Joab Rosenberg — Startups and Philosophy: Ep. 32
Hosted by Rabbi Dr Ari Lamm
The limits of AI make it clear that humans aren’t about to be replaced by robots. But anyone who reads the Bible understands that humans have a lot of room to improve and even transform themselves! And increasingly, we’re learning that the best way for humans to become better is for us to become closer, and that means more discussion, more conversation, even more disagreement!
So how do we get better disagreement? How do we bring the genius of scriptural and philosophical traditions into conversation with the know how of Silicon Valley (or Tel Aviv)? To unpack all this, Rabbi Lamm spoke with Joab Rosenberg, the Founder and CEO of ment.io, which has been called “Israel’s most philosophical startup.”
This is Good Faith Effort with Ari Lamm. And here’s your host, Rabbi Dr. Ari Lamm.
Ari Lamm [00:00:21]
Hello, hello and welcome to Good Faith Effort, The World’s Most Dangerous Bible Podcast, the podcast, where we show you how the values and ideas of the Bible can illuminate the most important conversations in society, from politics to pop culture and beyond. Folks, we got an amazing episode today. We have with us the founder and CEO of one of the coolest startups in Israel, Ment.io. It’s been called Israel’s most philosophical startup. The one and only Joab Rosenberg is with us today! And we’re going to talk about productive disagreement. So we’re in this period of the calendar right now, late spring and summer, when the Bible tells us that biblical civilization and ancient Israel imploded. Jerusalem was besieged and the temple eventually destroyed, right around this time of year. And even after the Jerusalem temple was rebuilt, it was destroyed again, by the Romans, also around this time of year. So both times ancient Israel was sacked, it shared a time of year in common. OK, but did it share anything else in common? Well, the answer is yes. Both times ancient Jewish society had pulled itself apart long before any invaders came along. So the first time Israelite society is destroyed, the Bible makes clear what the problem is. After King Solomon died, the Israelite kingdom actually split into two. The Israelites couldn’t hold themselves together for even one generation. The northern kingdom ends up being destroyed first, and they’re never heard from again. That’s what we now kind of call in pop culture “the 10 Lost Tibes,” and the southern kingdom had to fend for itself and found itself easy pickings for the Babylonians. The second time Jewish society is destroyed, we find the same problem of toxic disagreement multiplied a thousand-fold. You can read any record of ancient Jewish history from the Dead Sea Scrolls to the rabbis of the Talmud, and you immediately see how deep the divisions were in the ancient Jewish community. So the historian Josephus, who was a firsthand witness to the destruction of the Jerusalem temple by the Romans, writes tragically about how bitterly the Jewish defenders of Jerusalem fought with and disagreed with each other, all while the Roman armies camped serenely outside. So the most famous civilization in ancient history was torn apart by disagreement. And yet, on the flip side, the sacred literature of ancient Israel, and of two major religions to this day at least, valorizes disagreement. So in the Bible, the prophets argue with God. In the Talmud, the rabbis argue with each other. In the rest of Jewish literature, all the students argue with their teachers. The upshot, of course, is that discussion, dialog, even sharp disagreement, can be both the seed of destruction or the source of cultural renaissance. The responsibility of a healthy society is to incorporate as many voices as you can to create a symphony rather than a cacophony. And today’s society has both unique challenges and unique opportunities on this front. The nature of the Internet and social media means that we have more voices vying for attention than ever before and access to more information than ever before. The coordination problems are extraordinary, but because of those realities, we’re also capable of doing amazing things that our ancestors could only dream of. I mean, we each have the entire accumulated storehouse of human wisdom in a device in our pockets that we carry around every day. It’s insane. So how can we use the wonderful blessings of our age to meet its challenges? Well, to unpack all that, I brought on a brilliant founder who’s thought more about this than perhaps anyone out there. He’s the founder and CEO of Ment.io, an incredible thinker, the amazing Joab Rosenberg. Joab, thanks so much for being here.
Joab Rosenberg [00:03:46]
Thank you very much for the very relevant introduction.
