S20E2: Oren Etzioni - Founder of TrueMedia.org

It's not up to you to finish the work to solve the entire problem. But at the same time you can't just say, well, I can't solve it, so I won't do anything. You have to do your part. So I am driven to do my part in whatever it is that I'm doing, whether it's I'm teaching a class helping to educate the next generation, or nonprofit like true media.org which won't solve all of this information but make its contribution. And that drives me you're listening to.

What Fuels you where we deep dive with CEOs, entrepreneurs and business leaders to learn more about their stories and uncover nuggets of wisdom we can all use. I'm your host, Shawna Swerland, CEO of Fuel Talent, an award winning recruiting firm based in Seattle. Today's guest on the what Fuels you podcast is is Dr. Oren Etzioni. Oren is the founder of True Media.org, a free use product that allows critical election audiences from around the world to quickly and effectively detect deep fakes. He was the founding Chief Executive Officer at the Allen Institute for AI, having served as CEO from its inception in 2013 until late 2022. He is Professor Emeritus at the University of Washington where he helped to pioneer meta search, online comparison shopp, machine reading, and open information extraction. He has also authored several award winning technical papers and founded and sold many companies. In addition to his role@truemedia.org he is currently a technical director of the AI2 incubator and a venture partner at Madrona. So good to see you. Thank you for being on.

It is an absolute pleasure. Shauna.

I feel like I've hit the big leagues now. I'm like, I got Oren Etzione on the podcast. It's a big deal for me.

I feel I've hit the big leagues. What fuels you where you've had the star studded cast in past episodes.

That's sweet. I like to start with some rapid fire and I'm also trying to squeeze in as much as I can. First rapid fire question is, which three attributes do you most admire in a leader?

Integrity is an obvious one. Loyalty is so, so important. I've seen leaders who are very accomplished leaders, but they will discard you like a piece of tissue paper if you no longer suit their purposes. And I think that's that's not a great quality. And I would say the third one, interestingly is the right balance between vision and execution. So there's the famous saying, right vision without execution is hallucination I think that was. But there's the counterpart, which is execution without vision is a nightmare. So I think it's essential to strike that balance between vision and execution and to have both. When leaders have both and at the same time, they have integrity, they care about their people, not just as pawns on a chessboard, but as genuine people. Now you've got something magical.

I love all of those, and I especially love those examples. I've never heard that saying before, but it's so good. I'm not sure if you're a foodie or into food, but if you are and you knew that tonight was your last meal, what would it be?

I am a foodie, and I'm the kind of foodie that really enjoys two three star Michelin restaurants in Paris, also really enjoy just street dumplings in China. But if I just had one, one more meal, one last meal, I would resurrect my grandmother, may she rest in peace, and I would have the standard meal that she always made for Shabbat. Which was the best.

Which was what? What was it?

Well, you know, there was chicken, there was noodles, there was soup. It's a whole thing. There was homemade chocolate cake from scratch to end it all. It was simple, but every dish tasted like love.

Oh, I love that. Okay, so if you. I know you're famous. I mean, you're famous in tech, but if you could be famous as something else or for something else, what would that be?

I just turned 60 and one of my.

Happy birthday.

Thank you. One of my ambitions is to go back to something that I really wanted to do as a. As an errant youth in Manhattan, New York, and that is stand up comedy. I enjoy making people laugh and I enjoy fast repartee, improv, all that. And so my next career, if I get to launch it, is stand up comedy and improvisation.

Well, you can go practice now. My husband is in a football league where they did. What is it called, where you whatever in their football league where they pick the teams. And he decided that whoever lost had to go do stand up in front of everyone. And everybody was freaking out because it was like the biggest nightmare for people. But one guy ended up doing it and it was a huge deal. My husband thought he was going to lose, so he was practicing and practicing and luckily his friend lost. But that's, that's brave. It takes a lot of courage.

Right. The, the term I think you were talking about is fantasy league.

Fantasy football. Thank you. His fantasy football league.

This fantasy football with real consequences.

Yes. Oh, yeah. So I'm curious, what is your biggest pet peeve?

Oh, wow. I am such a. I have so many peeves, I have to think of what the pet one. You know, there's the old joke that the person comes to the peevish couple in the restaurant and says, is anything all right? Right. But as opposed to, is everything all right? So that reminds me, one of my pet peeves is when you've had a very nice meal in the restaurant, everything is good, and then it takes forever to get the bill. I'm like, okay, we're done. I'm ready to go home.

That's the New Yorker Israeli in you. You're like, yala, let's go, let's go.

Yeah. 20 minutes to get the bill. Come on.

Yeah. Do you have any guilty pleasures?

I have a lot of pleasures. You know the words of Oscar Wilde, I can resist everything except temptation. A guilty pleasure is chocolate. It's wine, which we're just learning is not as good for you as we thought. And I read so much about how wine is bad for you that I've stopped reading those articles. Not stopped.

I think you're ready for your. Stand up, Oren. Let's go.

I am. I am. Absolutely.

I'll be your agent. I've always wanted to be an agent. Let's go. I'm ready. We'll make some money.

Okay. We'll start right here and we'll go international. The other guilty pleasures are, I play a form of chess online@chess.com called Bug House, which is actually four people playing on two boards simultaneously. You're partnered with someone and you're playing two other people who you don't know because it's the Internet and the whole game. You only have three minutes to make your move, so it's crazy fast.

Oh, it's like pressure cooker.

Yeah. It's a lot of fun as well.

Yeah. And are you. Do you like to read, watch movies, listen to podcasts? Like, how do you spend your spare time if you're taking in kind of that type of entertainment? And if so, what do you recommend the most? Is there a book or a movie that you find yourself recommending a lot?

All of the above. I. I just love. I love reading. I love watching. My wife and I are just right now watching Sl. Horses on. I think it's Apple tv, actually. It's funny. One of my. I feel like my claim to fame is going to be that I coined a word. You know that feeling when you can't decide whether something's on Amazon or Is it on Netflix? And you're like, I want to see the next episode, but what's it on? So there's no word for that. So I coined the word amiflexed. I say, oh, I can't remember what it's on. I'm amiflexed. Amazon, Netflix.

