unstoppable podcast, episode 62, matt stephenson, planck
NF Ts

Launching the World’s First Science NFT with Matt Stephenson from Planck

Jul 07, 2021

この記事をシェア:

Diana Chen:  Hey, everybody.  Welcome back to The Unstoppable Podcast.  I'm your host, Diana Chen.  And I'm here today with our guest, Matt Stephenson.  Matt is an OG in the NFT space.  He is working on his PhD in Behavioral Economics at Columbia and he just sold his very first—not just his, but the very first science NFT ever through his company Planck for $24,000.  There's so much more I can say about Matt, but that's just a very brief introduction.  I'll bring him on and let him tell you a little bit about himself.  So welcome, Matt.  Thank you so much for being here.  

Matt Stephenson:  Thank you so much, Diana.  It's an honor to be here.  I appreciate it.

Host:  Of course.  So before we dive into all of the things that you've done, why don't you take us all the way back to when you first heard about crypto?  What was it about crypto that caught your attention initially?  And how did you start learning about it?

Matt Stephenson:  Yes, it starts really early for me actually.  It was some comments on a blog that introduced me to Bitcoin back when Bitcoin was a dollar.  And I, I bought it at a dollar and I sold it for $30.  And I patted myself on the back so hard, you know, about how good they're obviously. Now, I'm kicking myself, but I got a Master's degree partially funded out of my—out of my Bitcoin proceeds.  And I did a little—I did a thesis on Bitcoin and Econometrics thesis, which I would be terrified to revisit now and see what it actually says.  

I'm sure I like spelt Bitcoin with two words or somethings, but, but, yeah, what drew me into this space was just—I, I mean, I think I was—I, I was interested in like this new tech, like, the decentralization ethic has always appealed to me a little bit.  And so that caught my eye, but what really grabbed me, I mean, I was just kind of like a fellow traveler until CryptoKitties, and the NFT thing is what really grabbed me and hooked me and demanded that I get involved.

Host:  Okay.  So CryptoKitties is when you first got into NFTs?

Matt Stephenson:  I wasn't cool enough for cryptoPunks.  Few, few were—few were, but yeah, it was --

Host:  -- Few were?

Matt Stephenson:  -- it was December 2017, yeah.

Host:  So when you first heard about CryptoKitties and you started learning about NFTs, this was still very early days in the big scheme of things.  How did you go about wrapping your mind around it?  Because even in the last six months as NFTs have really blown up and gone into the mainstream, we've seen, you know, like so much rhetoric around NFTs and people are trying to understand what NFTs are, why is there a value behind basically a GIF or an image that I can download from my computer?  A lot of this questions.  And so you know, this was years ago that you got exposed to NFTs.  How did you go about wrapping your mind around?  Like, what is this concept?  And why is this valuable?

Matt Stephenson:  Yeah, I mean, I think there were two things that were probably on my mind at first.  I mean, one thing, there's the sense that like, we're so used to uniqueness in the regular world, that something will trace this sort of unique history so that if somebody gives you something, you have that specific thing, and that doesn't exist in the digital space.  CryptoKitties is the thing that sort of triggered that for me, you started to think like, okay, well, if somebody’s gifted that, this will always be the item they gifted.  And you can't really say that about anything else digital.  

But to sort of prove it to myself the analogy that really ended up sticking which I think everybody thinks is kind of weird, but this is the analogy that works for me so I might as well tell you is the Rules of Basketball, which I think sold for, I think $8million or something like that.  And the reason that analogy was so interesting to me is, this guy wrote it in a gym in 1908, or 1890 or something.  

His name's James Naismith, but he wrote it on a typewriter.  And the typewritten pages, which are just the rules of basketball, they got posted.  That is what sold for $2million or $3million.  And so this his idea that peopled would value, this unique object, it was actually just a digitally intermediated object, not digitally because it's a typewriter, but you know what I mean, like, it's not like his handwriting, it's not like his beautiful sketch, it's just—it could have been anybody who typed it.  It's just that people knew it was him.  And so when I found that example, that's what really kind of grabbed me and especially plugged me more into the like memorabilia of knowledge piece.

Host:  Got you.  That makes a lot of sense.  And I'm also curious because especially when I talk to people who you know, do have a lot of writing out there, do a lot of research and do a lot of thinking in this space. You're one of my go-to sources to learn about the space, but for you like, who are your go-to sources for learning in this space?  

Matt Stephenson:  Oh, well, thank you.  Yeah, I mean, Matt G Condon is somebody who has—he's one of the many Matts on Twitter.  He's a friend of mine and he, he taught me a lot about NFTs from the beginning.  But I would say in terms of art, especially like he has really good taste and pretty most of the stuff.  If you asked me what I really liked in the NFT space, I probably saw it on his Twitter.  Like that's, that's a pretty good, pretty good chance.  

I will say at the time, like in 2018, everything was very medium focused.  That's how you learned but it now—I mean, Twitter is kind of the go-to in and, and, and who you follow but, but nothing really good, obviously.  And I will say, most of the time to my great shame, aside from the like very mainstream sources and a couple of skeptics who I follow because they're skeptical, I've mostly kind of just get channeled through like people linking me stuff on Slack and seeing something good in Twitter and whatnot.    

Host:  That's totally fair.  That's pretty much how I get most of my information too.  So I, I think that's just the, the age that we live in today.  And so you know obviously, NFTs are sort of your thing.  You're one of the OGs in the NFT space.  Before we dive deeper into that, maybe let's just take a step back and explain what NFTs are to somebody who's listening and maybe still isn't entirely sure of what it is.  How would you explain NFTs to a newbie?

Matt Stephenson:  Yeah, well, so I really like this analogy, with dollars.  So everybody knows what it's like to go into a bar or a small business or something like that, and see that somebody has posted a dollar bill and framed it on the wall.  And it's the first dollar bill that they ever sold, right.  Like the first dollar bill—dollar that the business ever made.  And now you imagine calling a bank and saying, if it was a credit card transaction, that was your first transaction.  Like, Hey, that first dollar out of that transaction was really important for you to it set aside.  

