 It's great to be here. Thank you so much for joining us, Tim. Oh, thanks for having me. It's a pleasure to be concluding the first day of funding the Commons the first time it's happening in person. And I'm super honored to have Tim here with us. I was actually Tim's student way back. So in addition to all the great, amazing things, he also had the misfortune of having to deal with my probably bad problem sets and exams and so on. So thank you for imparting super amazing knowledge on all of us for many years. One of the things that I really appreciate about you and your work is that you really care about making sure that people learn. And you have been producing these amazing lectures online that have been teaching so many of us for many years. So thank you. Let's dive in. So I want to get a little bit of an interruption. One of the cool things about once you have some experience in your belt as a professor is you realize the payoff of all the hard work you put in on education and research, sometimes the payoff is very long time scale. That's true both for the research itself, but it also can be true for the personal connections. So I maybe see them when they're 19 and then when they're 10 years, fast forward 10 years later in the kind of running Silicon Valley or doing other amazing things. So it's a very cool part of the gig. So I kind of wanted to just start with your story. How did you get interested in Game Theory originally and computer science and so on? Sure, yeah. So I did my PhD at Cornell from 98 to 2002. And so if you think about 98, that's really when the internet was really starting to explode and get commercialized. And really all across, so I always started studying algorithms and theoretical computer science, but really all across computer science, it was obvious to lots of researchers that the discipline of computer science had to change in response to sort of this new artifact that everybody was using. Okay, it wasn't that new, it was sort of newly, sort of had newly crossed into the mainstream. A pattern, I might say I kind of expect to repeat with web three, computer sciences discipline sort of having to rethink itself in light of the technological developments. Anyways, one of the things that was new was rather than the 1980s where we were focused on like a single machine on somebody's desktop, but all of a sudden we had a bunch of machines that were connected together. And in particular we had interaction between multiple parties, possibly with conflicting objectives. And so you hear that and you're like, that sounds sort of like the, if you ever took Game Theory 101, you're kind of like, that sort of sounds like about what Game Theory is all about. So, almost all the computer scientists that started what's now called algorithm to game theory, we all had to be sort of autodidacts. Again, there's gonna be echoes that you'll hear in web three now. We basically had to teach ourselves the relevant parts of Game Theory and economics to sort of do the computer science research we wanted to do. So we were motivated by just giving computer scientists and engineers sort of good guidance, good mental models for how to think about the engineering problems that they were facing as builders. And to do that we really needed to acquire these new tools from economics. And in that time period, were there any kind of opportunities that you saw that were just greatly undervalued by either the rest of the field or computer science? I mean, there's a bridge building between the field that you were doing were there any kind of amazingly undervalued opportunities that you could recognize because you had expertise in some other area. Yeah, I mean, probably I was in, yeah, I guess I don't wanna take too much credit for any decisions. I was just sort of, I was lucky enough to get in on the ground floor of this movement. You know, probably the ones, where I saw I was like a beginning of my second year of grad school, something like that. And so I knew nothing, to be honest. But you know, my advisor who's a brilliant computer scientist named Eva Tardosh, she, there was a problem involving sort of game theoretic analysis of routing. Sort of motivated by computer networking. And she was as excited as I'd ever seen her about a problem. And I was like, I mean, I know Eva has great taste, so I'm just gonna start working on this and then kind of rode the wave from there. So a lot of it was just kind of right place, right time. But you know, there was also, I mean, it was a non-traditional topic to be working on. So I guess there was a willingness on both of our parts to sort of be early adopters to kind of take a bet in effect that this thing was gonna go somewhere. And again, I'm trying to phrase all of this so that maybe it sounds familiar to those of you that have been getting into web three in the last couple of years. And a lot of patterns were occurring, right? You need to teach yourself lots of things because the resources to learn them, you know, aren't yet. First of all, it's a lot of, I mean, it's just hard to, you know, nobody has like a completely, has a 100% perfect mental model of web three, that's for sure. And then even the people who have the best mental