Anyway, I'm going to write it down. So Slow Horses. And I'm guessing, guessing it's good because you're bringing it up.

Slow Horses is really terrific. And book wise, I read a lot of kind of just trashy stuff. But also right now I'm reading the Rumor Game, which is kind of novel mystery set in the US around the time of World War II. Very interesting kind of expose. It's a detective story, but also an expose of some of the xenophobia and antisemitism that we encountered back then. It's kind of a historical novel. And what I'm about to read, because I'm going to be on a different podcast with the author of that book is Stories are the History of Disinformation.

Oh, I cannot wait to listen to that podcast. That's okay. Okay. I'm writing all this down. I got a book to read. I got a show to watch on my. What are we, Emma Flex. I'm going to Amiflex it.

Yeah.

So good. So you said your wife. How long have you been married?

It's right around 14 years and yeah, very, very happy and we still adore each other. One of our favorite lines to each other, hey, you know what today is? And the other one goes, it's our honeymoon, you know.

Oh, that's nice. Yeah, that's great. Other couples need to learn from that. So you, you said Manhattan, but I know that you're Israeli. Where were you born and like raised?

I was born in New York City and when I was a year old, my parents who were Israeli Americans, they got divorced. So my mom moved with my brother and I to Israel when I was a year old. And my dad stayed here, he stayed in New York. He was teaching at Columbia University. Ah.

And what was. I know that he passed away last year. I'm really sorry that he passed away. Was he still involved at all or like connected to Columbia?

So some decades ago, he moved and became a university professor, George Washington University in Washington, D.C. he was in the. When I came to this country, he was working as an advisor in the Carter White House. And he became very interested in policy and politics from an academic point of view. So he moved to D.C. to be closer to these things. And when he Passed away. He was 94. And he literally just retired from the university months before that. So he was.

Lived a good life.

Yeah, lived a good life.

And so your mom took you and you said your sibling to Israel.

Exactly.

And it's just the two of you?

The. The two of us and her. Yeah, we were a tiny family. She got remarried, she had my sister. So there were five of us then. But yeah, we started out just the three of us on a new continent. Yeah.

Well, nice, because you were young and got to learn Hebrew and a whole new culture. And I love Israel. One of my favorite places in the world. So how was your upbringing there? How would you describe, you know, the environment and like, how growing up there, I guess, influenced your curiosity?

Israel is really a fantastic place to be a child. When I was growing up there, crime was very low. People kind of watch out for each other, was very much a community. And so Even as a 6 and 7 year old, I would go down to the neighborhood to play and I would come back home in time for dinner, usually covered with sweat and grime. When I was 10, right, it was time to really engage in chores. So I was in charge of grocery shopping. My brother very much bemoaned that he was in charge of taking out the garbage. We had different chores. But the point is, you know, I went to the store. It's very walkable, so you could walk everywhere and then, you know, bring. Bring the stuff home and all that. So it's really, really a wonderful upbringing. And people are connected in a different way than they are in Seattle. So curiosity. I was always curious about people, about topics around me. I would say the biggest source of my curiosity was less Israel and more. Both my parents were professors. My mom is still alive. She's 90 and she lives in Israel. She's cowering right now in Tel Aviv as Hezbollah, right. Has been firing rockets indiscriminately. And the first one targeted the city of Tel Aviv. You know, we could spend the whole time talking about that, but the fact of the matter is they were both very, very intellectual. And so around dinner in both homes. I would be in Israel during the year. I would spend the summers in the US and we were always talking about. Not about gossip or about, you know, this recipe or this person who's annoying, but much more about the economy.

Ideas.

Yeah, exactly.

Well, they say, I mean, it's easy to start talking about nonsense and basic things. And then it is interesting how much energy you get when you're around people and talking about ideas. It's so much more inspiring than the boring, catty stuff that sometimes people get caught up in. So they're both professors, obviously very bright and academic. Were there other people, like teachers or relatives or anyone else that was shaping your worldview?

My older brother, he's six years older. He was also a mentor to me, much more about things like sports. He taught me how to play basketball, how to ride a bicycle, various things like that, and help to remind me that there needs to be balance. He's successful, actually investment professional, so it's not like he was a personal trainer. But he always reminded me of the importance of balance between the intellectual pursuits and sports and other things like that. I've definitely had teachers over the years that were so influential. I think we all have. I had, actually, a poetry teacher, English teacher in High School, Dr. Galvin. I still remember him. He was quite a. I don't want to say pretentious, but very serious demeanor. He had a mustache, you know, not a mustache, it was a mustache. And he would talk about poetry in a way that was just so inspiring. And I'll tell you a little anecdote that you don't typically hear. So he got me involved in the school literary magazine and so on. And I'm one of these people because I'm a computer scientist, right? So fascinated with math and computers, but at the same time with poetry and so on. So I came to him, and one of the questions we studied is, what really is a poem, right? Because poetry is such a diverse topic, right? From e. Cummings to T.S. eliot to the Iliad, right? So many different things. So I said to him, okay, Dr. Galvin, is the following a poem? I said, 1729 is the smallest number that can be expressed as the sum of two cubes in two different ways, right? Cubes, like, you know, three squared, three cubed. I said to him, is that a poem? And he stroked his mustache, and he said to me, warren, is that true? Is that a fact? I said, yes, Dr. Galvin, it's a mathematical fact. And. And he said to me, well, if it's a fact, it's not a poem.

Oh, interesting. I've never had somebody explain it like that.

Neither have I. And I don't even know if it's. If it's true, but it was certainly very.

We can chatgpt it. We can ask, talk about AI. We can like AI that. I'm kidding. So were you. It sounds like it was a pretty idyllic childhood. You were there in Israel, all throughout. Did you. I know you went To Harvard, obviously. We're fast forwarding. Did you serve in the idf?

I wouldn't say that it was an idyllic childhood. One of my memories at the age of 10 was of the 1973 war, the Yom Kippur war, where suddenly, without warning, Arab armies from Syria, Egypt, Jordan and elsewhere attacked Israel with no notice or with unclear provocation.

And it was on the holiest day of the year.