In the digital space, there's no notion of what that dollar would be.  Like, even if you tried, if you got a banker to help you out and really want to find this dollar for you, it just doesn't exist.  It's an account in a database.  And so NFTs bring this notion of what a single dollar could be a unique object that we might like to put all of our sort of human values around.  This was the first dollar that my business got or this was—this something that my, my grandmother gave me or my, my mom gave me in the digital space.  And so that concept of digital uniqueness and digital scarcity is the right way to think about it.  In the same way that information transfer is the right way to think about the internet.  

And you can go deeper in terms of exploring.  Like what is the tech?  And what does it look like on chain?  But I really think the best way to talk about it in the same way that when you say what is the internet?  Somebody says well, the internet has to do with the ability to send information.  What are NFTs?  NFTs are about this scarcity and uniqueness with digital objects that allow you to sort of interact with them the same way we're used to interacting with real objects in the real world.

Host:  So it's interesting that you use the dollar bill as your explanation for what an NFT is because people often use fiat currency as an example of fungibility and then artwork as an example of non-fungibility.  So and this is something that you and I have talked about before, and that you wrote about in your Theory of Fungibility Paper is you know, like when you think about fungibility and non-fungibility, which you know, NFT non-fungible token, is it really like a black and white, like this is either fungible or non-fungible?  Or, or is there a spectrum?  And then are things inherently fungible or non-fungible or somewhere part of that spectrum or, or is it a value that humans prescribe onto that thing?

Matt Stephenson:  Yeah, those are—those are both really deep questions.  I'll try to do them justice.  So it's definitely a spectrum, right.  Like if you try to think of something that's maximally fungible, maximally interchangeable, that doesn't really make sense.  And when you try to think of something that's maximally non-fungible, maximally unique, that doesn't really make sense.  I mean, maybe you could imagine the universe itself is maximally non-fungible or something like there's some philosophy, but there's definitely a spectrum.

  There are things that tend to be treated as more unique, right.  Like, a painting or something like that and there are things that tend to be treated as more naturally, kind of like interchangeable.  So to your point dollars, right?  Like, I like that dollar example so much because it calls to mind the fact that like a dollar bill is almost always fungible to you.  When you're in a gas station buying a Gatorade, it's the same dollar bill, you don't care, right?  It just has the potential to be non-fungible and I think that's probably closer to the way we should think about non-fungibility in the digital space as well as something that doesn't confer value necessarily by itself.  It just allows the potential for value.  This sort of like scarcity, right?  

Similarly, I mean, one, one thing I like to go back to is, when I first learned about NFTs, I wanted to buy the Pizza Bitcoin.  I was like, oh, you know, I think there were probably 10,000 of them or after many there were.  I wanted to buy one of the Pizza Bitcoins and so I looked really hard to figure it out.  And because of the way Bitcoin itself is architected, you can't really do that.  There's no notion of it in the digital space either.  And so yeah, creating something where you could sort of hang that Pizza Bitcoin or that first Bitcoin in the digital space is the—is the same kind of thing.  Did I miss a part of that question?  I think I might have.

Host:  No, I think that's good.  But something else that I, I would like you to touch upon is your water analogy that you wrote in one of your papers as well.  I think that's, that's really interesting too.

Matt Stephenson:  Oh, yeah.  Well, thank you.  So I, I wrote this piece that was really kind of a response to some engineer friends I have.  Like, web three engineer friends I have, who are skeptical about NFTs.  And I think it's hard if you're an engineer because you're—you can be too close to the code.  If somebody tells you that they really value this T-shirt they got from a concert they went to with their dads or something, right?  Somebody who's way too close to that T-shirt and can only see it as like a collection of atoms that are interchangeable with other atoms, it would feel like what are you talking about?  Right?  Like, these atoms are the same atoms, don't you understand?  And of course, that's true right?  

Like, at some level, there's—these things are kind of like fungible as far as we can tell.  And I think engineers in the digital space felt the same about NFTs because it's like, I can just see it.  It's just a string of letters or a string of hex.  So the water analogy was supposed to just sort of provide a context for remembering how weird our human conceptions of like, a thing is.  

There was this brilliant cognitive psychologist in BC Malt, who wrote about—who asked people what they thought water was, and then—and also to name some things that weren't water.  And so what's a good example of something that isn't water?  Tea, but it's kind of water-like.  And what's another example lemonade, right?  Like those are all not water, but then she'd say, how much H2O do they have?  They'd say like, I don't know, probably like 99% which is true.  And then what's an example of something that's, that is water?  Well, they'd say like swamp water, river water, right?  Dishwater.  Well, how much H2O is in that?  They say like I don't know, 70%.

  And so you think the answer to what is water is going to be H2O.  And the same way that what is in NFT is going to be this string of hex that sits on a blockchain, right.  And when you're in that mode, it can seem weird that people have these weird distinctions such that if I dip a tea bag suddenly into this water, it becomes tea even though it's changed that much.  

In the same way that if an artist says this string of hex is the art or part of the art and separately, it can just become that.  It was—it was—the point was effectively that that's not that weird.  We're used to that in the real world all the time, right?  Like, I can dip tea—I can dip a tea bag in the—in the Hudson River and it doesn't become tea.  Lipton can pollute the Hudson River such that the tea content of the Hudson is the same as the tea in my cup.  It's still not tea, right?  And so this notion that the intent really matters a lot in terms of how we class something as water or not water, art, non-art, original scientific paper, a non-original authentic scientific paper, maybe.

Host:  Yeah, and so is it your position to almost say maybe we understand NFTs inherently more so than we think we do and we're sort of just like, overthinking it or fighting that or just trying to argue with ourselves on this, when inherently we do understand the, the value of NFTs?

Matt Stephenson:  I think absolutely.  I mean, I think there's a lot—there's—it's interesting to talk about it and it's interesting to experiment with.  And I love the artists who are pushing the boundaries of like, what, what NFT art could be.  And I, I applaud all that stuff, but at the same time, I do think pretty much everybody gets it when they have an NFT in their wallet and you have them send it to somebody else or send it to the burn address.  My experience is, they understand it pretty innately, then it starts to feel like a thing.  

