they clearly have a role. And there are influencers starting to play in these spaces. There are climate influencers, there are education influencers, those are usually teachers and stuff, but they who do really well on TikTok and Instagram. And I think just equipping them with these stories and figuring out ways to amplify them. I think part of it is also algorithmically we're getting we're shown the wrong things or we're shown things that only you care about. And we need a better way, like information diet is one of the things we also need to be thinking about. Like, we need a better way to show you things that you should know, not necessarily only the things that you're like rabbit holing on.

SPEAKER_00 7:29

Because um in Armenia, we usually complain, oh, this TikTok has become such a mess. And what when people say this, I tell them, look at your feed, what you see. What you see is what you watched for at least a couple of seconds.

SPEAKER_02 7:44

If you change it, and it just rabbit holes you down that load.

SPEAKER_00 7:49

Yeah. And if you change it, TikTok has an option. You can just make it like forget everything about me. Let's start from the scratch. And if you do that, actually you can find some educational comfort, some amazing things, not to complain.

unknown 8:01

Yeah.

SPEAKER_02 8:02

I mean, like I I actually started a company a few like two years ago where we were trying to, it was called Speakeasy, where we were trying to build a new social network to show people different types of content, have better conversations online. And one of the things was exactly that. Like, from the algorithmic stance, we actually didn't want to just rabbit hole you down the things you wanted to see. We actually wanted to show you conflicting opinions, different ideas, because the whole point was to get enriched by the internet, not be rabbit-holed by the internet. So, like, we've like, I was experimenting with this for a few years trying to figure out how we can break out of these molds.

SPEAKER_00 8:36

But we as people don't like it when the ideas contradict to what we believe, it will be very hard to convince us.

SPEAKER_02 8:43

Yeah. And so, like my theory there is that like we need to think about not showing you something that's 180 degrees different than what you believe, but maybe only five degrees different. We just need to start nudging people along the path. And I think we can tell those type of stories all the time of just finding people who are just like, you know, one degree different than you. Because I think that's really important. But like we still have some common ground by which to have a conversation, but that one degree different gets you to move slightly and gets you seeing a different, a different standpoint that could still be friendly, but it's a different viewpoint. Yeah. And then we just use that.

SPEAKER_00 9:17

And gradually, gradually you're just showing that there is a world outside your world. Let's speak about education because recently we have very contradicting opinions. You don't need a diploma, sure. And you need a diploma. I belong to the old school. You do need a diploma. You do need to go to the university.

SPEAKER_01 9:33

Sure.

SPEAKER_00 9:33

And then the argument is that, oh, Steve Jobs didn't go to the university. Mark Zuckerberg is a drop out. Uh and uh we do not realize that those people they dropped out of Harvard in the first place, which is very hard to get admitted to what's your opinion about education? Is it is it okay not to have a diploma or yeah, I mean let's let's remove the diploma for a second.

SPEAKER_02 9:54

Let's talk about why education is important. Education is important because it teaches you foundational skills, it gives you other viewpoints in the world, gives you perspective. And so, like, I think those are very important. Like, we still need, like, you know, I I'm of the school of thought that we need to teach kids computational thinking, not how to code, not how computers work, but how to like dissect a problem, how to think about problems as component parts, how to put them back together again, how to like manage a team of people working on problems, how to like like these like raw skills, you need to be taught them. Like very few people are born with those skills. So, in that sense, yes, we absolutely need schools. Like, we need schools to teach you these foundational things that you'll then like build on top of, like stand on the shoulders of the giants before you so you can build the next thing. I think for the people who are talking about we don't need like so then we could talk about diplomas. Are diplomas needed? Diplomas might be a more old school way of looking at what your education is. And instead, like maybe we need to figure out how to badge you correctly, like maybe it's a collection of badges of like different skill sets, and these badges make up what a new diploma could look like. But I think fundamentally we need to teach kids these skills that they can then build upon and move to the next level. So I think you're right. The people who are talking about like you don't need a diploma usually have already been really highly educated and are really try have already figured out how to like leverage that to their purpose. There are a few kids who will be excellent who don't need it, but I don't think that's the solution for everyone. And going back to my previous thing, we should be working to uplift everybody. Like if we are in a society where we're all highly educated, imagine the things we could go do instead of taking a bet on a few people who didn't need to have the education, were just genetically born to be amazing or like pre had a predilection to be able to do math or computers in a particular way. Like we should be working for all of us because then like we can then have like really great conversations. We can build upon each other. We can do all the safe. Yeah, we can rise together.

SPEAKER_00 11:57

And here we see Estelin paradox. When I'm better off compared to the rest of the society, I'm happy. When I'm when all the society is better off together together with me, I'm not that happy.

SPEAKER_02 12:08

Well, I mean, I don't know if that's true. Like, I think that like I think we can all have a conversation about what's the world that we want to live in. And I think that like we probably all, going back to your previous thing, we all probably can't all individually make the world we want to live in. But if we can all be in this in a community with all of us are building, all of us are designing, then the world can emerge. Like one person might not be able to make that entire society happen, but an entire group of people working together. So like I think in that case, like rising us all up actually allows us to live in that place that we want to all live in.