On the holiest day of the year, I was the one who, you know, grabbed my sister, who was about 2 years old at the time, grabbed her from her crib and, you know, took her to the bomb shelter. We didn't know how far their bombings, which have always been indiscriminately targeting civilians, how far they would penetrate. So we were terrified. We ran to the shelter. And that was actually the second war that I was in when I was three was the Six Day War in 1967. So there are wonderful things in Israel, but it's always been under a cloud of being surrounded by hostile neighbors attacking the civilian population. And of course we saw the most gruesome, horrific aspect of that on October 7th and the attack of the civilians there. So many wonderful things, but also some memories that we were always under a certain kind of siege.

Has that impacted you in a way where there's like PTSD or where you're in a constant state, A little bit of fight or flight anxiety or do you just. Because my Israeli friends are like, oh, this is just how it is. I mean, pre October 7, you know, oh, we're just used to the sirens going off. If you look at the interviews from the Nova Festival, it's like, oh, we kind of just, oh, they're always attacking and the Iron Dome will protect us.

I think that if you live in Israel and I've moved out and as the offensive technology, right, missiles and drones and such has become stronger, people have to make some kind of mental adjustment. You cannot live in a state of anxiety and fright all the time. And so, yeah, people kind of toughen up because I left when I was 14, so I didn't serve in the IDF and I'm. And I'm not used to it. I do find it very, very scary. I don't think people here can really imagine, imagine if, I don't know, they don't see how close things are. They don't understand the geography. Just imagine if Bellevue, right across Lake Washington was constantly firing mortar rounds and anti tank missiles at Seattle. That's what's happened with Lebanon over the last year. So over a period of a year, they were just firing mortar rounds, right? From Kirkland, from Bellevue, maybe, as far as, I don't know, Factoria, into Seattle, into your neighborhood. It's just. It's so. So difficult to imagine. But for me, again, if we're getting into this, it goes even further back. Both my parents were Holocaust survivors. They were both only children, and my mother was four, from Austria. She ran from the Nazis. She and her family hid in Italy during the war. There were many close calls. As the Nazis were in Italy. They were trying to hide from them. So she was a refugee. My father was from Germany. They came to Israel. So I don't feel that I have ptsd, but I do feel that I carry with me a heritage where the notion of antisemitism, of discrimination, of uncertainty, of being a refugee is not an academic thing. It's inherent in my family. I have no aunts and uncles because that was not a time to have more than one child. I could tell you so many stories, but let me give you an example of how it shapes me when people talk about undocumented immigrants in this country. And obviously, we have a challenge with immigration. We need to deal with it constructively. But when they're talking about these people who are seeking asylum, who are undocumented, illegal, that's my family. I relate to those people at a visceral level. That's who my parents were.

That makes sense. That really makes sense to me. So you left there when you're 14, you went to New York.

That's right.

And so you went to high school where in the city?

I went to Dalton High School.

Oh, you went to Dalton.

Yeah, which is featured in, you know, the Woody Allen movie Manhattan. And actually, I had a bit of a crazy time once I came here, because six months after I came, my dad was invited to the brooking Institute in D.C. so we traipsed over there. So I came here in the middle of eighth grade, spent half of eighth grade in Dalton, went to D.C. for ninth grade. Then he was headed to the White House, but then coming back, so he stayed in D.C. we came back to Dalton, but then my father and my stepmother decided, you know, this D.C. thing really makes sense. Crime was actually going up right where we live, which was the edge of Harlem. So we went back this time to Bethesda, Maryland. So I went to three different high school schools. And I. I like to joke that they just kept kicking me out, but, in fact, it was, you know, somewhat of a challenge.

Yeah, was that. Was that A tough thing. I've had a lot of friends and also just guests on the podcast where that is something that's like, you meet somebody now and they're this kind of strong, powerful, independent, confident person, and it's like, no, I had a shitty situation in high school or, you know, my childhood. Switching schools, especially if you're coming from another country, that many times can be super tough for kids. Nice. Or like typical teenagers, relentless.

I experienced a bit of culture shock and instability in a number of ways. Generally speaking, everybody was actually really nice, and I learned to make friends quickly. And overall, my high school years were very, very positive, but it did not contribute to my sense of kind of stability and for sure.

Yeah. So you obviously kept up your grades because then you ended up at Harvard. And what is it that I read about you being the first computer science major, like, you create. I didn't understand that, like, they didn't have the major.

So I was at Harvard starting in 1982, and back then they had an applied math major and there had people working with computers. But Harvard's point of view was that it wasn't clear that computer science was a bona fide discipline, a bona fide science. So they didn't have an official major. And I remember hearing a quote from a debate that the faculty had at the time, and they said, well, we don't have a major in automotive science. Right. If it's a real science, you have to call it a science, all these things. So right around 1982, 1983, they came to realize that this is not a fad or some sort of pseudoscience. This is actually a very real discipline. And they formally created a. What's called a concentration or a major in computer science. This was still paper forms, and I was being kind of an eager beaver. I was the first person to have my form signed by Harry Lewis, who was the person in charge of that at the time. And he looked up from the form, said, hey, you're the first computer science major. And it stayed with me because I felt like, okay, I meant to be a trailblazer in some way.

Yes, absolutely. Did you have a sense. I mean, how did you get interested in computer science? And did you have a sense of what you wanted to pursue as you. Did you figure it out at school, or did you kind of figure it out as you got closer to graduation?

So I'll tell you a funny personal story. And again, this is more of a personal podcast than the ones I usually do, so I'm very happy to share this when I was in high school in Bethesda, Walt Whitman High School, actually a terrific, terrific public high school. There we would have a mantle. And when I would come home from school, there'd be articles waiting for me, you know, orange scribbled on them. That meant that either my dad or my stepmother put them down for me to read. Now pick up the article and I could immediately tell if it was my stepmother because it would have things there about sexually transmitted diseases or drugs. She was trying to kind of warn me around the risks of being a teenager. And if they were about applied mathematics or computer science or AI, these would be my dad. Because what I was really fascinated with in high school was math. And he was very worried that I would become a pure mathematician, which he thought was too abstract, too divorced from reality. So his way of kind of cajoling me towards something more applied was to share these articles with me. And yes, I developed a taste for computers and AI a lot through that. And then at the end of high school I read a famous book, Godel, Escher, Bach, which also a lot of people read at the time. And by the time I was going to college, I was just absolutely fascinated with artificial intelligence.