models either haven't had the time or the inclination and necessarily spell them out in a way that's easy for people to consume. So teaching yourself stuff is sort of just part of the game when you're very, very early, early to a field. And then the multidisciplinary aspect also, I think, I mean, I think it's even more extreme actually in web three than it was in the work I was doing back then. Because in addition to all the microeconomics and mechanism design and so on, you've got sort of traditional finance. You've got political science when you're talking about governance. So I mean, you know, my vision is sort of both for the, you know, the lab that I'm now doing at A16Z and also just for the, just from an academic perspective of the field, as this proceeds, I envision this, you know, really sucking in lots of different disciplines, so. I'm personally super excited that you're bringing academic rigor from game theory into the field because the industry right now is just severely lacking the level of quality of analysis that we're used to in academia. And so I think it's really great. The, you may not recall this, but there's a fun anecdote where kind of like in 2011 or 12, I tried to get you into Bitcoin and blockchain. And I was like, Tim, this is like amazing mechanisms like in the sky, this is like super cool. And they were like, I don't know, this sounds kind of weird. Like the rest of the world, to be fairly, this is like very, very early, like there were a ton of other cryptocurrencies of the time and, you know, I had also dismissed Bitcoin even years before. What changed your mind? I think a lot of us have to go through that jump of, hey, actually, this is not that crazy. And this, oh wait, this is actually super interesting. Was there kind of like an aha moment for you or was it more like just gradual? Yeah, it's a good question. And many people will like tell you about the weekend they got crypto-pilled or went out in the rabbit hole. Everybody loves to say. And for me, it was actually a much more gradual journey, which I don't know, for me at least has been kind of healthy. So I don't remember which, I was certainly aware of Bitcoin by probably 2011. Maybe it's because you told me, totally possible. The first research paper I remember reading about Bitcoin was 2012. That was sort of an early paper in economics and computation, not by myself, by other researchers. And then, I guess what first, as with so many things, I first started incorporating it into my teaching. So because I was just, my calling card has always been sort of the connections between computer science and economics, and I'm in the computer science department, and I teach primarily computer science students. So my mindset is, let me teach computer scientists just enough economics and game theory to sort of make good decisions, sort of going forward. And over time, I evolved toward a more and more almost like case study-based approach to that. Where I talk about, it could be anything, it could be like ad auctions and sort of search engines or any number of other things, just examples of where game theory and economics really informed the platforms that the students are using every day anyways. So the Bitcoin White Paper, so first of all, whenever I did get around to reading it, I think like many people just struck me as kind of a work of art that just kind of dropped from the sky out of nowhere. So that's just kind of very, it's just a beautiful protocol, frankly, and I still read that White Paper probably once a year or something like that. And it was obvious there was like, game theory to guarantees or kind of what were being offered, Nakamoto is very explicit about that. But it wasn't clear to me, it's kind of like, okay, well, what's the teachable moment here? Like this all seems, I don't know, it's just like, but then another big moment was when selfish mining was invented. So this was, it's probably like the most well-cited academic research paper, actually in Web 3 by Sarir and Ayal. And they showed that actually there's really interesting game theory behind Bitcoin. So they kind of blew up the convention that the game theory was straightforward in Bitcoin, it showed it was actually much more subtle, much more nuanced. So that came out, I really liked it, and I was like, okay, this I can teach, this can be a cool case study, like for my courses, right? And, but I still wasn't like sure, it wasn't clear to me like, what would be a PhD thesis I would advise on this topic. So I was sort of starting to watch it from a distance. Out of curiosity. And then I think the next step was I just started getting, like a lot of people, I just started getting these cold calls in 2017 by people saying, we really, really need a computer scientist who deeply understands mechanism design. And the first call I got, I was like, what did you say? I was kind of like, are you pulling my leg? It's just like, yeah, I mean, it's just, it's not, you know, the usual dynamic as a professor doing research, even if you're working on like really, you know, sort of important problems is you kind of are in push mode. Like you're trying to like, you come up with what you think are great ideas and you try to