SPEAKER_00 12:43

Yeah, that's so true, but we have to come to that understanding that we need to move together. Like the the bigger the community is, the better we are going to.

SPEAKER_02 12:51

Yeah, like I mean, like there is it is not zero sum. It is not like you get educated at the trade-off of someone else. It's not that you get rich at the trade-off of someone else. We all can do it. And if we all do, we actually all might live in a place that we want to be in.

SPEAKER_00 13:06

Um, I want to talk about your podcasts. When I was writing, like trying to understand the topics. I thought that one of your ways to making the world a better place are those podcasts because you touch upon of many issues, and most of them are tech related. And personally, I found things that are very close to me, like the data, the online, the security, etc. etc. Let's speak a little bit about this. Um, how to stay safe online? Because the internet is amazing, it has amazing opportunities, uh endless. But when you look from the other side, it is scary.

SPEAKER_01 13:44

Yeah.

SPEAKER_00 13:45

I have my bank details everywhere in my phone, and I'm re I really don't know how it's used. And maybe it is used in a way that I'm not happy for it to be used, and I don't know about it. So, in general, what should we do to stay safe online because we do lack the literacy and understanding? And what to do not to become a victim of these kind of malicious stuff.

SPEAKER_02 14:09

Yeah, I mean, I think I mean I hate to say it this way. I mean, there's like a few ways to look at it. Like, there's what can a regular person do right now or an everyday person do right now when they go online. And I think it's a it's a few simple things. Like, I think there are questions around like setting up multi-factor authentication when you log into things. Things that you do in your bank, you should also probably do in your social network, you should probably do on your email, things like that. You should always keep your software up to date. That's actually quite important because there are always vulnerabilities. But then the last one is just being, let's call it, distrustworthy of anything that happens online. So, you know, we do this right now in our family. Like, if my wife ever gets a phone call from me being like, I'm in an emergency, I need you to transfer money, or gets an email from me, the first thing she's supposed to do is like ask what our code word is. Like, I have a word that only she and I know, and just try to effectively authenticate me and understand, make sure it's made. Similar to what we used to do with phishing attacks, where when someone sent you a phishing email, before you click on it, or if it's if it's asking you to do something, maybe call that person and just ask them, did you just send me this thing? So I think this level of suspicion is actually kind of required these days when you're online, because like we've lost all the social cues of being face-to-face or looking each other in the eye. So we need to like figure out how to get this back in. So I think there are those types of things you can do right now to keep yourself safe. The problem is that that's kind of a depressing way to live.

SPEAKER_00 15:35

Like over being suspicious of everyone is understanding.

SPEAKER_02 15:40

And the internet was supposed to be this amazing place that we are all supposed to be able to like learn from and engage with other people and communicate in different ways. So I'm mindful that being suspicious is counter to that. And I think the call to action here is for people like me, we need to reinvent the internet. Like the internet right now is not set up to be trustworthy in a lot of ways. It's not set up to protect you as a user, it's set up to protect the people who's creating the websites and people who are aggregating all the data. And so I think like there's there's a call for like we need to reinvent what that internet looks like. And so people need to start working on that. And there are organizations like you know, Frank McCord is doing something with Project Liberty, the Mozilla Foundation is trying to reinvent what the web browser looks like. I think we need to look to organizations like that to really lead the way to just reenvision what this market looks like, what this privacy looks like, and how interfacing with humans should look like.

SPEAKER_00 16:35

Is it possible? I mean, is it possible to just remove everything, start from the scratch, reinvent?

SPEAKER_02 16:43

It is not possible. I I think rebooting is not possible. I think we need to just go on a slow evolutionary journey. The problem is right now, the values of the internet. So the internet is basically like a three-legged stool. So you have commercial entities who control a lot of it. You have government regulators who try to impose every once in a while, whether it be section 230 or whatever it is. And then you have everyday people like you and me, civil society in it. And we've allowed it to get really imbalanced. We've allowed it to be d dominated almost purely by commercial entities. Not saying commercial entities are bad. I'm I used to work in the corporate in the social commercial sector all the time. It's just there's a particular set of incentives of just like if your only bottom line is making money, then that causes you to make a certain set of decisions. That's why you need government and civil society to counterbalance that a bit. And so I think it's not possible to reboot, but what we need to do is to start strengthening those other pillars. Like I need people like you to go on your TV show and be like the internet's broken and remind people there could be a better way and just start getting them to think about what that better way looks like. We need regulators to get better educated so that they can do smart regulation to try to nudge these systems back. I think it took 20 years to get to a place we are today. It might take us 20 years to unwind it. But I think we just need to start that journey. And sadly, it won't be flipping a switch. It's gonna be a long, slow road of us working towards it.