Yeah, I mean, people weren't even talking about it back then. Like, definitely a trailblazer. And so as you were getting closer to graduation from Harvard, what were you thinking about as far as how are you going to make money, what are you going to contribute?

At Harvard, I had the privilege of studying computer science and AI. There really wasn't a lot of AI, so I got to go to mit, the MIT AI lab. I ended up getting a job there as a research programmer. I took classes of visiting students. So I learned about AI at Harvard. I studied cognitive psychology, philosophy of mind, philosophy of language. And I found those just riveting. There's some great professors and students to discuss with. And so I knew already that I wanted to be a professor. Like I wanted to go in my parents footsteps, but they were both sociologists, political scientists. I went very far from that. And so I was very focused on getting into grad school, finding a graduate program that allowed me to pursue my interest in research. So I was kind of biting my nails to see where I would get into for grad school.

Yeah, and then you ended up getting a PhD, Carnegie Mellon. What did you write your dissertation on?

I wrote my dissertation on machine learning. I got to work with Professor Tom Mitchell. He was a phenomenal, phenomenal advisor and mentor to me, a towering intellectual. And he's really one of the fathers of the field of machine learning back then. Now, a lot of what we call AI is really what's called deep learning. And deep learning is part of machine learning and so on. But back then, it was a lot less clear. There were a lot of people who were skeptical about how well machine learning can do. It didn't. It wasn't nearly as powerful as it is today. And he was an early pioneer who both helped to define the field and argued the importance of machine learning. So my PhD dissertation was in machine learning. And yeah, it was relatively early in the field. One name for machine learning that sometimes people use is the field of big data. Right. Because these machine learning algorithms use so much data. And I like to joke that I've been working on big data ever since it was just little data.

That's funny. It's grown over time. Well, I mean, in my industry, you know, I started recruiting before there was the Internet, like, we. And then we. Then there's LinkedIn. And so it's my year, 30 years in. And so now part of my, like, spiel is that I can look at the data more clearly because it's accessible. Although it was accessible back then, just none of us knew how to access it except for someone like you. It was like, early. Early. And so, okay, what brought you to Seattle? You came here because of uw? Like, were you interviewing all over the place?

I was interviewing all over the place, and I was fortunate to get a few faculty offers all over the country from Cornell, which is an. Ithaca, New York. We're at a very cold, small place.

Then you'd be in Ithaca.

Right, To Urbana Champaign. Again, University of Illinois. Urbana Champaign. Terrific university, Terrific department, but again, kind of in the middle of nowhere. And I ultimately was trying to decide between University of Washington and Brown University. And I was much more familiar with the East Coast. I hadn't really spent time on the west coast, but I chose Seattle and never look back. It's been a wonderful choice for me and for my family.

So did they have the program when you came or. Tell me about the program that you came into and how it's changed. I'm a Husky. I went to University of Washington. It's obviously a different school now. And how have you seen it change over time as far as your department and the overall school.

University of Washington for many years had people working in computer science, but it was a small group and it had only become a department recently. They also only got into AI very recently. So this is 1990, 1991, when I'm interviewing and taking the job. They had hired again, another great colleague, friend and mentor of mine, Dan Weld. He's a professor at University of Washington. And he came here in part because his wife had excellent prospects here. She's a pediatric pulmonologist, she works at Children's. But also because of mountains, he's a veteran. He climbed the top 100 peaks in Washington over the years. And so he came here for that. And he took a big risk because usually you like to come where you have colleagues in your field. And then he was able to attract another colleague. And then I was really the third person to come to this fledgling AI group. Fast forward to now. The. What used to be called the Computer science department is the school. It's the Paul Allen School of Computer Science. It's very large, it has many specialties, and has very, very strong AI group. Really some of the best in. In. In the world.

Yeah. And very, very competitive. Yeah. So many people would have just stayed on that path, including, it sounds like your parents and other people who are just, you know, thrilled to be professors. But you've had this thriving, kind of like untouchable career as an entrepreneur. How have you done all of those things? And which influenced which first, I guess, does the teaching influence you as a entrepreneur or kind of both? Do you bring that into the school?

What happened is that I was very focused on being an academic and very focused on getting tenure. And you can actually connect the dots, right? The instability that my parents experienced during the Holocaust, moving continents as a young teenager in the culture shock and then bumping around all these different schools, I was really focused on being a successful academic and getting tenure, because that was like, okay, as a tenured professor, I'll have stability. And I worked very hard. I was the first person in my PhD program to graduate with a PhD. I got tenure early at the University of Washington, which I very much appreciated. And then all of a sudden, I was in my early 30s and I had achieved my main professional goal. And the question suddenly became, now what? And some academics, they're a little bit like hamsters on a treadmill. Not to be pejorative, but it's like, okay, I'll write one more paper, and sure, I'm associate professor, but I could be full professor. Whenever I hear the term full professor, I like to say, full of what? Right. It just. It just didn't make sense. So I actually was looking for a new challenge. And Dan Weld, as I mentioned, actually kind of introduced me to the world of startups. And we put some research projects together and launched the startup and it was actually incredibly rewarding. I'm a big fan of playing games and one of the things I like to say to people that academia is like playing bridge. It's very cerebral, it's very precise in certain ways. And startups is like playing poker, right? There's a lot of psychology, there's a lot of adrenaline, it can be high stakes. And so I found myself as a bridge player who's kind of reached maybe in some ways his peak at bridge. And I started playing poker and, and then people would ask, well, which one do you like better? I said, I like both. I like bridge and poker.

Yeah, well, I mean there's, yeah, there's the ideation phase, there's the growth phase, scaling, selling. I mean you've kind of, you've had experience several times at all of these that all of these phases, you can like them all, but where are you kind of in your flow the most, are getting the most energy?

I find that I actually get most of my energy and it took me a while to really understand this because again, I grew up with computers and math and ideas, but I've come to realize that I get most of my energy from people. So when I'm working with people, whether they're students or colleagues or co founders or startup doesn't matter. But when I'm working with people who are energetic, smart, visionary and capable and we're working together trying to achieve something, that's what gets me really excited. So it's not so much, you know, the money of it or the, it's just people working together, coming together to do something that's never been done before and that can be a research paper, that can be a startup, that can be a non profit, but that's what gets me really excited, being at the edge with a great team, linking arms.