convince people they're gonna be better off if they adopt your ideas, right? And all of a sudden, like I went through this whole time, the whole last five years, it's just been much more like pole mode. It's just like, people already believe that they really need expertise by people like me and they just kind of, you know, want help. Like they just have hard problems and they want help. And so it's exactly like when a student walks into my office and kind of wants advice about something, it felt exactly the same. So 2017, I was like, okay, this doesn't happen that much in a career, especially if you're on the sort of theoretical side, like I am. So I was like, I guess I gotta, you know, so I started just doing some like startup advising and like some of the startups were, you know, not good ideas, other ones were somewhat better ideas. But I started getting a feel for the ecosystem then. It's also probably when I started rocking the kind of world computer vision of Ethereum, which frankly is right there in their white paper, but it took me a couple of years to actually really get it. So I like to talk about the sort of virtual computer that sort of lives in the sky and sort of runs as a public good. And so, you know, as I got more embedded in the ecosystem, I started to understand that better. And then I started really teaching semester long classes just about Web 3 in 2019. And so that then I started really having to learn a lot of it. And, you know, the deeper I got, right? So like, you know, so it's a good sign when like the more you understand, the more excited you get. Like there's a lot of topics which maybe sound really cool, like superficially, you know, and then you spend like an hour or a day like thinking hard about it. And you're like, eh, it's actually just kind of like X, you know, and the more time I spent on this, the more obvious it became it was unlike anything else I'd ever seen. And at that point it's like, okay, this could be really fun. Were there some specific insights in there that either some particular mechanism, there's just some ideas that just kind of captivated you or that helped it? Yeah, I mean, I think probably more than, I mean, like I said, I mean, like, you know, Bitcoin was captivating in its own way, but was just kind of, you know, one thing. I guess what just was startling was, it was just, you know, it felt like all the computer science classes I took, like, you know, as a grad student and stuff, all of a sudden were newly relevant. Also the classes I'd studiously avoided because I didn't want to learn it, I found myself now having to teach myself. So like I never studied any distributed computing when I was in school. I never studied any cryptography frankly when I was in school either. And so I just never seen that before, right? So I mean, just like, you know, you have sort of the consensus stuff which historically has been like its own little community focused on kind of permission consensus protocols. You had kind of all the zero knowledge stuff kind of coming from cryptography. You had like succinct proofs which really, you know, their heritage is really from complexity theory. So that's a really important part. But then also I found myself, you know, teaching myself some traditional finance, just so I could sort of interpret sort of the developments in DeFi. And then there were a couple of things I already knew that were relevant happily like algorithms and economics. But so I think that, I mean, it just was assembling itself in this mosaic I'd just never seen before. And so it's just kind of like, okay, I mean, it's not once career is only so long and there's no way I'm gonna like miss this boat basically. So as you were sort of exploring all of this and teaching some of it, is that sort of when the EIP 1559 call to action came about or how? So let's see, so I was aware of the proposal already probably in 2019, just through the, just like I said, being part of the, a bigger part of the ecosystem doing some startup consulting. The Ethereum community reached out to me in probably spring 2020, I would say. And, you know, I think maybe some of them knew me through my algorithmic game theory lectures, for example, on the YouTube playlist. And yeah, so they were really, you know, so as you may know, right? So this was a, so Ethereum basically made a massive change to their transaction fee mechanism, which was deployed on mainnet about 10 months ago or so. But it was proposed, so that was 2021. It was proposed all the way back in 2018 by Vitalik Buterin. He started sort of circulating the proposal several years earlier. And it was very polarizing and for a few reasons. But the point is that, you know, some people in the Ethereum community really love the idea and some people really hated the idea. And the Ethereum community is very, very impressive as far as just how, you know, as far as their decision-making and their sort of passion and sort of coordination. And so credit to them. So they really said, okay, let's, let's try to make a good decision. Like what do we need to make a good decision? They went out and asked lots of community stakeholders. You know, what do you think, what, you know, what extra inputs do we need