SPEAKER_00 18:07

And all of us have to understand that we all have to have the same values.

SPEAKER_02 18:10

We all have to work together on it. We all need to have a shared understanding of like what we want to be like real leadership is needed here. And right now there isn't. There isn't that. And that's the thing we need to go for.

SPEAKER_00 18:21

But there is a wide consensus that it has to be fixed.

SPEAKER_02 18:24

Aaron Powell I mean, I think if you ask any parent right now of a kid who is online, they will say right now it needs to be fixed. And so I think that yes, I think there is. I think a bunch of people don't think about these problems every day. And that's another reason I do the podcast is because, like, you know, you understand why there are speed limits on the highway, but do you understand all the implications of a video doorbell on your door? That you like that do you understand that the police can subpoena that? Do you understand how all that could work? Probably not. Not saying that's bad. Maybe it's the right thing for police to subpoena it, but the fact that you don't think about it when you make tech buying decisions is part of the problem. And so we need people like you to understand what all the trade-offs you are making, it's more than just buying a device. You're buying into a particular way that the world works. And you could not. You could change it and you could choose something else. And those could sound real signals.

SPEAKER_00 19:18

We choose the easy way, like it's handy to have this, to have that, but we really don't understand what are the long-term implications.

SPEAKER_02 19:26

And that's why I do the podcast is to tell those stories so hopefully people can start to understand.

SPEAKER_00 19:30

And all your podcasts start with a case.

SPEAKER_02 19:33

Yeah.

SPEAKER_00 19:33

Yeah. This happened, and this is what And let's talk about it. Yeah, let's talk about it. I read and I agree to the terms and conditions. That is the biggest lie of the century. Nobody ever does it, right? No one does it. What what can go wrong with these terms and conditions? There's small tiny letters written.

SPEAKER_02 19:48

Well, anything. I mean, like, I think that one of the things that I'm really that I need to go do as part of my job, or when people when people ask me about these termic conditions, is can you ever delete your account? Like, and so like that's an honest question. Like, if I if I stop using a service, can you tell that service they need to delete all my information? So they can't use it. They can't use it in modeling, they can't use it in analytics, they can't use it in whatever. In some situations, the fine print says you can't delete it. And so, like, that seems wrong, right? That's my information, it's my data. I should be able to control it. So I think that's just like one example of what these terms and conditions could be hiding. I mean, there are plenty of others, especially with AI and generative AI. Like, these companies, these organizations want as much information as possible in order to train their models. Do you have a say of where your data goes inside those models? Do you potentially have any form of compensation if they make a lot of money off of your data? Like, all these things are buried in the terms and conditions. And I mean, yes, it is pages and pages of legal text. And so one thing that we could be doing on the smart regulation side is we could mandate that terms and conditions could be one page. We can mandate that terms and conditions could just be a simple bulleted list. We don't have those mandates right now, so it just gets long. We could just simplify this.

SPEAKER_00 21:08

And in this bullet points, they they can end something like similar to those and include this similar in a 10-pager.

SPEAKER_02 21:18

Well, I mean, like, but I think like I think the the thing I'm trying to say is that like there are levers that we can use to push the system around. We're just not doing it right now, so it comes at the expense of people like you and me. And instead of being for people like you and me.

SPEAKER_00 21:32

Yeah. What role does the government play in all this? We recently had Mark Zuckerberg have a big statement about what happened with Facebook. So what should the governments do in the States, in at every corner of the world, in Armenia, so that its citizens are somehow protected, feel somehow protected with all these malicious things that can happen.

SPEAKER_02 21:56

What's the I think the role of government here is smart regulation and smart. Legislation. I don't think it's overhanded, like tell you you can't do things. I think the government needs to do two things. One is it needs to remember who its constituents are, and its constituents are people. So we need to ensure that people's privacy is protected, people's livelihoods are protected, all that kind of stuff. And I think you can do that by mandating things like, again, I should be able to delete my account, I should be able to know where all my data went. I should be able to have to say whether or not you can use my information. I think a few of those type of things, as well as some transparency. Like when an algorithm presents me with content, I should be able, it should be transparent as to why it chose that. So I can be like, no, that's not the right thing. Like I don't want to see that. It should be able to see what the choices are being made for me. Government also needs to spur innovation. And so like I think a lot of people think that government's only role here is to like say, don't do things. But I think it needs to do that and it needs to say, but develop new things that are privacy preserving, that are open source, or all these things that could actually be preferred to benefit of the end user. So I think what government's role could be these two things to actually try to really fix what's what's going on.

SPEAKER_00 23:10

In Armenia, right now we are having this law that all the businesses are required to have CCTVs, cameras, and this camera footage is uh they are obliged to hand over to the Ministry of Internal Affairs and Police. It's a major thing. Like a lot of us do understand that that's a privacy issue uh because I mean maybe I don't want someone to see me walking in the street with uh whoever. And other parts like it's a safety regulation, it's it's for our safety. Are these kind of things well, they are obviously viol violation of a privacy, but what is the justification for this? Like on on to be on the government side, okay, they are doing this.