And doing it together and inspiring one another. So you've had several of these companies. I know we can't favor our children and say this is my favorite or like which, which of these experiences did you learn the most? I know you had metacrawler, Faircast. Decide. When I first moved to Seattle, that was right when decide I'd moved from New York and I remember decide because I loved that idea. It was such a good idea. Tell me a little bit about these ideas, how you came up with them. We can't go into all of them too much because we would just like talking about the Middle East. We'd go, that's a whole podcast in and of itself. But I am always curious about the ideas, like, which comes first? Are you sitting with someone and they have an idea and they're like, come on, Oren, are you the idea guy? And we talked about ideas and execution. Where do you fit into all that? And I want to hear about these ideas.

Each startup really is its own story and comes together in its own way. I'll tell you a very abridged version of two. So the first startup, which involved MetaCrawler, which, to those most people don't remember, MetaCrawler, before there was Google, there were quite a variety of search engines and they were called InfoSeek and AltaVista, that all kinds of things. And people didn't know which one to use. And each of them had its own quirks. Each of them covered only a small fraction of the Internet. And a grad student that I work with, Eric Selberg, had the idea to create something called metacrawler, which is something that covers all the search engines. And when you have a query, why don't you ask all of them simultaneously and get the results back in a unified report? That was the idea of MetaCrawler. It was a wonderful idea and ended up being the web's leading meta search engine. One of our lines was, anything you can do, I can do meta. Right. Alluding to the old, you know, anything you can do, anything you can do.

I can do better. Yeah, I like that.

Exactly. So we were like, anything you can do, I can do meta. Anyway, I learned a tremendous amount of that startup because we did so many things wrong. I learned, for example, that we had such technical talents, just brilliant technical people, but you have to have a business team that's at least as good and so on. So people often ask me for advice about startups, and tongue in cheek, I say, you know what, you should immediately go to your second startup because your second startup, you know so much more. It's kind of like second marriage.

I'm thinking that. I'm like, are we saying that? Yes, we are.

Well, yeah. There's actually a line I have to share with you about second marriages that people are sometimes very hesitant about second marriages. They say a second marriage is the triumph of hope over experience. And my perspective is actually, no, it's the opposite. A second marriage is the triumph of experience over hope because.

Yeah, that makes much more sense. Yeah. So MetaCrawler, do you remember Ask Jeeves?

Oh, I do. I was actually very involved in working with the Ask Jeeves team early on. I was a A consultant to them, building. Building on the MetaCrawler experience. So I. Yeah, I remember that.

I loved that company. And so tell me a little bit more about the MetaCrawler journey. And so you learned a lot. That's. Obviously you made some mistakes.

So, by the way, just apropos, Ask Jeeves, because a lot of people aren't familiar with it. Ask Jeeves was basically an early version of ChatGPT. The technology wasn't there, so it didn't work as well, but it was that same idea that you ask questions and advanced AI technology is going to give you the answer. You don't have to search, you can just Ask Jeeves. Right.

Ask Jeeves. You know what? I use, like, every day, and I don't know a lot about AI, but I use Perplexity, an app called Perplexity. I don't even know. A friend of mine told me about it, but I like it because. Shows me the source, which is very important.

Yes.

The information. Yeah. As opposed to. Yeah, because otherwise you don't really know where information is coming from. So tell me. So you sold that. Sold MetaCrawler.

MetaCrawler got integrated into a company called NetBot. Because we made so many mistakes, like not even successfully hiring a CEO, we ended up selling the company 18 months later. But it was a very positive exit. We didn't raise too much money, and it was kind of a hot space at the time. So we sold it to a company called Excitement, which then was bought by At Home, which was bought by AT&T. Kind of the bigger fish gobbling the smaller fish. I learned a lot from that about how to start a company, about how to work with venture capitalists. I worked with Madrona Venture Group, a group in town that I've been working with for almost 25 years now. And my second company, Faircast, is a company that's probably my dearest one and my most successful one. I founded it myself with some of the background knowledge, and I wasn't one of these people. Sometimes people talk about a serial entrepreneur. Right. You found the company. Then six months later or two years later, you found another one. It was actually seven years between when Netbot, my first company, was started, and Faircast in 2003. And Faircast was started with the following very personal experience. I was on a plane and I got to talking to the people seated next to me, and I realized that I had paid more for my ticket than they did. And I bought it later. I'm sorry, they bought it later. I bought it earlier. I always thought, okay, you know, you buy your ticket early.

Early bird gets the worm.

Yeah, to get the best fare. Right. And it turned out that that's not always the case. And I got exposed to a practice called yield management, which is still done today by the airlines. Right. The prices go up and they go down. It's quite an unpredictable process. And I realized that maybe we can use AI and big data and machine learning, all those technologies, to help figure out when is the right time to buy your ticket. So I started initially a research project at the University of Washington, and it was called Hamlet. And we said, to buy or not to buy. That is the question.

You have all these amazing lines. First of all, I can't believe you remember them all. Half the time I've got brain fog. And I'm like, I know it was something brilliant. I just can't remember what it was. But the fact that you remember them all. I think we started to talk before the podcast came on. I was like, do you have any weaknesses? And you said, sense of direction, but your memory is just, like, wicked. It's. It's unbelievable. So why is that one your baby? Because it was, like, in your mind, the best idea.

It was a very creative and innovative idea. Nobody before that had the idea to empower consumers with the knowledge of when to buy their. Their ticket. And I'm always pro consumer, so this was kind of a David versus Goliath situation where the airlines are armed with tremendous troves of historical data, with all kinds of information, and they're changing the prices on you. And our work tapped into this frustration that people have, like, the airlines are playing games, and people really want to know, particularly if they're on a budget, when's the right time to buy your ticket to get the best price? And it's frustrating because you're buying the same good, you know, a seat on an airplane to fly from New Seattle to New York. Well, why should the price be different by $300 just because you bought it on a different day? So it was really what's called a pain point. Right? When I talk to people about it, they'd inevitably have a reaction like, yeah, I need this. I want you. Once this got a little bit publicized, students would come up to me after class and say, professor, I need to go visit my mom, and I'm on a limited budget. Can you tell me when's the right time to buy my ticket? And I said, look, I can't, but what I'm going to do is I'm going to build a website to support that. And so this was actually incubated by Madrona. My friend and colleague Matt McElwain, who's a leading venture capitalist, invited me to launch this effort. And. And people talk about working in the garage, working in the basement. We were working on it in Madrona's daylight basement because they're on the 37th floor.