Like oh, I can send this to the burn address and nobody will ever own it again.  That feels a lot like a thing you destroyed, right?  I can transfer this to you, and I'll never own it again, and I won't have the rights.  Like it just feels thing—like a minimum viable thingness that, that is pretty natural and intuitive when people experience NFTs and don't overthink them.  So yeah, I wouldn't be inclined to agree that your experience is probably the is probably the best guide.  

And I think, in, in Economics anyway, we are used to thinking that, well, if somebody paid $68 million for something, they're probably valuing it and given the amount of money that's been paid for NFTs, I mean, at first pass, you just go, there's something people like about it, right?  They, they didn't pay for it as a JPEG, they paid for it as an NFT.  There's something to this and I find that just fascinating that you make a world computer with—that has these characteristics that Ethereum has and the objects on it start to get treated kind of like real world objects.  I mean, I find that fascinating and amazing and really cool.

Host:  A 100 percent, yeah.  I'm totally with you on that.  And so having been in the space in the last few years or it's probably been three, three or four years now since CryptoKitties came out, how have you seen NFTs evolve?  And I think this will be a good transition into talking about the science NFTs that you're working on now.

Matt Stephenson:  I mean, so I got in right as—right as in some sense E or NFTs were defined, which was the ERC721 specification.  I think that actually happened a little after I played with CryptoKitties, if I remember the timing right.  I have to look again.  And so a lot of the evolution you know, there's a—there's, there's standards that have changed it a little bit but 721 that initial standard has remained pretty constant.  I think there has been a lot more thinking about ways to kind of like marry the-what the NFT is supposed to represent, in some sense to the on-chain rule.  

I'm a big fan of so I think Zora has done some really interesting work on this.  And one of the things that I think is, is now being pretty regularly considered and thought about is both where you keep the sort of visual data, which is not going to be on a blockchain because it's too expensive.  Like making sure that's on IPFS or are we for something that's kind of decentralized.  So the, the—so the image stays there that's, that's been a newer innovation and then maybe trying to make the on-chain representation, a hash or a signed hash of the contents.  I think enforcing that as more of a standard and making that more normative is a thing.  Those two things are really positive directions, but also obviously there's just been this tremendous explosion in creative work around it.  I mean, we could talk about all the sorts of cool generative art projects and things like that they are, I mean, that so all amazing. 

Host:  Yeah, and then another newer application of it is Science NFTs and that's what you're working on with Planck.  And so tell us a little bit more about Planck?  What it is you've been working out for a couple of years, but it's really just blown up in the last few months or, or maybe I just heard about --

Matt Stephenson:  -- Yeah, no, no.

Host:   -- it in the last few months and I think—okay.

Matt Stephenson:  No, no, you, you—it's blown up in the last few months.  So it started in 2018.  I mean, I started working on Planck effectively after I was in my PhD working on innovation and incentives for innovation, and then learning about CryptoKitties and thinking about this weird kind of memorabilia way that scientists used to have to fund their work.  Like Galileo was trying to convince people that it was interesting to study a frictionless plane or like the motion of a planet.  And people were like, you should study, like how we can shoot the standard better buddy.  And so he had—he had to get creative and name stars after the De' Medici Family.  And right, like there's this aspect of memorabilia and collectibles, that's connected a little bit to science.  

And so I started working on it.  I wrote an essay that was pretty well received.  I was—I was lucky enough to get some appreciation out of it.  And it was canonized in the A16Z NFT Canon, which is really cool.  That was on science NFTs.  But then, Science NFTs were never really going to be a thing before Art NFTs.  And so there was a—there was like biting, biting our time and waiting for the moment to launch the coolest science NFT that went on in 2019 and 2020.  And I did some work for Dapper and some other companies in the meantime to sort of like, wait.  

And then when it happened in 2021, as we remember.  So we were ready.  We had conducted a pretty cool study that we released and sold as a science NFT.  And then there was a decent amount of other backlog and things that had been planned for this moment that suddenly rolled out really, really quickly.  And we were talking about this a little earlier that like, in some sense, there's a lot to discuss and a lot to cover about Planck in part, because it's a two—there's portions of like a two-year rollout plan that had been compressed into like two or three months and a lot's come up pretty quickly.  But that was our first thing, was the science NFT. 

Host:  Yeah, okay.  So let's talk a little bit more about this science NFT.  And this is the one that you, you just sold for $24,000.  So you were conducting research.  What was the research that you were conducting for the study? 

Matt Stephenson:  Yeah, so it's weird because I, I—so I'm a behavioral economist.  I'm conducting research on something in a totally different field.  And the connection is that it was about trying to identify, like what kind of ideas could slip through the cracks under current incentives, right?  Like demonstrating the way that NFTs might be able to reward things that the system couldn't have rewarded before.  It felt like a really cool and noble thing.  So looking for a good example of like a theory or a piece of knowledge or a piece of science, that just had been forgotten.  

And having upon this one from Seth Roberts, who was a Berkeley psychologist and kind of a rat researcher actually.  That was around the idea that flavorless calories could reduce people's appetite.  And when you release this, there was actually a decent amount of fanfare.  He was a pretty serious thinker, even though he wasn't a nutritionist.  And his statistics were a little weird.  They were self-experimentation for the most part, but at the same time, some really credible people like Andrew Gilman and Alex Tabarak and Aaron Swartz, the co-founder of Reddit and Eliezer Kelsey, and some other people took it pretty seriously.  And they were like, there's something interesting here.  We all can't wait until the clinical trials will obviously be done because obesity kills two billion people a year—two million people a year.  

There's a lot of like personal hardship faced by people who can't control their appetite, but also just people who want to be vegan and they're tempted by cheese pizza, right?  Like there's this way that controlling your appetite could be a really powerful unlock.  Somebody thinks they can do it.  It's published in the New York Times, right? Credible people are saying this should be studied.  Just never happens.  It's not patentable, right?  Like you can't patent a flavorless calorie.  