SPEAKER_02 23:55

Well, the government size is ease of use, right? Like they want easy access to all the information. And like I think like it's a natural tendency for most governmental agencies to be pro-surveillance, right? It's just simpler for them to do policing work in that case. Like this happened even in the front for the French. The French have had the French have a massive privacy loss, and it's part of the EU, all that stuff. Um, but they inserted the clause that for the Olympics that they could do massive surveillance. And so like that already triggered all the privacy problem uh advocates. I think in these situations, like there needs to be appropriate checks and balances. I'm not saying that police shouldn't have access to it, but there needs to be a very concrete, potentially slow process by which to get access to it so that lots of people can weigh in. Like in the US, if you wanted to get CCTV access, you need to file for, you need to get a judge to step in, they need a subpoena to get that information, they need to do all these things. And I think like that doesn't solve all our surveillance issues, but at least adds a little bit of friction, which then causes police to have to do something else. And like that something else might be the better answer in a lot of situations. But I think like those small pieces of friction are actually good. Like technologists usually think that you want all the friction out of the system so things can move really quickly. That's a slippery slope to getting to disinformation, to get to algorithmic bias and things like that. Friction actually causes you to have to think and take a breath. And that might be the way to go approach something like this.

SPEAKER_00 25:31

Gradually, slowly and steadily. Uh, we spoke about misinformation. You mentioned Facebook is the main tool of spreading misinformation. You can have as many fake accounts as you want.

SPEAKER_03 25:42

I believe so too.

SPEAKER_00 25:44

You can do whatever you want. And there's potentially no way to tackle you back and see who is this user one, two, three, four, five. What should we do? Not to I mean, I do understand how to fact check if something is popping up on my feed. I know, okay, this is fake, I'm not clicking on it. This is not fake, I can shouldn't do it. Uh what an average citizen with no high level of media literacy has to know so that not to become this the the the victim of these clickbaits.

SPEAKER_02 26:18

Yeah, I mean I think this gets back to my point of like you need to be we need to teach people to be suspicious. Like, and I think that's sadly the only thing that we can do right now in the current system, or at least that's what we can teach people to do. But it's not the only thing we can do. What we should also be doing is trying to push for again smart regulation to require transparency. Like we should be able to understand why the algorithm or the set of algorithms chose a particular piece of content to put in front of me. Like, what was the decision factor? And like I feel like if we if companies were started were required to start disclosing that, it would change behavior because they would either have to be forced to reckon for themselves that like we have built an algorithm that everyone can see that's choosing to cause you to have violent tendencies or choosing to cause you to get depressed, or truly choosing to cause you, a teenage girl, to commit suicide. Like it would force companies to reckon with that. And I contend that would change a lot of behavior right now. So I think if we're looking for the two things to do, one, everyone needs to be more suspicious. That's part of being digitally literate these days. You need to have a better understanding of why you're being shown things. But two, we need to put pressure from the outside. Transparency is probably what seems to be this the lowest possible bar, but I think could actually have outsize effects.

SPEAKER_00 27:35

Does Facebook know a little too much about all of us?

SPEAKER_02 27:38

Oh, 100%. Like I think that like, I think the the beauty of these deep learning algorithms, these machine learning algorithms, is that they can make decisions that no human could ever ever made. They can spot patterns that no human could ever spot because they just can amass more data than our brains can process. So on that thinking alone, then yeah, Facebook has probably spotted patterns in the way you digest information, the way I digest information that I could never even articulate. And that's the beauty of all the math that's going on beyond it. But that's also the power and the problem. And so, like, that's the thing that again, people like you should be more, people like people who are watching, people like me should be more aware of that as we make decisions that these systems are watching us. These systems are watching us, learning from us, adapting to us, and they're doing their best to keep us engaged, to keep us clicking, to keep us doing all these things. And it's part of our jobs as human citizens, citizens of the world, to then not be sucked into it. And sort of like look for other points of view, connect with humans outside the computer, talk to regular people, and just like change like this can't be the only view you have in the world. It might be one input, but that's our job in terms of behave differently in face of this.

SPEAKER_00 28:57

We sort of, especially after COVID, we became very close. We don't want to speak to other people. I mean, I find it hard to call people. Sure. And many people of my age and younger, they do not do this. Um and for children, all of children are somehow, some way one a little bit more, other a little bit less. We the they have access to internet. And the safety of children in internet is a major issue. I mean, you cannot monitor all the time, you can do some parental controls, but it's not still going to do all the work. And children are easy targets for any any hacker, any bad guy in the internet. So what should we do as parents for the children not to, I mean, to be safe online?