Ari Lamm [00:03:49]
Amazing, I love it. That’s how we roll on Good Faith Effort. So I want to kind of step back for a moment and think about the space that you’re in as a whole, education and conversation, dialog. So if you look at a list of the longest continuous institutions in existence, like an absurdly large number of our universities. So like the University of Bologna, it’s still a prominent school today, it was founded in 1088. And yet in terms of operation, universities today are essentially identical to what they were a millennium ago, give or take a cafeteria here and a gym there and a fitness center here. So they have classrooms in which professors lecture and students study. So things have changed at the margins, but the product is shockingly similar. So why have we not seen more innovation in terms of how education is delivered? And is this changing now?
Joab Rosenberg [00:04:39]
Well, I think the straight answer to that is that people feel that it’s good enough, the professors are usually conservative, so they wouldn’t want to change much even if they say so. And the more prominent professors are usually older. So it means they would less like the idea of implementing very new technologies, which usually the students like no better than the professors. And of course, as a professor, you would usually prefer to know better than your students. I used to lecture myself in two Israel universities. I do think the system is working well enough, which I assume, by the way, Trae Stephens and Peter Thiel will not agree, but I think it works well enough. But still, there is much to be done and there’s a lot of progress that can be done if we do allow some technology to go into higher education. And I think Covid has done well in that respect. It did very badly in other respects for the health of many, many people. But in terms of adjusting the universities, the very conservative universities to online teaching, I think there was a major push to adopt new technologies because there was just no other choice.
Ari Lamm [00:05:51]
So can you tell us what is the intervention that Ment.io is trying to achieve?
Joab Rosenberg [00:05:57]
Well, I would say initially the most important thing is our vision. On the one hand, we really like the Internet. As you said at the outset, it’s an amazing tool. We can speak to one another all around the world in a very easy, straightforward way. But on the other hand, I really hate the Internet because I think up to now it really promoted shallowness in the way people think. And it was much more cacophony than symphony. So we know, first of all, it was all about the value of freedom of speech. And back in 2011, I think even Mark Zuckerberg was very proud of the Arab Spring that he thought was helped by Facebook and by other platforms because of freedom of speech. But then I think since 2014, and especially in the last three years or so, we’ve seen post-truth. Right. And we’ve seen the price of this very shallow thinking and the fact that we allow people to argue that the earth is flat and we allow people to argue that immunization is really risky and all kind of stupid stuff that the younger generation, you know, they cannot judge because they hear every different view that they would like. So we decided we want to do something on that. And of course, there are many people who are trying to do things, so we try to do our little contribution. You know, it’s not for us to finish the world, but we cannot decide not to do it. So we built a platform that was designed from the first moment to promote deeper discussions, more reliable discussions, more diverse discussions. So we actually took all kinds of principles that we learn and we know from philosophy, from Judaism, from the Talmud, from decision-making processes and studies, and we tried to implement them within the software. It’s a discussion board, but where we implemented in the software some major principles, very simple ones, but still very important ones, to allow a more rigorous, rational, deeper discussion. And then we implemented some very unique algorithms that we actually invented and patented that will analyze the structure, the rigorous discussion, and will assign a score to each of the different views, will monitor how diverse the discussion is, will try to manage and streamline the discussion into a better, more reliable way of discussing things.
Ari Lamm [00:08:18]
So is the upshot of what you’re learning, I guess you can kind of think about the philosophical history of decision making, if you like really want to oversimplify it, it’s kind of like Aristotle versus Hume or maybe Descartes versus Hume. So are humans reasoners with some annoying emotions around the margins, or are they emotional creatures with some facade of reason at the storefront? So you can go back and forth about who’s right, who’s wrong. But what you guys are doing is essentially saying, well, whoever is right, even if you assume that Hume is right, we can help humans be more reasonable by using sort of extra-human tools. Right. So is that essentially what you guys are doing?
Joab Rosenberg [00:08:57]
Exactly. And I do think they’re both right, right? There are some sides of the human which are not so reasonable. They’re emotional, if you would like, or irrational. But I think mainly when we discuss things and we try to take decisions together, it’s more the rational side of the human being that is being exercised. By the way, also, in the modern time, you could think of two Nobel Laureates in Behavioral Psychology. So Kahneman was more on the side of, you know, presenting to us the irrational side of human beings.
