Image credit: Ken Flerlage
This article is a guest post by Joey DeBruin with a contribution from PrimeDAO. Joey is a neuroscientist turned product builder, previously head of product at ResearchGate and currently the cofounder of Backdrop. Backdrop is a network that helps people discover and leverage the magic of tokenized communities.
“Perverse incentives work like an ill-tempered genie, giving you exactly what you asked for but not necessarily what you wanted.”
- Stuart Ritchie, “Science Fictions”
Scientists are some of the most selfless, anti-competitive people on the planet, but the incentive structure in academia makes it a difficult and competitive place to build a career. As a result, biographies of scientists read like novels, and few are more gripping than “Code Breaker:” Walter Isaacson’s biography of Jennifer Doudna and the race to bring the magic of CRISPR to the masses.
One of the central story lines in the history of CRISPR was the rivalry between Jennifer Doudna’s lab at Berkeley and Feng Zhang’s lab at Harvard. I won’t spoil the book, but it includes down-to-the-wire battles for being the first to publish, lawsuits over patents, and more.
Rivalries like Zhang and Doudna’s are so competitive for the same reason that journalists worry about getting “scooped” or 99.9% of startups will only ever dream of being Google sized — these systems have strong winner-take-all dynamics. In science, the first to publish will often get the majority of clout and citations, which translates to funding, patents, and more.
Participants in a winner-take-all system typically think about two different options: either I win, or they win. The tragedy which usually happens without most people noticing is that a third option is often more likely: nobody wins. Participants in a hyper competitive system don’t think about that third option because it’s out of their control - why waste time thinking about a world in which everyone loses?
In order to increase the amount of innovation we get as a society, we should care a lot about that third option. There are countless accounts of discoveries that were so close to being discovered only to disappear for decades before finally making it to mainstream knowledge. The work of Gregor Mendel, the father of modern genetics, never made waves until many years after his death. We see the same in tech (see “The Boneyard Principle”).
Should we just accept this lost time as a natural component of discovery? I don’t think so. I’m going to argue, with some simple math, that the optimal thing for Zhang and Doudna to do in this situation would essentially be to ‘bet’ on each other. And I think in the future they’ll be able to do that. One of the most powerful trends that’s happening — everywhere from investment platforms like Republic to tokenization of new projects in crypto — is increased liquidity at very early stages of projects. “Swapping” ownership could provide a real way to increase the size of the pie at the early stages of the innovation pipeline.
To make the case for ownership swaps, we need a bit of simple math. Let’s say in a winner-take-all system there’s a 20% chance Doudna makes the breakthrough, a 20% chance Zhang makes it, and a 60% chance that neither of them do. This may not have been the right odds for CRISPR, but all that matters is that there’s some non-zero chance that discovery stays locked away.
For simplicity’s sake let’s also pretend that citations are the only thing that matters. If either Doudna or Zhang wins, they get 100 citations and the other gets 10. If neither of them win, both get 0. So each of their expected values is (100*.2 + 10*.2) = 22 citations.
Now imagine if these two leading scientists were to work together, to share insights rather than hide them. If Doudna makes some progress, she shares it with Zhang fully open source to help him move forward, and vice versa. Let’s assume that full collaboration doubles the chance that one of them makes the discovery, so 40% chance of Doudna and 40% chance of Zhang. Now, each of their expected citations is (100*.4 + 10*.4) = 44.
You might think “why don’t Zhang and Doudna just merge labs if this is so much better?” The answer to that is that reputation accrues mostly to the individual in science - even in the same lab, Zhang and Doudna can’t share the credit. So if a full merge isn’t possible, how might we get closer to that positive sum outcome? One answer is a version of the age old saying “keep your friends close and your enemies closer.” Keep your enemies’ tokens. Of course we would prefer in this case to not think of them as enemies at all, so a better phrase might be “Keep your allies’ tokens,” but you get the point.