SPEAKER_02 29:46

Yeah, I mean, like my kids, for example, I mean, I I can't tell every parent what to do to their kids. My kids are not allowed to use social networking on the internet. Like they are eight and eleven. There's I think they're way too young for it. So they are not allowed to use any piece of technology that involves that could potentially be speaking to another person on the internet, all the way down to video games. Like the video games that they are allowed to play, if it has any internet component, it has to be that you can't, no one can just talk to you, like either text or voice or whatever. Like all that is banned from my children's lives. And I think that is probably the smart thing to do for almost every parent. It's just like, I'm not saying we can have a whole other conversation of whether kids should be using screens and computers, whatever.

SPEAKER_00 30:29

That's a different thing.

SPEAKER_02 30:30

When assume that they're using it, you could just say that like the kinds of stuff they're gonna do in the internet um does not allow for communication with other people. Like it's just not the thing. You can be playing games, whatever. And I think that's probably step number one is sort of like allow them to understand the technology, like you and I understand the technology before they have to then grapple with all the social constructs and things like that. So postpone that as long as you can. One of the things I've always wanted to do is, you know, in the United States, we teach a class called home economics in in schools, where home economics is meant to be it's literally when I was in school 30 years ago or however old I am, it would teach you how to sew, it would teach you how to do wash the dishes, do all the stuff around the house. But we need to revamp that. Like it's home economics now, it should be digital literacy, it should be how to use computer applications, it should be how to converse online, it should be this level of suspicion, teaching kids this level of suspicion I can. So I think like from a parent standpoint, the the number one thing you can do is delay your kids as long as possible before you allow them to get into social networking. But from the school standpoint, we need to then start teaching computer literacy, internet literacy, artificial intelligence literacy. From the government perspective, we need to start putting in a bunch of different guardrails about like child safety online regulation. In the US, we have COSA that's stuck in our Congress right now. But it's specifically meant to be that kids can't see targeted ads, all these changes that can actually be put on the platforms. So it's not falling only on the parents to push off all these things in the computer. So I think we can attack this from lots of angles. But for parents, getting back to your first question, delay as long as you can. Like my kids play on the computer all day long, just they don't talk to their people on it. They have to talk to me, they have to talk to their friends at school. That's the way they communicate.

SPEAKER_00 32:21

Okay. So no, no, I mean, no WhatsApp, no chat. No, they have no WhatsApp, no chat. Oh my god. I mean, imagine like saying something like that to an Armenian. I mean, in Armenian schools, they have this WhatsApp chat.

SPEAKER_02 32:32

Yeah, come on. I mean, like a lot, it's hard. I'm not saying this is easy. Like it's taken us to have effectively deals with other parents of kids in their school, being like, we're all not gonna do it. We're all gonna delay it as long as possible. Like, so for example, our school, we've all have an agreement effectively that no kid is gonna have a cell phone until they're 13 or 14.

SPEAKER_00 32:54

Okay.

SPEAKER_02 32:54

Not all parents are following it, but enough parents are following it that like text messaging each other is not a thing. It's not it's not a reliable way to communicate because kids know that not everyone has it. So I think that these are things communities could be doing together in order to try to do it. You know, when this when the city of San Francisco, they were trying to figure out how to do earthquake preparedness. And the number one thing they realized during earthquake to prepare citizens for earthquake preparedness was just know your neighbor. Like literally have a conversation with your neighbor, face to face, know that person because like then you have a social community, you have a construct that like something happens during earthquake, you have someone you can turn to. But if you can do that, then you can also have block parties, you can have do all these things that actually could be wonderful.

SPEAKER_00 33:43

Social life, right?

SPEAKER_02 33:44

Social life. Like we need to somehow teach people the right balance again of just like everything is gone to socializing over the computer. You need to bring it back. And look, I helped build one of these systems, right? Like I was like, I was an executive at Twitter. I was one of the first, like I was the employee number 50, give or take. So like I helped build these systems, and I'm I'm saying that now we need to find balance again.

SPEAKER_00 34:08

Yeah. Um, let's speak about AI. Um, my grandmother used to say that whatever is built by a man can never be more clever and more developed than a human.

SPEAKER_01 34:22

Okay.

SPEAKER_00 34:22

But now we kind of see something different. Um and sometimes it's scary, sometimes it's not. But I want to focus um and you as uh pioneer who did the AI in integrated AI into politics. I want to speak about this. Um, what's the future of AI in politics, especially in the verge of elections right now in the United States and in the world? And um of course AI is going to have different impacts in different parts of the world. What's AI doing in the US is not the same as what's going on in Japan or in Armenia. I just want to hear your insight of what is the future of AI in politics, in climate change, and in medicine.