Ari Lamm [00:09:30]
Joab Rosenberg [00:09:31]
And then Professor Ullman was much more about Game Theory and the rational side of thinking. So we have both of them, but I think we can still apply algorithms to try and first of all, streamline the discussion in more of a rational way, because I think at the end of the day, we’ll just drive better and learnings and better decisions, and we can try and find out the irrationalities of certain participants or subgroups of participants, all kind of biases such as groupthink. We are, I think, still the only platform that allows you to alert the manager of the discussion to a groupthink within one of the subgroups of the participants, which is very unique and I think very important.
Ari Lamm [00:10:14]
That’s fascinating. That kind of gets me to something that I was so captivated by as I was learning more about Ment.io, and that is that when I think of Jewish pedagogy, so it’s extremely dialectical, extremely dialectical, I think even more so than the Greek tradition. Right. So in the Greek tradition, obviously, it’s, you know, the earliest kind of philosophical texts. You know, we talk about pre-philosophical, the earliest philosophical texts are constructed as dialogs, but it often reads, at least on the surface, and I think even if you dig a little bit deeper, it’s sort of like very smart people exposing the idiocy of lots of other folks. Whereas in Jewish tradition, at least in the Talmudic tradition, it’s really just people arguing with each other, very oftentimes, loudly and harshly, but in good faith. So the Talmud imagines heaven as a place where scholars can argue with each other and with God. Right. It’s hell for some people, but it’s heaven for the Talmud. So in that respect, the whole concept of Ment.io felt very familiar to me as a student of Jewish wisdom. Right, you’re trying to distill a position or distill the truth from many different voices and contributions that are trying to pull you in many different directions. So it seems to me like it’s not an accident that Ment.io is an Israeli company. So what do you think about that? Is there something sort of quintessentially Israeli about Menti.io or the concept?
Joab Rosenberg [00:11:28]
Well, I agree. I think the Greeks somehow were first to understand the importance of dialog, but we have to remember what they’ve done to Socrates by the end of it. So he was killed. In the Jewish tradition, we had one was killed because he was too argumentative. That was Reish Lakish, right. But then you remember, Rebbe Yochanan was very sad about that and he regretted to the point that he himself died. So we kind of think of the people who are extremely argumentative as helping a lot to develop the knowledge. And of course, the Talmud is all about documenting disagreements for the end of time. And it’s a very interesting and it goes very deep. Why do we want to document all of these disagreements? Because at the end of the day, it’s about Halacha, right. You need to know what you do when you wake up in the morning. So why would you want to know that one rabbi thought this way and the other rabbi thought the other way and get all of the discussion, the full discussion, instead of just getting the end point of it? And I think it is really important and it’s very, very different, you know, from today. How do people take decisions today? So the most you get as a software is polling, right. And what is polling? It’s just vote. And, you know, the 20 percent thought A and 80 percent thought B. We’ll never remember why did they think A and why did they think B. And sometimes you think A and I think A. But we think A for very different reasons, right. If the decision was wrong and we tried to learn our lessons, we will not even remember that. So the fact that we allow people to actually document the rationale, the reasoning behind the decision is to me very, very important. And when we try to distill what a good discussion is about, as you said, because of the Talmud, we distilled it into not just commenting on one another as you find on regular social media, but actually separating your arguments from your counter-arguments. So on our platform, you don’t just comment. You either ask for clarification because you didn’t understand what I’m saying, or you say I agree with you because of A, or I disagree with you because of B. And then someone else can chime in and say, I disagree with Joab about this because of another reason. And then you present evidence and then you can put on data and do things to prove yourself to be right. But we are documenting the rationale and the reasoning. And in that way, it’s very, very different. And by the way, because we do that, we have a very precious dataset where we can see all kinds of psychological and cultural differences. So you will not be surprised to your question to learn that the Israeli users are much more argumentative on the average. So they’re using the disagree much more than the Americans.
Ari Lamm [00:14:13]
I am shocked.
Joab Rosenberg [00:14:15]
And Americans on average, of course, it’s all averages, are much more argumentative than the people in the Far East. So understanding that disagreeing with one another is not a personal insulting thing, but actually something that promotes knowledge and advances the understandings of human beings is of greatest importance to learning, but also to taking the right decisions.
Ari Lamm [00:14:41]
So I would have predicted that, you know, Israelis are extremely argumentative, you know, Americans less so, and people in the Far East, maybe less so than that. I guess that would have been like my default assumption. But it’s fascinating that you can actually demonstrate it. But one question that I actually don’t have a default assumption about would be, do you find differences in how quickly different cultures are to come to agreement? And are there different ways that that shakes out?