to make a good decision about whether we should make the switch or not? And one of the most frequently things mentioned was a sort of more formal description and analysis of the new mechanism. And so, you know, part, you know, just sort of in response to that. So that's why they reached out to me. So they're like, okay, who should we contact for to do a formal analysis? And I'm very, very, very grateful that they contacted me. I already knew about the mechanism. I knew a little bit about it. You know, obviously I've done mechanisms in my whole career. By that point, I was getting pretty in the blockchains. So it was just kind of like, again, like just perfect timing and just very, you know, just feel very lucky that I got that opportunity. It was really good that you did the analysis when you did because so we were developing Falkland and we were in like the Testament era and so Falkland ended up having to be in L1 as well. And we were running into problems with the traditional gas model week because we followed a lot of the same kind of structures. And so then we just moved into EIP 1559 and it worked dramatically better because we were already congested at that point. And so it just worked very smoothly but the analysis hadn't dropped yet. So we're like, ooh, are we about to like launch the mainnet without like very strong grounding. And I think I don't remember the exact timing. I think we might have launched before or maybe it came out before it. But anyway, we were suddenly one of the largest networks deployed for the EIP 1559. The analysis dropped, we got a massive confidence boost. And we're like, oh, this is great. Like amazing ecosystem bridging. We're like, Ethereum reached out to you, got it all figured out. Then we got to benefit from the drop. So it was really good. And yeah, it's like, now there's a bunch of super interesting gas theory changes when you start thinking about like sharpening and all of the different layers. We're just getting started with the research, that's for sure. Yeah, actually, so on that, like you, so you also share the view that we're super early still in this space. Yeah, I mean, so right. So people talk about, you know, it's still early and usually they mean either from an investment perspective or just an adoption perspective. And it's not that I don't agree with that. I do agree with that in both those dimensions but like from a science perspective, my God, are we early? Right. Which is normal, right? If you think about it like, you know, I mean, people were trying to build compilers for early programming languages, like by the early 1950s. And we didn't have like the dragon book. We didn't have like a good theory of compilers till a good, I don't know, 10, 15 years later. So it's a very normal pattern where, you know, the, so I mean, it goes, but sometimes you have the theory pushing the frontiers, right? So like zero knowledge proofs would be a great example where the theory came first. But it's also very common that people build stuff and they have, you know, inevitably they're like super smart people, like encoded in their brain is like all of this accurate intuition about like why they did what they did. But then the question is like, how do you articulate that? And, you know, sort of make it modular, break it down so that that knowledge can be sort of passed on to the next generations. That takes time, so. So right now we have this model where we're just shipping a lot of things with some amount of analysis, probably not nearly as much as we should have. But it's also really healthy for the environment because we get to progress really quickly and explore a lot of things. There's cryptography now being deployed that, you know, 10 years ago nobody would have dared touch with a 10 foot pole, right? Like we're shipping in a crypto world, think, you know, zero knowledge tools that were invented like the same year, right? And so like this is unheard of before it would take like five to 10 years for some cryptography to be really trusted. So now there's of course like a danger onto the other side of it, which is like we're now deploying and relying on like potentially mushy foundations that may turn out to be broken. How do you sort of like thread the needle? Would you kind of like orient more towards having slowing down, doing, applying more rigor or still ship but maybe bound the value flows and then apply the rigor or how? Yeah, I mean, it sort of depends on the risk profile of the project I'd say. So I don't have any universal advice around that. You know, I feel like my job is more to kind of say here are the risks if you do this. Here would be the reduced risk you would enjoy if you spend an extra six months doing like detailed simulations or an extra year sort of, you know, getting sort of more mathematically inclined people to do a formal analysis. That's kind of their call on it. You know, it's often gonna be like a business call, right? So like it's gonna depend on like how urgent is the need to go to market? It's gonna depend on like what's the worst case sort of, you know, lost from an attack. So yeah, so I feel like all I can do is give people guidance about the various options and ultimately it's kind of on