And in fact, the researcher who, who is no longer with us, was pretty—he was a pretty good scientist on this front.  He did not try to say like, "Oh, it's this special oil that does it."  He actually went and studied.  Gosh, could I hold my nose and eat?  Does that count as a flavorless calorie?  According to him turned out, yeah.  Like according to his data.  

So - - randomized controlled trial of it, crowdsourced it, and we had people buy olive oil and get randomized into conditions with a website that I built.  And I programmed up e-mail protocols so that they'd be e-mailed to track their hunger and see whether the flavorless calorie conditions was different substantially than the like flavored calorie condition.  We ran the design by a prominent statistician on the stats blog as well.  And then collected those results, looked at them and put, put a graph together.  And that graph was the visual representation of the NFT. 

Host:  Got you.  Got you.  So what were the results?

Matt Stephenson:  Okay.  So the results—I'm finishing up writing something on this right now.  So it's on the brain.  The results I will say in my opinion look very good.  And I will say that I think most anybody who is looking at it would say that it's more likely having seen the results that this theory is true than before having seen the results.  But I do think it falls short of a full scientific finding because when you run a crowdsourced study, you can't control the number of people who sign up.  And this will take us into the scientific weeds.  

I'll let you guide me whether, whether you want to circle back and have me detail this part.  But I guess I'll just say that there's a thing you try to do before you run a study where you figured out how many people you want.  And I wanted this many people, like 800 people.  And I got 200 people, but actually, 150 people who stayed on and did it, which I find remarkable.  I mean, they had to buy olive oil and drink it, like that's—and do it for five days.  And like respond to my annoying e-mails three times a day for this random theory they just heard about.  And so I am eternally grateful to them.  

At the same time, I knew go as soon as the study was run, like we're not going to have what's called the statistical power to get a full result.  So we have a statistically significant results.  I ran it by a couple of statisticians who took a look.  I mean, I think everybody who seen the results agree that it ranges from like, not bad to like, oh, this is actually pretty promising.  And more importantly, we're taking the money that the NFT sold for the $24,000 and we're running like a full-scale replication that will ensure we get the, the power available to really test it.  That'll happen this summer. 

Host:  Got it.  And so was that sort of the goal from the start, is like, let me mint this as an NFT, sell it for however, much I can so that I can conduct a full-blown study?  And is that the model that you hope to replicate in the future with more science NFTs that come out of Planck? 

Matt Stephenson:  Yes, exactly right.  Because at the end of the day, even though this, this one didn't get the power.  Even if we hadn't got the power, I think it—or if we had gotten it, I think it would be the case that it would still just feel a little bit weird, right?  It's nice to have this research conducted by an independent lab.  

And so modelling this way to bridge-like citizen science or perspective science in the direction of real science.  So that even if it does fall a little short for whatever reason, you have this ability to sell the NFT and then generate the thing that can actually close the gap and bring it to like a full scientific result, or reject it, right?  As pseudoscience, which of course we want to—we certainly want to do that too, right?  Like it's real science.  

So we want to make sure if it's false, we falsify it.  But yeah, that path of you sell the NFT of either the prospective study or the earlier preliminary study, and that can fund the connection to the research.  And then that early NFT that the person bought can grow in value if the research is validated.  Because, I mean, if this thing changes the world, I would expect the person who owns the NFT, that to have more value to them at least personally.  I mean, it's a cool thing, right?  Like it's proof, they rewarded this research that really like could change things quite a bit.  I mean, Aaron Swartz called it like a future without obesity or something like that.  And you just basically imagined it was going to be this crazy transformative thing that goes on.  Yeah, I would think this NFT would mean a lot. 

Matt Stephenson:  Yeah, very cool.  And so who ended up buying your NFT?  Was it like an institution or was it just a person that was like, I really want to know the results of this, this question?

Matt Stephenson:  Yeah, so I was—so I was—we, we actually got a decent number of bids, and I knew a few of the people who were bidding.  And so a few of them who I got—they got some texts from me.  Like you better not win this.  But the person who won, I did not know him.  I talked to him the day after though, his name is Paul Koolhaas.  And he was working on a really cool project himself on NFTs that sort of like hold intellectual property around health.  So like NFTs that actually hold patents, real patents, like legally enforced patents. 

 And it was a thrill to like talk to him.  It was a great gift that he thought this was cool.  I guess it's kind of no surprise that he would think cool because he was working at least adjacent enough.  And then we, we hit off great.  And I'm a big fan of his and I feel like it's, it's mutual and at any rate.  yeah, I really liked his stuff and it was cool to meet him. 

Host:  Yeah, very cool.  And then when you—when you mint to this and put it on the market to sell it as an NFT, what were your expectations?  Did you think it would sell for—was 24,000 like in your ballpark like it --

Matt Stephenson:  -- That was outside. 

Host:  It was outside.

Matt Stephenson:  I didn't think that was going to happen.

Host:  Okay.

Matt Stephenson:  Yeah, yeah, that was that was better than I thought.  I mean, my hope was that it would—it would be a number enough that it could make a little bit of a splash.  So I was hoping for at least a few 100 or something where it would be like, Oh, that's interesting.  If it's like we sold the first science NFT for $2, I feel like less exciting.  I would have still been proud.  I mean, it was—it's a cool thing, I think, to have done it.  But it was nice that the number is enough to actually start to, to prove out and probably fund replication.  

I mean, the full ideal lab replication that we would—we wanted to do and the design which has like, we wanted to test people's Ghrelin hormones.  And there's a way to really, really do this and figure out what the mechanism is.  That's going to be—that's going to be well outside the bounds of 24,000, but, that will get us enough to do a pretty good replication.  So that's beyond the expectations and ultimately our wildest dreams, I think.

Host:  That's awesome.  And so how far along are you with conducting the, the, the fuller research on this study? 

Matt Stephenson:  So validated what the lab—what the two options for labs would be.  And what we have going on now is, we're going to do a design competition because again, I'm not an expert in this, I design the test to the best of my ability.  I mean, I've run a lot of experiments, but I'm not a nutritionist.  And so I think the design was fine.  But we want to make sure that if this test hasn't been run for 15 years that people clamoring for it.  So if we have a chance to do it, we want to make sure we do it as, as make sure it is well specified as we can.  