Joab Rosenberg [00:15:10]
Yeah, there’s a lot in the data and we still didn’t do all the research. We have several professors who are also very interested in this kind of data because you cannot find it on Slack or on Facebook or anywhere.
Ari Lamm [00:15:21]
I can imagine. This is amazing.
Joab Rosenberg [00:15:22]
So we could also, we can and we monitor the polarization of discussion. So whether they converge into consensus or not, because you can see that. You can see, and that’s another very interesting topic, it’s the dynamics of the discussion. How it starts, who initiated it, who came in first and who came in later, who made the people change their minds. So you can see the paradigm shift, because we assign as I said, we assign a reliability score to each of the views based on the responses that you get, the peer review, if you’d like. And you can see the changes in time. So sometimes you see, let’s say that the structure of the discussion starts with a question, then the alternative answers like you find in old school forums. And then, as I said, there’s a debate instead of just commenting. And then you can see that a certain answer, let’s say answer number two, initially got most of the support and then something happens and people change thire mind and shift towards answer number five. And you can go back and trace what made them change their mind. Is it a new participant? Is it the new evidence? Is it just a certain dynamic that changed their mind? And this is something that we monitor, and, you know, all of the research in psychology that you find about these dynamics was always done in a classroom or in a meeting where you get thirty people. You monitor and you record what’s happening along two hours of a discussion and then you go back to your room and you just play the recording and write down what you found. And we have tens of thousands of users using the platform continuously. So it’s a big data set that I think is very unique and I think by now we have the biggest ever data set of arguments and counter-arguments, any company in the world has.
Ari Lamm [00:17:10]
It’s the most Jewish story I’ve ever heard. It’s amazing. I love this. So am I correct? You’re a PhD candidate in philosophy.
Joab Rosenberg [00:17:16]
Ari Lamm [00:17:16]
Should more people in tech study philosophy? Like, what do we lose by not studying philosophy?
Joab Rosenberg [00:17:22]
First of all, I think everybody should study philosophy. Not only in tech. And everybody should study the Talmud. You know that the Koreans are teaching the Talmud in their schools. That’s quite amazing.
Ari Lamm [00:17:33]
I remember when I was at Yeshiva University, we had a delegation from South Korea come and visit the study hall, the Beit Midrash. It was unbelievable.
Joab Rosenberg [00:17:41]
And I tend to think that especially in tech, where people speak about artificial intelligence and data analytics, they must study philosophy because there are certain prejudices from the computer sciences schools that are just ridiculous if you read Hume, you know, just like 101 Hume. The thinking that, you know, we just collect more data and suddenly the answer will come out of the data automatically by blackbox algorithms is ridiculous. It will never happen. Correlation is not causation. The endless amount of very, very basic ideas in philosophy that just falsify this very basic idea that people pour money into, because the illusion that, you know, very soon you just collect a bit more of data and all of the problems of humanities will be solved automatically without the human beings, who are confused creatures. They do not understand that, it’s only the computers that can do that. And you hear that. You hear that a lot. You hear people raising money on this story. And it’s very clear why. Because you won’t need sales-people anymore, right? If computers can do everything. You don’t need medical doctors because computers will just monitor your temperature and whatever and they will know immediately what to do with you. And to my mind, it’s so obvious that this is not going to happen any time soon because statistics is just not enough, and Big Data is about statistics. So humans are there to stay. And in that sense, also, I agree completely with Peter Thiel, who said, you know, you need to use A.I. to accelerate human thinking and not replace human thinking. And in that sense, I think we devised a very important machine in the stack. The technological stack, what do you do with data? So there are a lot of companies investing in cleaning data, storing data, finding correlations in data, doing all kinds of things with the raw data, the information. But information does not amount to knowledge and cannot drive, major decisions. You know, recommending my next movie on Netflix? OK. My next book on Amazon? Who knows? But I would not want to be treated my illnesses with a machine that they don’t understand what’s happening and is judging by using all kinds of regressions and anomaly detection. It’s important. It’s very important. It can help the medical doctor. But I don’t think it will ever, well, let’s say at least in the next 20 years, it will not replace a medical doctor. And we can see what’s happening with autonomous cars, for example. People thought it will be, you know, few months and we’ll get there. And we’re not getting there because the human mind can do many, many amazing things that computers are still not able to do.