the project to decide. So let's talk about ACC Crypto Research like super exciting lab. We were super stoked to hear the announcement. Yeah, what is your vision for it? Yeah, so, you know, I really probably already become clear from what I've been saying so far. I mean, I really believe that we're kind of witnessing a new area of computer science like just kind of materialized, like before our eyes kind of in real time. And so, you know, I'm fundamentally an academic, my passions are sort of research and education. And so I really, you know, my focus is always on sort of shaping, you know, kind of new fields, right? Both in the sense of trying to do fundamental work, but also just in the sense of, you know, on the education side just giving people what I think are the right mental models for sort of to think about it. And historically I've done that in the ivory tower, like with our algorithmic game theory, all the work that I did there was as a professor. But my approach sort of is, you know, my vision for the lab is sort of similar here. Like so I would love 20 years from now people to go back and say like a turning point in the development of sort of, you know, web three, whatever it's gonna be called 20 years from now. A turning point in its intellectual development, you know, was the A16Z Crypto Research Lab, right? So, you know, to give a, yeah. I mean, so, cause you can, there's other parts of computer science where you can point to, you know, like, you know, when Von Neumann first, you know, did the ENIAC in the late 40s, no one thought there needed to be like a field of computer science, right? It was just like a machine, right? We don't have a field of toaster science, right? So like what's so special about a computer as a machine that we need a whole scientific field around it? You know, now we understand that there's like a universality to computers, right? And so, turn completeness if you like, right? So in some sense, computers capture just, you know, in the same spirit of like physics or something, they capture something just very fundamental about the universe we live in. So, plus obviously the, just the sort of impact it has on all of our day-to-day lives is obviously just, you know, tremendous. So no one doubts that it should be an intellectual field now. But honestly, like when I was an undergrad in the mid 90s, it was a fairly recent development. Like I think Stanford had had an undergrad CS program for at most 10 years at that point, right? It was not necessarily regarded as, you know, a topic sufficiently, I don't know, you know, whatever for sort of undergrads to study at that point. So that would be the best case scenario, right? That that's people can point to that. Now obviously, you know, just on the short term, right? The big mission statement of the research lab is just to help kind of A16Z's portfolio sort of succeed. Now, what's cool about it is their portfolio is so big it has such kind of thorough coverage of the whole Web 3 space. I mean, I, you know, basically, as far as, you know, I think A16Z's objective function in Web 3 is basically to have the space succeed. Like they're gonna do well if the space succeeds, they're not gonna do well if the space doesn't succeed just because of their kind of sort of thorough coverage of it. So that's kind of the, so, you know, often we will talk specifically to our portfolio companies about their specific problems, but, you know, one thing that's fun about having kind of a team of academically minded people is, you know, you see these, you know, on the one hand, in the short term, you help companies to start us with their specific problems. On the other hand, you start seeing the patterns and you start sort of zooming out a little bit and you start asking, like, what are the fundamental challenges, the fundamental barriers that seem to be holding up all of these different projects? Okay, can we have a model for what that is? Can we talk about impossibility results, possibility results, trade-offs between the solutions? So it kind of, it's, you know, it's, you know, things evolve simultaneously, I think, top-down and bottom-up. So bottom-up meaning you just take concrete problems faced by concrete people building things and then top-down looking for kind of what, hopefully, will be the, you know, the dragon book for Web3, so. Do you plan to, so beyond, I mean, definitely the S&Z portfolio is super expensive. It's extensive, so, probably expensive, probably expensive too. But probably less expensive now. So, good, good, yeah. Now we're used to it, like, you know, we've been through, like, what, I've been through four roller coaster up and downs. It's like, you get used to the volatility. The, and actually, you get good at planning, so good advice for anyone out there building a company. Get good at planning your runway. In these moments, you don't have to be, you don't want to be scrambling to figure out what to do. So, the, so as you think about maybe collaborations with groups in the ecosystem, S&Z portfolio, and beyond, you sort of envision maybe, like, larger scale projects, like, either, whether it's, you know, things like EIP559, but maybe, like, designing some system like that, or