In essence, what Zhang and Doudna should do is to make winning a more even split. Let’s say that there’s a way for Doudna to own a 30% stake in Zhang’s future citations, and Zhang does the same. Now if one wins, they get (100*.7+10*.3) = 73 and if the other gets (10*.7 + 100*.3) = 37. In this situation you would predict they would be slightly more collaborative, and therefore the chances either of them would win go up, which as we know grows both of their expected citations.
Clearly, there’s no way to literally invest in citations, but this general framework already happens in many ways — what I’m arguing for is just to take it further.
One of the articles I cite all of the time is Packy McCormick’s “Power to the person” which compellingly lays out the argument that the modern world shifts power from institutions to individuals. We trust individuals more than institutions, and modern technology gives individuals the tools to do amazing things.
It’s inspiring to think about the power people have to follow their dreams, but science has seen the opposite side of this coin from a reputation perspective. If reward accrues mostly to the individual, it builds walls between small groups of people that are focused on something new and risky. It hasn’t always been this way — before the modern publication system came to dominate academic reputation, academic societies like the Royal Society or one of the thousands of others that now exist helped individuals spread their risk/reward amongst a larger group.
One of the trends in science I’m most excited about is what I would describe as the rebirth of powerful scientific societies — including Arc Institute and Arcadia Science — that spread risk/reward across larger groups of scientists in a number of different ways, including taking a share of the downstream economic benefit of their work. In crypto, Molecule is increasing the ease of patenting very early research via IP-NFTs to create more liquidity in risky basic science, while VitaDAO collectively vets and funds potential projects in longevity on Molecule’s marketplace (more on this later).
Both Doudna and Zhang have gone on to create lots of successful patents and biotech companies, so if they had issued an IP NFT on their future work and co-invested in it together, it would have been quite a valuable investment. Again we don’t have the counterfactual of if neither of them had succeeded, but surely there are many discoveries and products still stuck in the graveyard somewhere.
This, to me, is the opportunity with tokens and token swaps. For better and at times for worse, crypto has created highly liquid markets in things that don’t exist yet — yes, even memes. I wrote for example about how Alex Masmej created and sold a token that represented a share of his future income. The holders of that token did very well.
There are at least three ways that tokens can create liquidity very early on in the creation/discovery process, and enable the type of ownership swaps I’m arguing for here:
There is no right answer to any of these three options. Each has different advantages and challenges. But all of them provide the ability to create liquidity in things that previously were fairly illiquid, and that creates the ability for grow-the-pie risk balancing.
Ownership swaps are more than theory. I’m publishing this article in collaboration with PrimeDAO because they’re already facilitating these token swaps at scale. VitaDAO recently announced a $50,000 swap with PrimeDAO in order to align incentives as they leverage some of PrimeDAOs tools in order to help scientists and investors develop longevity focused biotech IP.
Consider what happens if VitaDAO encounters another DAO working in a similar space. Rather than simply thinking of them as a competitor, they can work with Prime to align incentives with that DAO via token swaps or even simply buying tokens of their competitor. Longevity research is the type of moonshot where if any of the pioneering startups succeeds, the impact on our lives would be huge — we’re all better off it they find ways to work together.
It’s easy to dismiss areas of crypto like this that are “already possible” with existing technology. It’s true that patents existed before IP-NFTs, and that you can agree to a stock option swap with existing startup laws. But ask a scientist how easy it is to sell fractional ownership of the downstream IP of their lab, or a startup founder how easy it is to coordinate a stock option swap with a company in another country. The earlier up the innovation pipeline you go, the worse the existing liquidity is because the proportional cost of implementing ownership is higher. The legal fees of patenting an idea or creating a startup are usually tens of thousands or even hundreds of thousands of dollars, something that’s not possible for most people unless you already have investment.
Where blockchains are truly a transformative technology is in lowering the cost of implementing and managing distributed ownership of something, and that allows us to push ownership and liquidity further towards the source of innovation - places like science where incentivizing collaboration can make it less likely that by playing a winner-take-all game we end up with everyone losing.
We should all hope for more experiments like VitaDAO and PrimeDAO, and for more tools to help create a more efficient market for risky bets.