SPEAKER_02 35:02

I mean, so those are big topics. Let's do start with politics for a second. I mean, a lot of this we can choose. Like again, like I'm I'm I sound like a broken record, but we can choose what's the direction we want to go. I think it's really easy for us in politics to think that the only way this is gonna go is we're gonna have problems with disinformation and deep fakes. I mean, if you look, this already happened, right? Like in the US cycle, we had a deep fake of President Biden making a robocall in New Hampshire trying to tell people not to vote. In Indonesia, it was like Disneyland of deep fakes, as far as I could tell. Like their entire election was mired with deep fakes. India had this problem as well, where like literally celebrities were giving endorsements that they had never given because a computer just generated for them. So it's easy to think that this is the way it's all gonna go. Um, but as I choose to be more optimistic about it, we can choose that this is not the way we want it to go. We can just say we can start putting simple guardrails in place. Like, for example, there was a a woman running for a parliament in Ireland where she was then assaulted effectively online with synthesis, synthetic sexualized imagery of her. Yeah, let's they tried to drive her out of the race because they were trying to, they were, they just flooded Ireland with this content. She didn't leave the race. Um, but we could just say that should be illegal. Like we could just say this is not the world we want to live in, that should be illegal. So we can start bending the curve away from this hellscape of disinformation and things like that. It's not 100%. People look for solutions that are 100%. I'm saying even 5% better, 10% better. We should just start. We should start doing the simple things. Now, what could we do instead if we used AI properly in politics? There's a few things. One is every single voter could actually get a customized outreach on the things they care about. Like a candidate has no ability to talk to 100,000 constituents. There just isn't enough time. But maybe every single one of those constituents deserves to hear what that candidate's thinking. Could we then extract from the candidates' positions, from the candidate's public statements, what are the things that every one of those 100,000 people should care about and then give them a personalized message so they can learn the candidate better? That would be kind of cool. Should we figure out how to do personalized and customized outreach so every single person can figure out how they should get involved in this election, where their voting booth is, where what the process should look like? That would be easier for them. So I think we can choose to live in this world. We just need to, it's really easy for us to fall into the bad places. We just need to regulate and legislate them out and set values that what we want to do instead is this better, this better idea of the way that we want to live in the world. So I think right now, if you say where is politics going to go with elections, it's probably gonna go bad. We are lucky that we haven't seen a lot of it in the United States so far, although other parts of the world have. But there again, I gave you some examples. President Biden, um, Donald, the former president Trump tried to do deep fake imagery of Taylor Swift endorsing him. Yes. Then she came out and said that no, I'm actually endorsing the other person. So I think we are gonna be battling that forever unless we can figure out ways to regulate against it. And so I think that's one thing we can do, we can get to a better place. So that's politics. I think on climate, I think we have a different set of issues when it comes to climate. I think on the climate side, like I think right now these AI systems need a lot of energy. They need a lot of energy and they need a lot of water in order to be trained. You know, I've I've had conversations, joking conversations that you know, GPT-5, the next version after GPT-4, might require the largest data center ever built, and it might require a nuclear power plant to power it. So like it just requires enormous sums of energy. Microsoft has been deploying uh data centers that require a lot of water to cool them. So therefore, all these generative AM models are going to require more and more computational power, just gonna require more water to cool them. We're already in a water crisis in a bunch of parts of the world. So all this is could potentially go badly. And again, I think we can just choose to do differently. So I think that we could be talking about questions around like local models or smaller models that fit on phones, that fit on devices, instead of relying purely on the cloud. We're also building all these tools without exactly knowing what we're gonna use them for. So instead, we can be much more directed in the kind of work that we do. We could be inspiring researchers to instead of being thinking about what's the possibilities of like if you had infinite computation, what does limited computation mean and how could we actually make the most of it? So I think there's like opportunities here for us to change it. I also do feel that like these AI systems will allow us to explore the solutions for climate change in a lot better ways. So I think that we need to figure out this middle ground of like how do we like restrict but also keep innovation going.

SPEAKER_00 40:01

See, we see the top of the iceberg, like it's chat, GPT, whatever. We ask questions, it r responds to us, we converse, we have conversations, we joke, but we don't see the what's going on with the colours.

SPEAKER_02 40:13

We don't see the cost of it. It's all hidden from us. And these companies are actually losing a lot of money right now.

SPEAKER_00 40:19

It seems so easy to us, like, hey, yeah, it's cool.

SPEAKER_02 40:21

Exactly. It seems free in a lot of ways. But the reason why they're losing money is because it takes so much energy, it takes so much stuff. And look, I support invasion. Innovation probably takes investment, you'll lose money in the beginning. But I think we should start opening our eyes on exactly how much this is costing because it makes you have to wonder whether or not we're spending our resources in the right way.

SPEAKER_00 40:41

Yeah, rather than looking cool, look smart. Exactly. That's right. And some insights on medicine. Oh, yeah.

SPEAKER_02 40:49

I mean, medicine, again, like I think there are real opportunities on when it comes on the medical front. Like I was talking to a bunch of researchers up in Seattle who are actually trying to figure out different ways to model protein folding because we can just explore lots of different drugs simultaneously. There's another set of researchers, also coincidentally, in Seattle, that started to work on robotics specifically so they can um they can automate all the lab skills that go on inside a biological lab, a research lab. Again, not to replace humans, but imagine if we could just speed up the process. Like the robots could be experimenting in ways that humans just can't because they have to sleep, they have to do all these things. We could just be rapidly trying to find new drugs, new opportunities for all of us. So I think there's a huge opportunity in AI to help just search the biological space to find place, find things that could benefit us as people.