designing next generation systems, and then suddenly you're like, hey, by the way, world, here's a whole new structure. So beyond kind of like analyzing and tweaking algorithms, and then suddenly realizing something fundamental, and then like, or maybe working with some groups to develop something like that, and maybe spending it out of the lab, or? Absolutely, yeah. So, and indeed, another big part of the mission statement of the research team is just, so obviously the, you know, the academic papers that we produce, textbooks that we write, you know, those are in a form of public good, but also just more on the engineering side. And in fact, we're gonna be collaborating closely with Eddie Lazarin's team, who runs the engineering and protocol design team. Actually, we were just talking this morning about some super cool projects around sort of, you know, privacy in auctions, your knowledge in auctions, to an auction design formats, which we're gonna be working on. So absolutely, the plan is to, you know, just in a, you know, resources for the space to succeed, you know, in every imaginable dimension, right? So it's not just the academic research papers, it's not just the textbooks, you know, it's also, you know, open source, you know, GitHub repos, you know, stuff that people can use as starter code for their own projects of various forms. We have a research seminar series that we're running. We're having something like three to four seminars by a top researchers all over the world coming through the lab, recording all of those. Those will start getting rolled down in a YouTube channel, probably, you know, next month or August at the latest. That'll obviously be, you know, free for everybody. That's again meant to be, you know, kind of, you know, just more resources out there for people to just stay on the cutting edge, so. So we're running close on time, so let's switch to the last thing, which is the public education stuff. So I think we're both very strong proponents in knowledge diffusion and learning, and online learning specifically. I tend to see learning as a great public good that we can, the more we can do there and then we can benefit the world. So how did you get started doing the lectures online and so on? Yeah, so it started, so there's this algorithms MOOC that's been running ever since 2011, I think it launched. Again, it's just kind of, you know, right place at the right time and then taking advantage of that fact. So I was at Stanford as a professor from 2004 to 2018 and 2011, my colleagues, Andrew Ng and Daphne Kohler were found in Coursera. And I'd already actually worked some with Andrew around just sort of videotaping my Blackboard lectures at Stanford, so had, you know, had that connection already. They, you know, needed their first batch of courses, basically, right, and I volunteered to do algorithms. I taught it, I don't know, seven, eight years in a row. Had it down, I was pretty proud of the algorithms course that I'd sort of eventually kind of converged upon and, yeah, it was just one of those things, it was again, it was just like, you know, I don't, this might be a moment in time, right? So it felt like that in 2011 when Coursera was first coming out the same way, like I hope to all of you now feels like a bit of a moment in time, kind of in the Web 3 space. And then that just, honestly, that just took on a life of its own, right? So definitely at events like these, it's very common for people to come tell me I helped them, you know, pass their algorithms class. You know, I get emails, you know, all the time, so I'm just like, you know what? I was stuck in Bangladesh and I learned algorithms and programming from your course and now I've moved to Bangalore and I'm just like so much happier, you know? I get emails like that all the time, which is really, I don't know, I mean, it's obviously this tremendous gratification and like, you know, just doing the on-campus course, you know, and having people really appreciate that. But I mean, it does feel great to feel like maybe beyond just, you know, I mean, obviously there's lots of people who don't have the opportunity to take a class in person from me at a Stanford or Columbia or something like that. But I think we all know there's sort of just utterly brilliant people all over the world. So just anything that gets more of them into the fold, like just gives more of them a chance to kind of succeed. I mean, that'll get you up in the morning, that's for sure. So for everyone out there, it's Tim Ruffner and his lectures on YouTube and there's algorithms, algorithm and game theory, mechanism design, incentives and computer science, and now foundations and blockchains, which I think is still going, right? Yeah, still working on it, yeah. Probably gonna be a lot of foundations for like a few years. A while, yeah, yeah, yeah, exactly. Yeah, so for an entry point, I would suggest incentives and computer science. That's meant to be easy listening, as it were. Are you in touch with any other YouTube educators, or are you just sort of like doing, of course like all the Coursera peers and so on, but have you seen maybe like some of the changes and just approaches to explanations? There's for example, an