So there's a design competition, which should be rolling out soon.  I had to wait on Judge timing a little bit and get a few things arranged.  But we have two really cool, amazing judges who will be judging the design and that's the design we'll be rolling with and submitting to the labs.  And hopefully we get something that's, that's pretty cool and surprising.  But worst case, I mean, it will just be an improvement on the original design and, and that's something that I think will run pretty quick.

Host:  Yeah, and then once you have the results on this full-blown study, are you going to publish that again, as minted as an NFT and, and sell that idea?

Matt Stephenson:  Indeed. 

Host:  Okay. 

Matt Stephenson:  Indeed.  Yes, and in fact, that's doing something kind of cool.  That is the next thing that we plan rolling out actually.  The mode where that money goes is, is to a prize, which I think will be kind of a cool thing.  So the original—I think I mentioned clearing like 2 years back, like the original NFT paper we wrote was called, Post venture capital in the crypto Nobel Prizes.  And it was supposed to be about this idea that NFTs could have any value but also that prizes could be NFTs as well and prizes could be exchanged for NFTs.  

So we're taking the money for—which will hopefully be—hopefully be non-trivial.  I mean, you never know, but at least it'll be something.  And, and that will go to seed fund a prize.  And these prizes are kind of neat.  They—the prize design, not only goes to reward new innovations in science, particularly, things that were sort of like—we're looking for things that were forgotten or left behind.  Like studies and that are what the prizes can be designed to award.  

We have some really cool judges for that as well.  And the cool thing about the prize design is that it's designed to sort of be self-perpetuating, such that the prize funds get topped up by the NFTs that you exchange the prizes for overtime.  So the hope is that you get this kind of like autopoietic is the $10 word.  I think the better word is like self-sustaining.  And that we're seeding what's hopefully this like self-sustaining prize fund that will go to award this stuff in the future.

Host:  Got you.  Got you.  Okay.  So—so explain to me a little bit more how Planck fits into all of this.  Is Planck the platform where people can mint their NFTs or is it sort of just like the, the breeding ground for this community of scientists who want to run these experiments or tell me a little bit more about that?

Matt Stephenson:  Yeah, you know, so when, when we decided to launch it and go live in 2021, I think one of the real guiding things was, I didn't want to duplicate efforts with anybody.  I wanted everything to feel very, like collaborative to the furthest extent possible.  And so our goal for the year has been to try and use tech to the furthest extent we can rather than reinvent any wheel.  

So we used OpenSea as a platform.  First, we used Mirror as a platform for a secondary NFT sale.  There’s a zero-knowledge proof protocol that we used so we could have some statistician's query results.  And we did that with an existing company called Elio who they were really good to work with.  

And so I think what's going to happen to your point, is Planck will probably eventually have to be a protocol in the sense that there are parts of our design that are far enough away from what exists now.  Then we're going to have to start to build the stuff ourselves.  But - - let's do as many things as we can with the tech that already exists.  Because I mean, it's just amazing what people have built in the empty space over the last year or two and it would be a shame to just like, I'm so thrilled with what Devon's done with OpenSea and what Dennis has done with Mirror.  The opportunity to play with that tooling.  Is, is like fascinating and amazing.  So we've tried to do that as much as possible. 

Host:  Yeah, for sure—for sure.  And then another question that people have with NFTs and I think this is maybe and even bigger question with science NFTs is, how does IP apply to that?  So normally, if you conduct scientific research, you publish a paper, you know, you might want to patent whatever discovery that you've made.  And so with the science NFTs—I know you've talked about something with like an alt IP system.

Matt Stephenson:  Yeah.

Host:  Okay.  so can you talk a little bit more about that. Like how that's similar to or different from our traditional IP system? 

Matt Stephenson: Yeah, so I, I call Planck sort of an alt IP system.  That's the role of it is kind of similar to the role that IP plays, but it's explicitly alt because it's not trying to use any governmental mechanisms of enforcement, so right.  Like the way patents work are, the government gives you a legally enforced right to excludability such that if anybody uses the thing that you patent—patented, you get to sue them, right?  So that's—they’re excluded from use.  

Planck doesn't involve—isn't involved in patents that are in working in that sense.  And there's a company Paul, who I just mentioned, he's doing something that does and I, I mean, I think there's definitely—that's a very cool thing and a very cool use case.  But you think about what a patent is supposed to do—what a patent is ultimately supposed to do.  You go back to the early history, it's attempting to just ensure that people who did high fixed cost innovations, right?  Like, these things that take a long time, but then once you reveal them to somebody, then it gets free to copy this thing that you see on the internet all the time.  

Early innovations, like somebody would invent the Marine chronometer, like this incredible navigational tool that would allow ships to sail.  But then once you show somebody a Marine chronometer, they can just take it apart and figure out how you did it.  So yeah, it took you five years, but now you have like a thousand competitors.  And so the original patent idea was just to make sure that the person who did the original thing is rewarded sufficiently that they could do it again.  

And so Planck is attempting to be a system that can do that, right.  And so you can see that like even to take the first example, Seth Roberts spent a lot of time on this.  We spent a decent amount of time, nothing nowhere near the amount he did, right, but trying to run the replication.  Whatever process eventually generates this as a finding if it is indeed a finding.  The finding will just be there, right.  Like, there's no patentable enforcement of it.  

An Olive oil company can start tomorrow and start highlighting this study and saying this proves that you should be drinking extra light olive oil and you lose weight or something like that, right.  And, and there's no defensibility for another kind of—for the person who necessarily did the work, nor should there be.  I mean, I can't run an Olive oil company.  I'm sure Seth could neither, right.  And so you actually want to enable this sort of division of labor which I think is what patents where all about, right.  

This ability for somebody who invented the Marine chronometer to open it up so people could use it and improve on it and so on, right.  While also having a protected Planck's trying to effectively do the same thing with NFTs through these two channels.  Channel one is just this, this connection with the future that memorabilia value can give you, right?  You can tell the NFT to somebody who might see the potential in it in the future means that even if, if it gets spread and change the world through companies that don't care at all, about what we do, the idea will at least matter.  Sort of anti-rival property and person who funded and started the idea and sort of has the NFT, will have a little extra because they own it.  