Oh, yeah, it's beautiful.

The 36th floor.

Matt was on the podcast, by the way, if you want to listen to his. Was fun. We had a good talk. Lots of the Madrona team has been on. But what was the business model? How did Faircast make money?

The business model, we went through a number of different ones. I would say the most exciting one we actually were not able to successfully implement. So the typical business model was what's called a bounty model, which is ultimately, you sell the ticket, it's on Alaska or United, and they would prefer you to buy the ticket from them as opposed to on Expedia or a place like that. So they would provide us with what's called bounties. Right. So that's a very typical business model in the industry. If you bring customer to an E commerce site and travel, right, you're buying tickets. It's E commerce. That's a very typical business model. The really exciting one I had was the idea that, let's say you're looking at a price of a ticket and it could be international, coast to coast. It's an $800 ticket. And we predict that the price is going to drop over the next three weeks to $600 or less. What we could do, if our model is good enough, is offer you that ticket right now for $700. Best price you'll find.

They're like buying futures.

Exactly. Basically, it was kind of a.

It's like a hedge fund.

Exactly. So. So we would offer it to you for $700. You'd say, hey, this is great. I'm going to lock in the price. Then we would buy it for 600. And what would happen then is you would get a great price and we would actually make, you know, quite a lot of money on the ticket, which is essential because these models are never always right. So sometimes the price would go to $900 and we'd lose. But we ran extensive simulations showing that this approach was actually a very strong. Very strong money maker.

Yeah. Oh, for sure. And decide. Similar business model, I'm guessing, with electronics and. And what else? I mean, I remember it was electronics. It was also like dishwashers. It was Everything, Right.

We started with electronics and we were very much. The side was in many ways Faircast 2.0. And in fact, we hired Mike Fridgen, who was the CEO of the side was the VP of product and marketing at Faircast. So it had elements of, you know, the bend back together again. And the site has its own story. I started it actually with some very talented and tenacious UW undergrads who came to my office to discuss their ideas. But. And we started with consumer electronics because you have this experience, right, where you buy something and then three weeks later they come up with a new model and like, why did I buy that? You know, I should have just created it. Yes. So a lot of interesting issues there. And we kind of tried to expand the value proposition from know when to buy to also know what to buy. Right. There's also, like, if you're buying a camera, there's so many models and so many cameras. Which one? Which one should I buy? And then we went from consumer electronics to a variety of other categories, including appliances.

So my question for you is, like, you know, you're bringing up people like Mike Fridgen. Awesome. I love Mike. And then he kind of came with you. How. How do you think about building teams, recruiting? Is there a methodology? Or do you go with gut? Or are you more kind of background? How much. How much weight do you put on experience versus potential?

I have two sayings that really guide me just to set the stage here. And the first one is, life is too short to work on a project you're not incredibly excited about. So when I get into a project, the first question I ask is, am I really, really excited about this? I'm going to put so much energy, time, blood, sweat and tears into this. Is this worth it? And then I've developed the saying, life is too short to work with people you're not incredibly excited about. And so over time, I've worked with people, I've hired people, and, you know, they're very smart, but they're difficult to work with. Or there's trust issues. And I've inevitably regretted that. So now when I go to a process of putting together a team, hiring, my first question is, do I want to work with this person independent of qualifications, experience, Do I want to work with this person? Will other people want to work with this person? And if the answer is maybe that's not good, that's not.

Yeah. I've had people say, like, I'm on a business trip and the flight gets delayed five hours, and I'm like, Stuck in the airport, do I want to sit next to this person or am I avoiding them? You know, that type of thing.

Yeah. The delayed flight test was literally one of our hiring criteria at Decide, as we were hiring people, we would formally ask, okay, how did they do it for you on the delayed flight test? So, absolutely, this was just part of the process.

Yeah, I think it's an important one. But it also can play into bias if you're not careful that, like, we all want, you know, the whole saying, we want to be around people who think and look and walk and talk like us, and we need diversity for diversity of thought. And, you know, our gut goes to like, oh, these are my people. These are not my people. But at the end of the day, I think it's more the. A lot of companies have the, like, no asshole rule, and that's the bear, you know, or they'll take somebody out to lunch and see how they treat the weight staff or just kind of, are you a good human being?

Are you a mensch, as we might say? You're absolutely right, though, that these softer criteria do have a potential for bias there, and there are ways to work with that. But it's tricky, and I do acknowledge that to deal with bias and to deal with the fact that our gut may clear somebody, but that's not the only criterion. There is this element of thinking fast and slow. And so in addition to that, I would also try to spend enough time, because hiring is so important, particularly for senior roles, particularly if you're working with co founders, to make sure that I have enough information. So one of my favorite observations in life is that sometimes when we look at big decisions that we make, whether it's buying a house or a co founder for your startup or a key executive role, on a proportionate basis, we spend more time trying to choose, like, which battery to buy or which ice cream flavor to choose than we do for this key role. Right. I mean, obviously ice cream, you don't spend weeks and weeks, but sometimes, like, okay, we'll interview the person for a day, we'll check a couple of references. They look good, let's hire them. And there's the famous thing, hire slowly. Right. So you really want to get as much information and that you gave example of the lunch test, but really understanding what they did in previous roles, I'm very fond of giving them a task. Say, hey, you know, can you write up your thoughts for me about this topic? Having a disagreement? Say, you know, respectfully, I really disagree with you about this. And Seeing how that goes, doing reference checking very, very deeply. Not just their references, but backchannel. There's just so many things you can do to increase the amount of data you have. Another favorite saying of mine is an ounce of data is worth a pound of my intuition. So yeah, that personal connection, that's the intuition. But data can be very important in coming to realize, wow, this person is amazing. Or realizing, hey, you know, they sound good, they feel good, but they're not really, they're not going to deliver. So I'm a great believer in hire slowly. And of course there's different roles and there's different things, but I don't just trust my gut and I don't just trust what's on the piece of paper, what's on the reference. I am just incredibly hungry for as much data as I can to get on a critical hire.