Ari Lamm [00:20:20]
That’s such a point well taken. And it actually leads me to my last question, which is, you know, I am a big believer and I’ve said this on the podcast many, many times, that the world of tech sort of has a natural ally in the world of tradition. The world of tech is sort of about what humans can do. And I think the world of tradition is about what humans are capable of. So if I’m a person in the world of tech, I’ve never encountered the Talmud before. I may have encountered the Bible, but I’ve certainly never encountered the Talmud, which by my lights is probably one of the most underrated works of intellectual accomplishment in the history of civilization. And I just want to know where should I get started? What’s one passage I should read, a tract that I should read, something I should do to get started, a book I should read. How would you guide someone to get started?
Joab Rosenberg [00:21:03]
It’s a bit difficult. You need some help. But luckily enough we have the Rabbi Steinsaltz, who actually translated the Talmud into English. And I actually bought the story of Rebbe Yochanan and Reish Lakish I brought to my investor, the partner of Peter Thiel, when they invested in me because I wanted to let them know of the story of a very argumentative tradition. And I think by now, again, on the Internet, you should be able to find quite a lot of help on that. And I’m sure on YouTube you can find good speakers speaking about it. I don’t know if the Rabbi Steinsaltz lectures are also in English or did he only speak in Hebrew. But, you know, Rabbi Sacks and others that you know, Rabbi Soloveitchik, most of it is in Yiddish, I think. But there must be something in English too.
Ari Lamm [00:21:52]
No, there’s, there’s a lot of stuff in English. Those are good recommendations. I mean, you can start on Sefaria. Those are all amazing. Unbelievable. Joab, thank you so much for joining us. This was fantastic. Thank you.
Joab Rosenberg [00:22:02]
And let’s hope that we can bring this tradition to life and to technological life. That would be even more amazing. And I really think we can accelerate not only decision making in learning, we can actually accelerate science if we apply algorithms in the right way, because people can quickly find what others are thinking, what kind of evidence they have in favor or against a certain idea. And that will be the next step. I’m sure it will be the next step of Internet, you know, putting algorithms to work. I don’t know if our algorithm will be the one to win, but we’re certainly going to do the effort to win. But I’m completely convinced that in the next few years we will see more and more people doing knowledge analytics and not only data analytics.
Ari Lamm [00:22:46]
Amen. And where can people learn more about Ment.io?
Joab Rosenberg [00:22:49]
First of all, on our website, Ment.io. And there we have all kinds of resources like webinars, introductory videos, blog posts and also on YouTube. If you just search for Ment.io you will be able to find all kinds of videos. And of course just write to us. You can write to me personally, firstname.lastname@example.org or email@example.com. And we’re always happy to speak about our vision because we find it so exciting and important in these specific days.
Ari Lamm [00:23:22]
Unbelievable. The world’s most philosophical startup. Joab, thank you so much for joining us.
Joab Rosenberg [00:23:27]
Thank you, Ari. Thanks so much.
Ari Lamm [00:23:36]
Look, you heard Joab. Humans are here to stay. But anyone who reads the Bible understand that humans have so much room for virtuous improvement. And that’s where tech becomes a wonderful partner. Because the best way for we humans to become better is for us to become closer. And that means more discussion, more conversation, even more disagreement. So how do we get better disagreement? Well, I can’t imagine a better partnership than that between the wisdom of tradition, the Talmud and Socrates, Jerusalem and Athens, and the know-how of Silicon Valley and Tel Aviv. That’s it for today’s episode. If you enjoyed it, well, give us a rating and a review on iTunes. It really helps people find the show. So that’s it for now. This is Ari Lamm making a good faith effort. I’ll see you next time.
Good Faith Effort was created and written by Ari Lamm, if you enjoyed the show, please rate and reviews on Apple podcasts or your podcast app of choice, because it really helps others find the show. Our executive producer is Josh Kross. The show is produced and edited by Paul Rueste. This is a Joshua Network podcast presented by Bnai Zion. Follow us on Twitter @gfaitheffort. Follow Ari @AriLamm and sign up for our email list at theJoshuaNetwork.com.