And then secondly, there's this piece called the slipstream, which is designed to flow value back toward things that aren't naturally appreciated by the market.  And those two things in concert, are supposed to replicate some of the things that patents are supposed to solve around for awarding these, these high five fixed cost, knowledge generation sort of goods. 

Host:  For sure.  Very, very interesting stuff.  And so zooming out of this a little bit, how do you think allowing scientists to mint their scientific studies as NFTs is going to further scientific research or, or what is this going to do for like, the broader science field in general?

Matt Stephenson:  Yes, so I mean, I feel really positively about this.  I, I, I guess as you can imagine, in Economics, we think we know a couple of things.  Not that many things.  One thing we know we think, is that ideas are the only source of wealth in the worlds, right.  Like we—that can sound a little weird because you think of wealth as being money, but wealth in terms of growth is just new ideas, what Paul Romer calls new recipes, new ways of organizing matter in the world such that we used to be travelling this way by horse - - become more efficient, right?  That—those are ideas.  

We also think that ideas are this kind of ultimate public good.  Meaning ideas are the sort of things where two—if you've quoted them well, if you've expressed them well, if you've explained them to somebody, they're just owned by everybody immediately, right?  And you can't exclude other people from them once you've communicated it.  And we know that public goods are underfunded and under, under-provisioned.  

So if that's the case those two things together give you a pretty, either a terrible story or a very promising story about what you can do if you change the incentives by a millionth of a percent.  Because it tells you, ideas are pretty much all we have to improve the world over time and we are absolutely terrible at rewarding encoded ideas, right?  And if we think incentives matter which that's the other sort of poor principle of economics then you just stack those things together.  We're terrible at rewarding incentives or incentivizing innovation and knowledge sharing incentives matter.  Thus, people are not sharing as much knowledge or generating as much knowledge as they should be and this knowledge is what generate growth.  

We are just incredibly far away from the optimum, right?  There is potentially a lot of wealth idea generation sharing and so on that could and should be generated if the incentives were different and if we can get even a little bit there like a little bit closer to that, that would be this big whole world changing thing.  And so if scientist are able to fund things that otherwise wouldn't have been able to be funded and other scientist are able to look over and go, "oh I could actually kind of tinker around with this."  And generate may something kind of cool and that research turns into something cool and they do this because they see the NFT sales and they otherwise wouldn't have.  Yeah, I think that could really be amazing.

Host:  Super cool.  And so in your perfect world where would we be with NFTs in let say ten years?

Matt Stephenson:  I think in my perfect world, NFTs are normalized.  They're maybe called NFTs maybe not.  It maybe start to feel weird for things to be fungible online more than non-fungible.  Just because I think, as the digital space becomes more normalize and people, people kind of lose their biases that are like, oh your weird for looking at art on your phone, you should look out on our wallets like we just—we are more and more on the digital space, right?  

So the idea that the digital space will become more reflective of our natural human impulses and desires and the sorts of things we really love and value might look around what, what's—this is a $10 phrase but like the biographical indexicality is what it's called when you look at, like these people who study collectibles, like why do people collect?  Why do they treasure—I keep saying a scarf or a tea cup or whatever it is?  It's because these objects provide a biographical index to events that really have a lot of meaning for us.  

Like, this is a beautiful human vein that we are really used to in life and having that on the digital space is—to me feels like so natural and so desirable that I really expect that to become more of the norm—in so far as it can be the norm.  I mean, it remains to be understood like what you need to meet that sort of like threshold for a minimally viable NFT but, yeah that would be my dream.

Host:  Awesome.  Yeah, I certainly hope that we'd be there in ten years too.  I would love to see that as well.  So switching gears a little bit, something else I want to talk to you about is your, your PhD in Behavioral Economics and how that ties into web three and Dows in particular and so maybe a good place to start is what is Behavioral Economics?  What do you actually study?

Matt Stephenson:  Yeah, so Behavioral Economics is some field of economics that was designed at—it started with sort of poking at some of the assumptions of rationality in economics in the 80's.  Infamously economics has this assumption of like perfectly rational selfish agents.  And Behavioral Economists started to say like, "well people aren't always that rational and people aren't always selfish, and we can design some cool experiments that demonstrate the ways in which people would predictively and repeatedly violate this sort of countenance."  

So that's Behavioral Econ.  It tends to be really of an experimental discipline for the most part and I intended to run and, and design a lot of experiments.  And I particularly worked I this area of trust in division of labor like the ways that people naturally like cooperated in interacted in tasks, and the way that they like to work with people that they trusted or they'd work with previously and so on.  So that was my field, but Behavioral Economic is typically like other—what are called other regarding preferences and then like kind of biases and rationalities.

Host:  Right.  And so super interesting how you've been able to tie that into crypto because there are a lot of question about how people work together, what incentivizes people, and all of these things especially in the realm of Dows but what are some things that you've studied I guess in the crypto space that you can really apply behavioral economics to?

Matt Stephenson:  Yeah, well actually, I mean, NFTs are an interesting example where one of the famous behavioral economics experiments is, is what sometimes just called The Mug Study.  Nobel Laureates, Richard Thaler and co-authors did a study where they gave students these mugs and then they tried to buy them back from the students effectively.  And they recorded all the values, right.  And the values were like $5 a mug.  And then with a different class, there was also a really large.  So you can sort of expected that that if they are comparable. They just auction off the mugs in the first place.  

So both of the models and the, the values were much lower, right?  It's like $2 a mug or $3 a mug or something like that, right.  So what you get is the sense that once the mug is owned, once it becomes part of your story, it's valued differently.  So that study formed the basis of what's called Prospect theory and, and the endowment of fact, like this Nobel prize winning research.  But if you thing about it, there's a way in which that mug study is about non-fungibility.  About the way that this mug that economist think should be perfectly fungible within other mug, is made different because of its unique history and that unique history is I own this mug, right?  Therefore, it's valued differently, right?  And that's a—that's a neat thing to think about it in the NFT concept.  