That's really important. And do you believe in or subscribe to like personality assessments?

I've read about that, particularly from Ray Dalio. Right. In his book Principles where they had extensive tests like that.

Yeah, Bridgewater. That's a whole other crazy culture.

Right, Right. I've never used those personally. That doesn't mean that they're not, that they're not good. So I guess I would say I'm open minded about it, but that's not an approach that I've used personally.

Yeah, tell me. 2013, a little over. That's the same year I started Fuel and I worked a lot with Vulcan, so I knew a lot about when AI2 was starting. Tell me how that idea came to be and Paul Allen's involvement, all about it. I mean where, where it was and where it is now, like crazy growth. It's been really inspiring to watch.

Paul Allen, the late Paul Allen is, is somebody that I respect enormously. And his autobiography was actually called Idea Man. So if you want to talk about. You asked earlier about ideas. He is the, you know, the genuine article. He was such a, such a visionary and he had in his head the idea to build an AI institute out of Vulcan. It was kind of a bolt of lightning. On a clear day, I was suddenly approached by a recruiter and said, would you be interested in doing this? The timing was wonderful because I started bemoaning the incremental nature of academia. You know, one more paper, one more algorithm. And I was looking for an opportunity to do something bigger. The side was winding down. It was being sold to ebay. And around that time they said, well, you know, come talk to Paul Allen. And so I did and we had just a fantastic conversation because he was a genuine intellectual. He was asking the hardest, boldest question. He wasn't asking, hey, would an AI institute make me look good or will it make me more money? Not at all. He was asking, how many Turing Awards, right, the Nobel Prize of computer Science do you need to get to artificial intelligence? And it was just so inspiring and invigorating to talk to him about his vision. And of course, he had the resources to back them up. So I went through an interview process, accepted the position, and people asked me in the media, oren, why are you leaving a tenured position at the University of Washington to go and just, you know, start from nothing? Right. There wasn't even a plan for the institute when I was hired. I was hired into Vulcan, his holding company, to write the plan for the institute. And I said, the sky's the limit. Paul Allen gave you the feeling that if you were able to think boldly and execute effectively, the sky's the limit. And for me, that was just incredibly invigorating.

Well, and it's been the case. I mean, you've grown it into a significant force, like major research. So many incredible companies coming out of AI too. Like, what are you most proud of and what have you learned from that experience?

So we started in 2013. By January 1, 2014, we were launched as an independent nonprofit called the Allen Institute for AI. Over the years, it grew to, over the nine years or so that I was running, it grew to 250 people, more than $100 million budget. We created an incubator, the AI2 incubator, which has then been spun out. It's now run by Jacob Kolker and a number of folks who were employees at the institute and now running the incubator. And I'm really proud of the different projects that we did. We could talk about them, them, the different people. I'm also really proud of the fact that it's a sustainable institution. So I haven't been running it for a couple of years now. Ali Farhadi, who's somebody, another UW professor that I hired into AI2 and helped him launch his company. Xnor, which was a spin out of AI2, was acquired by Apple, became the basis of Apple's office here in Seattle. And now Ali came back after being hired by Apple, came back to run AI2. He's doing a phenomenal job, and I really hope that he gets the kind of support that I got from the late Paul Allen, real visionary. Unfortunately, Paul Allen passed away and we're still waiting for a foundation to be created that would implement his vision. That's what been a more, a slower and more involved process than I think anyone expected.

Yeah. And well, it also has just contributed so much to our community, to this area. I mean the talent coming out and the development of ideas. It's a legacy thing that you should be, I would imagine, super proud of. I'm curious how over the course of early days, one of the OGs and AI and talking about it, you know, it's its potential to really benefit human now at this point and obviously we're going to get into true media.org, like how do you think about AI as far as optimism or pessimism or what it could do for humanity?

AI often framed negatively and people often talk about the risks, the downsides. Will AI take over the world? Will it take our jobs? All these things. I am an AI optimist and AI optimist for very specific and very Jewish reason. Judaism really believes in the sanctity of human life. Human life is so important and I see so many opportunities to save lives using AI. I'll give you quick examples. One of them self driving cars, a project we were working on for decades and it's getting closer all the time now in San Francisco, Austin and elsewhere. Right. You can actually get into a self driving Waymo Taxi. The important thing about self driving cars is not to put Uber drivers out of a job. It's the fact that we have 40,000 deaths each year in this country on our highways. We have more than a million injuries. Everybody knows somebody who's hurt in a car accident. And the Studies show that AI based driving can reduce that by 80 to 90%. We won't have distracted drivers, people driving under the influence, so we can save lives using AI. The same thing is true in our medical system. In hospitals, people don't have access to appropriate medical care. Nurses are burnt out, are out of jobs. So again, AI systems can really help improve the quality and reduce the cost of medical care to save lives in hospitals. Third leading cause of death in American hospitals is physician error. So it's funny, people talk. Sometimes people ask me, do you feel like Oppenheimer, you're working on inventing the atom bomb? And I say, no, it's quite the opposite. I have a moral imperative to work on effective AI systems to save lives.

Yeah, but there is the threat. I mean, well, just the frustration and the threat of the AI generated deep fakes, which is now what you're addressing with true media Dot org. How did you. I mean, the timing is just incredible. And I'm grateful as somebody who's just always confused about what's actually true, that it exists. So tell me more about what sparked your interest in this specific area and what made you feel, like, compelled to do something.