SPEAKER_00 41:39

Let's speak a little bit about your career. Okay. Because your path is motivating and is a dream to anyone who wants to be successful, from Uber to Twitter, from Emerson Collective to uh being the pioneer of the AI in the DNC. How did you do that?

SPEAKER_02 41:59

Yeah, I mean, it's a good that's a good question. I mean, I think for me, the number one thing I look for whenever I try something new in my career is can I learn something different? And so, like, I you know, I joined Twitter early in my career, but like I became part of the executive team because frankly, I was just starting to ask myself different questions than engineers asked themselves. Just like, how can we actually marshal all of engineering to try to build something differently? So just be that like for me, cultivating this notion of curiosity and just asking questions is what always drove me to the next thing. So then for when I left Twitter, there were lots of opportunities to join other social networking companies, like a bunch of social networking companies. I was fortunate, they all called being like, come work for us instead. But I figured I wouldn't learn anything new. And then when Uber came around and I finally understand what they really wanted to go do, I was like, I don't know anything about robotics. I could learn something about robotics. We could build self-driving cars. I didn't know, like, I drove a car and I don't know how a car works. And so, like, for me, I just wanted to dive in because like my my passion and my curiosities, my passion drives me to learning. And I was so curious that it just like this was a huge problem space. I'm willing to like lose tons of sleep to work as hard as I can because I'm learning something new. Same thing came around for politics. So just like I felt the world was broken. We needed to go solve the issue of President Donald Trump and we need to figure out how to get him out of office. I didn't know anything about politics, but I'm like absolutely willing to roll up my sleeves and learn as much as I can in order to pull that off. So I think for me, it's always been this combination of like, what can I learn and never get worried, never get hung up on, I but I won't do that work. I'm like, if you need me to go fix your printer, I'll fix your printer. But then fixing my your printer is a way for me to then realize what's really going on. Like ask questions. Like the people who are who are told me about the broken printer will probably tell me what's broken next. I can just work on that and just keep on going from there. So You know, it's weird to say humility, but like really it's about willing to do anything. Like I'll roll up my sleeves and just do anything because that gives me a way to learn. And then from there I'll figure it out. Do do something else and do something interesting from there. So that for me, that's always been it.

SPEAKER_00 44:16

Oh yeah. I mean when we have this joke, when we see like anyone sees a tech guy, it's like, can you fix this for me? And I mean, that's not, you know, um basically, yes, and many people they do not, after a certain level in their own.

SPEAKER_02 44:33

They just don't want to do it anymore. But not do this. But this is the thing I talk to my team here at Emerson Collective about all the time. Like a lot of our jobs is to go and help nonprofit partners or people doing social work or whatever it might be, trying to make social change in the world. But if you just ask them what your problem is, they won't ever tell you what's really their problem. They'll tell you what they think is their problem or they'll tell you the thing that they want to be their problem. But if they have a broken printer, just go fix their printer. And then while you're fixing it, you can be like, what's going on over there? I'll fix that next. What's going on over there? And you get a better picture and you build trust. You learn, you learn how to talk to each other, you learn how to work with each other, fixing small things, and then something big will come of it. But you have to start at the bottom. Like you can't just come into the top. You'll just have the wrong picture of what's going on. Yeah. So I've always, you know, my wife, my wife says this about me all the time. Just like, I'm willing to kind of do anything. Just like throw me in there. I'll figure out a technical problem. In like, I'll work a I'll work coffee counter, I'll work fast food. There's some interesting technical problem there. Sure. And then something will come of it. But this is the fastest way to learn is just get really in contact with the problems.

SPEAKER_00 45:43

I mean, you really know, I mean, you you can be a good leader, in my opinion, if you have been in a very low level and raised up from there. Like I volunteered, I did this, I did that, and now I I have a firm and I know exactly what works there.

SPEAKER_01 45:59

Yeah.

SPEAKER_00 46:00

And it's really good to understand, it's really good that all the businessmen and leaders do understand that it's okay to fix a printer, to fix a cable, to do this and that it's a part of who we are.

SPEAKER_02 46:10

Like, yeah, I mean, like I again I have this with engineers all the time of just like, you don't want me writing code, but I'm happy to do it. Um just because like one, it gives me credibility also of just like we can speak the same language. I'm willing to do it, and then we'll just we'll figure it out from there.

SPEAKER_00 46:26

Yeah, that's so cool. When do you know that it's time to leave and start a new venture, new company?