So beyond that I mean organizations and corporation and the ways that like monetary incentives get complicated and attention, attention is scarce and social incentives can kind of be crowded out by monetary incentives and all these fun behavioral economics things really come to head a lot in the way that Dows are obviously a big example of place where you think about it a lot.  

And I mean, public goods in general which is something that there's a lot of talking, a lot of discussion in this space around solving, there is work done on sort of like the experimental economics and behavioral economics and public good.  So I use it pretty much in literally everything I do in crypto on some level or another but, but I think the, the Mug Study is like the one that probably helps me understand why the NFTs might not be that weird.  

And the rest of it is just like, when you do behavioral econ. experiments what you're really doing is you're doing—you're sitting people at a computer and you're giving them a very abstract incentive game.  You're not telling them—you're not saying like hey, you're going to play a banker, and you're going to play a bank teller, and you're going to play—like, you're just giving them—you have $15, you have $10 here's what you can send.   You don't give them a story.  You see how people interact over this very abstract set of incentives.  

And I do think that in my work—that in my work, it, it block science doing some incentive design and obviously a lot of my work are played.  I do think that's also just been helpful.  The way that when you interact with small contracts and you interact with these incentives structures and mechanism designs that are supposed to scale up really big like you see in block chains, a lot of that is going to present to people as this person gets this amount of money—this person gets this amount of money and the context in which that changes can—or that—the context that defects that can vary a lot.  So it's nice to have this kind of like intuitions and experiments to flow back on.

Host:  Yeah, so I, I want to dig a little bit deeper into public goods because you've brought it up a few times already and you wrote this article, Public Goods Problems versus Coordination Problems which I think you co-wrote with Scott Moore from Bitcoin who we have --

Matt Stephenson:  -- Yeah and Michael Zargham.

Host:  [Interposing] on podcast recently.  Yeah, okay.  Great, so I'm going to ask you these same questions I asked him but, I guess before we do that why don't you just give us a little background on what is the Public Goods Problems?  What is Coordination Problem?  What's the difference between the two or how are they similar?

Matt Stephenson:  Yeah, so there's interesting relationship between Public Goods and Coordination Problems especially, you see it a lot in like web3, and the space where people will talk about them in these ways that sort of confuse the two things.  And in experimental economics, they're, there're different and they're different types of games.  So public goods games are the sorts of games where there is an optimal solution that would make everybody in the game better off and you just can't quite get there.  And that's what a public good problem feels like.  

And to me that's the important part is when you play these games and when you interact with these and in these public goods problems, they have a feeling.  They have a feeling where you see the way - - off, and you also would just find as you try to do it, you can't get there.  So the analogy we use in the paper is like you've all—everybody's probably lived with some slobby roommate at some point.  And sometimes it would be like gosh, wouldn't we all just be better off—everybody in the house wants clean dishes, right.  That's a public good problem because we can see the state of the world which is better for all of us, right?  

And we can all just get there if we will all just wash the dishes and maybe someday, you're like let's all do it.  Today we're going to do.  We're going to clean up and the dishes are going to be clean and we're going to feel so good about it that we never let the dishes get dirty and fill up the sink again.  And of course, it's just what happens typically is that a week or two later you just let the dishes start filling up again.  And that's—that to me is the best way to sort of explain the difference between a public goods problem and a coordination problem.  

When you think you can get to the optimum and you try to get everybody there but you can't stay there, that's a public goods kind of feeling, right?  And that's the frustration of somebody who thinks they're in a coordination game.  Coordination game being one where if you just got people to that state, they'd stay there.  Like if you just made the dishes clean one time, people would clean them forever, that's a coordination game.  

Public Goods game, if you clean everything up, it'll just get dirty again.  And the interplay between those two, I find it's fascinating and complex and in some sense kind of undecidable but those are the two concepts and how they're different.

Host:  So what's the solution to the public goods problem?

Matt Stephenson:  The—so weirdly and here's the kind of catch, there's no such thing as a solution to a public goods problem.  As defined, public goods problems cannot be solved and that sounds like a sort of pedantic, annoying like game theory approach.  And in some sense, it is, right.  But the real trick is—it's a—it's a creative act to solve what we call a public goods problem and the creative act is to suddenly see the way that the public goods problem is actually a coordination problem, right?  That there is a solution that exists.  

So I'm not—I'm not sure if we want to stack analogies on analogies but the other one, we used is standing up at a concert which has this similar sort of thing, right.  So if you're sitting down on a concert and enjoying it, somebody stands up two rows in front of you, you have to stand to see.  If everybody suddenly stands, their view isn't any better off than it was before, right?  But now you're stuck here because if the person three rows in front of you sits, it doesn't change anything your view still bad, right?  And so there's the stickiness in the bad equilibrium.  And so let's say everybody wishes they could be sitting, how do you get everybody there?  Well, you can't just do it, right.  Like you can't just sit.  

And so how do we reimagine this situation?  Well, I don't know if you have a bullhorn, you could maybe yell at everybody, right?  But there's also I think Andy who of course, we, we both know and you had on the podcast who's, who's excellent at Colonel he had the suggestion that, well, why don't you just try to make everybody dance like why don't you make them enjoy standing up.  Totally brilliant solution, right?  Like it's just another way of reimagining what seemed like a public goods problem where we had to get here and coordinate everybody there as a coordination problem where this is now the socially optimal equilibrium, right?  And so these creative solutions actually want to be in the way.  

The only reason I'm being a little careful in my language is if you know any of my old game theory teachers who are listening to this, they're going to go.  There's no such thing as transforming a game from public good problem to a coordination problem.  That's correct.  Like the game is what it is, right?  And the real way to think about solving it is to think that we are fooled into thinking that public goods problems are in fact public goods problems and they must be unmasked and revealed as the coordination problems that they are.  And hopefully there's no such thing as a public goods problem.  Because the public goods problem are terrible.  We don't want them to exist, right?  Like there are these unsolvable problems where we can make ourselves better off.