amazing YouTube channel, Three Blue One Brown, I'm actually wearing a Three Blue One Brown shirt. It's a great public good. There's amazingly beautiful visualizations of math to help people grok concepts that otherwise would take you a while to really get. I don't know if you've, have you explored any of that kind of stuff? Yeah, I mean, there's definitely other instructors who I admire. Sometimes it's just they're at the blackboard and they're brilliant at the blackboard. And then like you, I have seen examples where I'm saying calculus or something, all of a sudden like you get something in five minutes that normally it would be very hard to get otherwise. I agree. And you know, if any of you have watched my videos, it's like the opposite of production of that, right? It's literally like me and my chicken scratch hand, right? And like filling up the table, right? Which on the one hand, I think some people find like appreciate the homespun, unproduced aspect, but on the other hand, yeah. I mean, it is tying my hands a bit with the visual aids. So I, you know, but I... Maybe we can get a grant. That was just, I was literally just gonna say that. Actually, because you know, I do all, you know, I just do this for free in my spare time or whatever, and it's hard enough just to get out the chicken scratch. But I think the... He was also your student by the way, I think. Is that right? I think he took 161. So yeah. Everybody took 161, that's crazy. That was it, yeah, great. Anyway, so I think it could be, you know, very high return on investment. You know, from the public good perspective of having, if you get the right instructor, you get the right animator, you know, I think it can be very powerful. Yeah. Well, any thoughts or an advice to aspiring blockchain engineers and mechanism designers out there, whether from the academic field, trying to get into practice or from the Web3 crypto space with lower or grounding in terms of the foundations, but you know, that wants to get more theoretical? Sure, maybe, I mean, the first thing that comes to mind is maybe more comment for people coming from the academia side. You know, probably for people who are, you know, in this audience, this is just preaching to the converted, but you know, to tie a couple of those threads together, I mean, when I did those MOOCs, and so for the first several years, I was like super engaged. It's like on the discussion forums. And like, when the first version came out, like there was almost no competition for online courses. So there were like hundreds of posts to the discussion forums every day to the point I couldn't really keep up with them. So I interacted with all kinds of different people all over the world at all different kind of life stages taking that class and I just think I got a much better understanding for what was valuable to them. And I also, you know, and something that I've, you know, something I learned over time is just, where I teach kind of mathematical stuff and increasingly I just realize that, you know, the proofs that we fill up a lot of these classes with, you know, there's some tiny percentage of the population we're training to be able to produce those proofs themselves, but the primary objective for 99.9% of the audience is to basically rewire their brain a little bit, like just sharpen their understanding and intuition for how to solve various problems. And so I think that's, you know, I think that's the right kind of attitude if you're coming from the ivory tower trying to engage. It's like, look, you know, like, so what did they can't do a proof by induction off the top of their head? Like, so what? Like they're building something cool, like meet in the middle, like understand what they're trying to do and then just think about how what the stuff you find obvious, right? But it's maybe very valuable to them just, you know, start getting that transfer of communication. And so I guess on both sides what I would just say is like what's amazing right now, and I'm not sure this will be true still 10 years from now, is how much like low lying fruit like win-wins between the builders side and the sort of researcher side there are right now, where on the one hand like what people are building is something the researchers have never seen before. So they find it fascinating. On the other hand, the researchers sort of, you know, have skills, you know, just topics that they eat for breakfast that are tremendously valuable back for the projects that's being built. So I would just say, you know, I would definitely strongly support Engage, Engage, Engage, you know, and yeah, it's just, it's, I mean, the final comment I just say, I really do, you know, I said it before, but I'll say it again. I do think it is a, I think it's a moment in time. I think, you know, all of us in this room will inspire jealousy in people 10 years younger than us when we say that we were here at this moment. So I think we're in a golden age and we should enjoy it. Yeah. Well, thank you so much for being with us and thank you so much for all of the knowledge that you've given us over the decades now. So, thanks for having me. Thanks for having me. Thank you so much. Run and applaud. Thank you.