I'm an AI optimist, but I'm also an AI realist, so there's actually quite a story here. I was invited to a meeting of some AI leaders with President Biden in the summer of 2023. It was in San Francisco. We flew there, and I went there with some of my exciting ideas about how AI can benefit humanity. The mission of the Allen Institute for AI was AI for the common good. And I've always thought about, how do we use AI to help people? My idea had to do with using AI to help people see better. Right. So many people have difficulty seeing, particularly as they get older. We could use AI technology to help with that. So I was there to pitch that idea to President Biden and senior people in his administration. Remarkably, I came out of that meeting with concern raised by somebody else, Tristan Harris, who talked about how AI could influence the election. Deepfakes. AI generated content that's the source of disinformation, could really tip the scales in an election year. And timing wise, I started saying, well, let's look at what tools are available. And I was shocked to find out that for the general public and for many, many media organizations, even some of the largest ones, there aren't adequate tools available. And so I said, well, okay, let's build one. Let's build a deepfake detector. That's simple. You can just take a social media URL where you've seen something that you're not sure is right. Maybe it came from Ukraine or Gaza, or maybe it's a clip of an interaction that Vice President Harris had or former President Trump. And you want to see, is this video real? Is this image actually bonafide? And we've built a technology to take that URL, extract the image, video or audio, and analyze it and give you an assessment within minutes, a minute or so. Is this real or is this fake? That was our vision. We got backing from Garrett Camp, who was the co founder of Uber, and this is through his philanthropic organization, camp.org and we were in a race against time to build this in 2024, in time for the election. We launched in beta in April. The New York Times covered that. It was a very exciting moment for us. Fast forward to September. We launched this to the general Public. If you go to truemedia.org right now, anybody can issue a query and see whether what you're looking at is fake, manipulated by AI, or if it's real and bonafide. And it's become a really important thing because the election is very close. And in the run up to the election, we've seen this all over the world time and time again. You see deepfakes rearing their ugly head. And already in this election, if you want, I'm happy to share examples we've seen.

Yeah, I wanted to ask, are there examples that you're like, oh, my gosh, this is the one that went viral. I mean, I've heard about a few, but even when I hear them, I'm like, is that true?

We are often not the people who are being targeted by disinformation, by influence operations, and so we don't see them. But now that we are scouring the Internet and have thousands of people uploading them, I cannot believe how much garbage is out there, whether it's scams featuring Elon Musk or the CEO of Google pushing some questionable investment, or whether, Let me give you maybe the best example. Taylor Swift, when she announced her endorsement of Vice President Harris, said that one of the reasons was deepfakes, and specifically the deep fake of her endorsing Trump that he forwarded, that he shared in true social. So we've seen that we detected that as fake. We've seen images of Taylor Swift baking cookies for Trump. Again, fake. We've seen so many of these. There are fakes on both sides of the aisle. Don't get me wrong. We've seen plenty of fakes of Trump, but the point is that they are often viewed by millions, if not tens of millions of people. It's hard to track exactly how much effect they have, but it's a very worrying trend.

Yeah, well, the other part that's so hard is that information is being consumed so quickly. I mean, even just in the early days of the, of the war, talking about the hospital that got hit, you know, that Israel hit the hospital, and it was like by the time that kind of went viral, they then were course correcting after the fact. Like, it's, it's, it's crazy how quickly you got to stay on top of this stuff. And it's incredible also that you came out with it being accessible to the public now. Like, we got to get this, we got to get the podcast out asap.

I hope you do. We will. It's not a coincidence. It's a very Hard. It's very hard work on the part of the team. And from day one we said we are deadline driven. You know, even an early stage startup, when you're working towards launching a product or getting a financing round, you can always say, well, okay, we'll get a bridge loan, we'll move this by a few weeks. I told the team from day one, nobody's going to move the election for us. No, we have to hit our deadlines.

We have a hard deadline. How many people do you have working.

With you right now altogether? Including interns and volunteers and contractors? About 20 people. It's a very hard working, very idealistic group. We all really care about doing the best we can to solve this problem or at least to address it before the elections.

Yeah. Well, I'm really excited. I'm going to put in some videos and check them out because I'm constantly looking and just shocked by how toxic and polarized the world feels right now. And I'm hopeful that we get on the other side of this and people can kind of come together and exhale a little. But this will give me some confidence. What is the long term plan for the business?

It's a non profit, so we are really focused on what's going to happen in the run up to the election, which is weeks away. So very focused on that. After that we're going to take a deep breath and look at different options. We've had a lot of outreach from major media companies, we've had outreach from the legal system. Deepfakes are not just about politics. What happens to evidence in a world where pictures are so easily manipulated and faked and likewise video and audio. So there are many arenas and we're going to do a very careful assessment of what's next. But frankly, we don't have time for that because it's still a very small team running hard to make the tool as effective as possible before early November.

Yeah. Well, thank you for doing it. Okay, my final question for you is what fuels you?

Well, other than chocolate and, and no more wine. Right. And food. Right. We mentioned, you know, I love food. Really what fuels me is a combination of two things. One is the feeling that we can have an impact. Right. And it doesn't have to be that I'm going to solve deep fakes or I'm going to save everybody's life. Again, the Jewish saying is if you save one life, it's as if you saved the whole world. And another Jewish saying, I'll actually say this in Hebrew, Which means it's not up to you to finish the work to solve the entire problem. But at the same time, you can't just say, well, I can't solve it, so I won't do anything. You have to do your part. So I am driven to do my part in whatever it is that I'm doing, whether it's I'm teaching a class, helping to educate the next generation, or nonprofit like True Media.org, which won't solve all of this information but make its contribution that drives me. And then the second piece, as I mentioned earlier, it's the people. So when I'm working on that really important problem with a great team, we're in this together, we're pushing the envelope and we're doing the best we can and sometimes having a wonderful time doing it right. A lot of laughs, a lot of camaraderie. I can do that all day and all night.

Yeah. Well, this has been so much fun for me. I could do this all day and all night. So thank you. Thank you, Oren. So nice to see you. Really appreciate it.

Shauna, thank you for having me on. Thank you for your conversational style, which really brings out both the most personal and the best in people. I love your podcast and thank you. I really appreciate that the opportunity.

Thank you for listening to the what Fuels yous Podcast. If you'd like to check out past episodes to hear from more business leaders, go to fueltalent.com podcast and if you have a minute, please leave a review and rating on your favorite podcast app or share this episode with a friend or colleague. Please share any feedback or interview suggestions for other guests guests by sending a message to podcastueltalent.com I'm Shauna Sperland and thanks again for listening.