F2:  Okay.  So one application of this that—I think about a lot in web 3 is with open-source.  Everything is open-source and web 3 whether it's code—whatever the case may be research anything.  And so we're talking about open-source everything I think we face this problem of free riders where you know people—some people may not contribute their share of it and yes, it's like in a perfect world.  Everybody would just contribute all their knowledge and then this—the whole world would be a much better place.  It would be full of knowledge.  It would be like two heads is better than one, however, many billion heads is better than one.  But that's never going to be the case because you're always going to get people who don't want to contribute and just want to take out.  So how do we solve this—or how would you frame this for me?  Is this a public goods problem or how would you frame this?

Matt Stephenson:  No, that's, that's definitely the right way to think about it, right?  Like the analogy to the dishes people—those are the people who don't individually wash their own dishes, right?  Those are the free riders in that analogy and so yes, with like open-source software and with knowledge sharing and so on the problem is with—there's always a strategic advantage to—like in terms of monetary incentives that is always as constructive an advantage.  So what do you do to solve the problem?  I think there are these cool, like creative solutions.  

Obviously, bitcoin has a really cool on the days implemented with quadratic funding which is—there's a lot to explain about it, right?  Like one of the things is it's a very efficient way of trying to like get people to reveal their preferences and express their preferences in a way that doesn't just depend a lot on their wealth.  And so that's a cool way of building and rewarding public goods that people want.  But in terms of knowledge sharing and rewarding things that have already been put out there, that's kind of a different problem in some ways a backward-looking problem.  

And weirdly in some ways it's actually kind of like a problem that has to be solved by some people before you can even do quadratic funding because you have to make a quadratic funding proposal which is in some sense sharing the knowledge.  So if the knowledge is itself a public good then it's never actually clear why the person who made the proposal is the one who should receive the quadratic funds.  Because why shouldn't they go to the person who can build it the best?  Those two people shouldn't be the same.  

And so I mean, for the things that are funded—I mean, there's a lot of things that go into that where a lot of times the people who are suggesting it are the best ones to do it or there's this sense of reciprocity and the reward.  And so the system works, but really analyzing like how the system works and what it sort of depends on like these norms around, well, the person who should share it should get rewarded because they shared the proposal.  

That's kind of a cool and interesting thing to poke at and recognize as actually not—it's not literally irrational, but if you tried to look at it in terms of like the very dry economic incentives it is kind of irrational.  It's like well, no, we should take this idea.  And we should instead give the funding to the person who can execute it.  And if that's not you, then you shouldn't get anything.  So there is this weird like knowledge sharing problem that even exists at the root of, of that as well.  That is, I think in fact solved, but it's very interesting to just think about the way that is solved.

Host:  Yeah, so I'm going to ask you the same question asked about NFTs.  In your perfect world, how would open-source work with—obviously, being realistic too.  It's not like it's not like a totally utopian world where everybody just willingly contributes everything that they can and everything's perfect, but in your perfect world, what would be the best solution to the public goods problem as we move forward and building web 3 and moving everything towards an open-source environment?

Matt Stephenson:  Yeah, I mean, I think—I do think that public goods problems are going to have to be probably solved like application by application or domain by domain, but for knowledge sharing which is the one that's near and dear to me and close to my heart, I mean, I think the best solution and, in some ways, one of the only solutions that you can really work it with is to try and commit to solving it backwards.  There's two ways to look at it, right?  One is you try to say we're going to like algorithmically bounty some knowledge sharing, right.  

And we're going to have some way to determine what that is, right?  What the best possible thing is.  And I think that can work and it does work, but it's—it relies on people expecting or knowing what they want.  If you would have asked people what they wanted in 1900 they would have set up a faster horse rather than a car or something like that, right?  And so there is this way that forward looking bounties have a limitation and that limitation is like human expectation.  And for things that are really transformative, it's often just really hard to reward that stuff by guessing at it, right.  

You can't say I want the theory of relativity to exist because one day, I'll have GPS on my phone.  That happens to be true, but you can't understand what GPS is and how the theory about it.  Einstein can't tell you probably the GPS is going to exist if you get the theory of relativity.  And so design is around trying to work backwards from that.  And trying to basically say if we can connect this thing that we like a lot to this earlier thing, let's try and make sure that a little bit of the value flows.  

And that commitment over time can allow more people to share and maybe some, some entrepreneurs—like, ideal entrepreneurs to try and collect these early NFTs of ideas which can maybe come one day to mean a lot more.  So that's kind of our approach, but I think both, both work in concert and the one that's very forward looking I think obviously like bounties are amazing for that.

Host:  Yeah, for sure—for sure.  All right.  Well, thanks so much Matt, for being here.  I have to cut it off here, be, be cognizant of your time.  I, I feel like I could just talk to you for hours about all this stuff.  You have so many interesting articles out there, but so before you go, tell people where they can connect with you personally if they want to chat with you more or also like where they can go to find your writings because you've got a ton of those and all those are really good.  And then tell people where they can go to learn more about Planck as well.  

Matt Stephenson:  Yeah, so you can follow me on Twitter @Stevenson H Matt.  If you are interested in trying to fund a—fund a self-sustaining prize and you have—and you think that's a cool thing you can—you can get in touch.  My DMs are open and maybe we can talk.  And then my website for Planck which we're going to push live by the time you hear this is, I hope is P1anck.com, but the, the L is a one.  So P-1-A-N-C-K.com.  It's like planck.

Host:  Cool, cool.  We'll include that in the show notes too.  So people can just click through to add easily if they get confused.  Well, thanks so much Matt for coming on the Podcast for taking the time out of your day.  I really appreciate it.  We'll have to have you back for part two of this because we have a lot more to talk about.

Matt Stephenson:  Yeah, I will love that.  Thanks so much for having me.

Host:  Awesome.

Matt Stephenson:  I love your podcast.  You, You, you do the best interview So thank you.  This is great.

Host:  Thank you.  Thank you so much.  All right.  Chat with you all later.

Matt Stephenson:  All right.  Take care.

ソーシャルメディアで繋がる

ヘルプセンター