SPEAKER_02 46:35

So I think it's two things. There's one which is looking in the inside. And for looking at the inside, for me, it's about have I stopped learning? Is there like is there nothing new for me to learn here? So that's a very inward way of looking. But that's the reason why I left Twitter, for example. Like I felt I learned, I I mean, not that there wasn't new things to learn, but my rate at which I was learning at Twitter had slowed down. But then there's also looking at the outside, which is like, what are the big problems of the world? And if you're in a fortunate position to make a shot at them, why aren't you? Like, and so, for example, like I gave the example of when Donald Trump was elected president in the US, like literally his inauguration day, I was like, fixing this problem of having a president I didn't agree with was way more important than making self-driving cars work. And so, like, uh-huh, if I had an opportunity to fix on that, why wouldn't I work on that? So that's the reason I went to the DNC, is because like it like it that seemed like a more important problem than I'm working on today. So I should go work on the more important thing. And so, like, I look at usually on those two lenses like, have I stopped learning? And there's something more important I should be doing in my time. I have, I mean, I'll tell my funny anecdote on this is mostly around email. Like a lot of people, like they view their life as like, I need to clear my inbox. I spent all day in email, like make sure it's clean, empty, and things like that. But like when you're 60, do you how would do you want to say that you spent a good portion of your life in your inbox? Like that clearly wasn't the most important thing to be doing with your life. So, like for me, like if there's if I have an opportunity to do something more important, it's almost my responsibility to go and do that instead of what I'm doing now.

SPEAKER_00 48:16

Are we most of us living inside a box? Uh, where are to take risks?

SPEAKER_02 48:23

Yeah. I mean, like, it's really easy. You know, companies see this all the time. It's called in the innovator's dilemma of just like, I know how to do this one thing, I'm doing it pretty well. It's too risky to do the next thing. And look, I get it. Like, people come from different backgrounds, people come from different financial statuses, people come from these things. But I think that like getting into a mindset of being able to take small risks that then build up to bigger risks, that bring up bigger risks. Because again, if you're not doing the most important thing, like why are you doing it? Like, if there's something more important you can be doing for your family and you can make it financially work out, then you should just do that next. And like just this that that train of thinking can can lead you down very interesting roads.

SPEAKER_00 49:06

Yeah, no, it's so I mean taking risks, thinking outside the box is what's going to make the drop bigger and bigger and change the ocean.

SPEAKER_01 49:15

That's what I think.

SPEAKER_00 49:15

Um, you are very active in Armenia, you are a board member of Tumul and supporting kids who want to get education. Just a little bit share about this as well. Like, where do you see Armenia's uh like younger generation going? There's a promising generation, I know I've seen many brilliant faces. So, what's next for Armenia?

SPEAKER_02 49:39

Yeah, I mean it's a great question. So I've been involved with Tumo for 12 years now. Um, and so I remember the day I first walked into the door, and it was mind-blowing what sort of what the team had built there, what Sam had built there, and things like that. Um, and so for me, seeing the opportunity, and now 12 years later, I've like I invested in some startups that TUMO graduates had created and things like that. So there is a positive trajectory there. And I think that things like TUMO are there to teach kids how to dream and how to dream and to think about problems they've never seen before and how to solve those types of problems. So I think if we can empower kids not to be stuck in their box, to be creative, to be risk-taking, to try things, to search for the most important thing they could be doing, I think a lot could come out of it. And I think that this combination of like a TUMO with like an old Soviet educational system and things like that actually set us up really well. If you asked me five years ago, I was already saying that because of our math training, Armenia is well poised to be the AI powerhouse of the world. And that was five years ago, before generative AI and ChatGPD came out. I still believe that to be true. We just need to do some targeted investments to really enable that to happen because I think we have a generation of kids who are mathematics people, who are scientists, who are thinkers. We just need to inspire them with creativity and open their eyes to what the real problems are in the world and they'll solve it. I have faith in these kids, they will solve it.

SPEAKER_00 51:11

Yeah, they are so amazing. And in the end, and I know a lot of people are going to watch tech people, people who are just doing their first steps in any venture, any endeavor. What would your advice to them be if they want to start something new?

SPEAKER_02 51:26

Well, one, make sure it's worthy of your time. Like you have to do the things that like you're gonna be excited by, passionate about, because like anything hard is gonna take all your energy. And so, like, you have to be worthy of your time, you have to be curious about it, you have to be excited about it, and then just throw yourself at it. Just go for it. Because I feel like that's the way you're gonna like no, you will not predict the end from the beginning. So you just gotta go and be excited to go for that ride.

SPEAKER_00 51:54

You will fail, you will fail, and then you will pick yourself back up.

SPEAKER_02 51:57

Do it again.

SPEAKER_00 51:58

Thank you so so much.

SPEAKER_02 51:59

This was so much fun.

SPEAKER_00 52:00

Yeah, it was. I wish we had more time. Thank you very much. All the way from Palo Alto San Francisco. This was Inside Brooke Your Screens. I was speaking with managing director and chief technology officer of Emerson Collective, Rafi Krikorian. Subscribe to Tribune AM not to miss the programs of Inside. See you soon.