Tokenized Thought

Thank you to Jocelynn Pearl for reviewing a preliminary version of this article and to Dani Bergey for figure design.

Utility of the NFT expands far beyond fine art and collectibles. Today’s mania does not reflect the more holistic vision and ubiquitous tokenization that is about to take hold. This includes physical assets like the deed to a house and the title to a car, capital assets like stocks and bonds, and as I will argue here, intellectual assets like patents and publications. Intellectual property as a digital object written to the blockchain will change the way we think about publishing science, opening the process to be more transparent and realigning incentives within academic communities.

The Past and Future of Scientific Publication

What we think of today as ‘science’ formed around a culture of natural philosophy, where a small group of thinkers shared ideas through both spoken and written word. As time progressed and groups of thinkers got bigger, individuals wanted to benchmark ideas as their own. The first official patents were developed through the Venetian Patent Statute in 1474. Eventually, patenting inventions alone was not enough, thinkers wanted to document their understanding of the universe, and the first scientific journal was published in 1665 by the Royal Society. This exclusive group initially consisted of only 40 members. Finally, in the late 19th Century, many more subscription and/or for-profit journals began publishing scientific articles. Notably, Science and Nature magazines were both started at this time. While much of the science during this era was published as somewhat informal letters between authors, it is also when the peer-review process was formalized and publishing became what it is today.

Many academics likely get the feeling of decentralization even now, especially with the myriad open science tools like Python, Github, etc.; unfortunately, the reality is that a few publishers control the space and that a few critical data centers hold most of the publicly accessible data. The majority of prominent journals still institute a paywall, publishing can be prohibitively expensive, and data products are often concealed well after the associated manuscript is published. Throughout history, unexpected events where the state seized control of information flow, like book burnings, have shown us that centralized powers can quickly and unilaterally manipulate free thought.

As a next logical step forward in the history of science and intellectual property, we can imagine publishing thoughts and inventions as NFTs on a dedicated Ethereum rollup. In this way, scientific information and the scientific process reach beyond a singular elite to become a public good. The landscape of blockchain-based science will include but strengthen all of the scientific platforms that we already use today. As open science builds upon the subscription journals that we have used for hundreds of years, so too will the future of blockchain publishing take advantage of the open science tools that we have been building in the last decade. The critical piece that will enable truly ‘open’ science is trustless ownership on the blockchain. Scientists will rest assured that a digital book burning is not possible.

Blockchain-based publishing will be an improvement to the user experience as well. Any person will be able to benchmark their idea from any place and at any time. Since the blockchain cannot be edited, there will forever be evidence that the person had their idea at the time it was published. This takes power away from publishers, whose sole purpose is to act as an intermediary between the idea creators and the idea consumers. Then, with a less restrictive publication process, scientists will tend to publish more incremental progress and more continuously. Continuous publication will slow down the rush of publishing a written manuscript to avoid getting ‘scooped’, and it will give credibility in publishing investigations that do not result in a peer-reviewed manuscript or those that give an ambiguous result. Today, those efforts generally go uncredited. Instead, scientists will focus on incremental publications like software updates through git (which can currently be thought of as a private and centralized blockchain), data-sharing through a decentralized database (IPFS), and of course written articles through blockchain-based journals.

Academic Coordination through DAOs

Academic communities will coordinate through Decentralized Autonomous Organizations (DAOs). Members of a given DAO will share common interests, they will coordinate to progress understanding in their given field. As an example, I will use a community to which I personally belong, the American Geophysical Union (AGU). As a DAO, the AGU would control a small treasury to fund annual meetings, workshops, etc. The AGU DAO would publish several distinct journals, as it does today, and would pay journal editors, reviewers, etc. from the treasury as well. The DAO would be governed by its community, and those who contribute the most would earn the largest portion of the voting rights. While geophysicists coordinate around the AGU DAO, chemists will have slightly different priorities and perspectives; they will form a distinct DAO around the American Chemical Society (sorry for the America-centrism, in reality these will be mostly international groups). The same goes for biologists, mathematicians, you get the point. DAOs are a global coordination tool, and who needs to leverage global coordination more than a dispersed group of thinkers trying to debate the nature of the universe?

Token-Based Incentivization

Ideas and inventions will be published on the blockchain as intellectual property, this much is clear. Further though, I envision the complete tokenization of academic achievement. Digital tokens will be used to represent not only publications but educational achievement, teaching performance, citation metrics, and effectiveness as a peer reviewer. Taken together, these tokens will be a continuous record of one’s performance. Hence, they will act as a public and ever-evolving CV. While the scientist of today is often fixated on boosting their h-index, a measure of citation count only, the scientist of tomorrow should be incentivized to continue learning, they should be more rigorously assessed on their teaching effort, they should be adequately compensated for their peer review. Each of these achievements will be tokenized as either fungible or non-fungible tokens. None of the tokens will have any monetary value or be tradeable at all. They will be desirable for the same reason that a high h-index is desirable, because they are a public display of one’s scientific status.

Take the example shown below. Scientists A and B publish a series of scientific work, some of which is reviewed by scientist C, and is eventually cited by scientist D. First, scientist A writes the software for their model and publishes that on her own, receiving an NFT representing her ownership. Scientist B, meanwhile, collects data to test their model and publishes that as a separate NFT. To benchmark their progress they then work together to create a figure illustrating their work so far. They publish this with the code from A and the data from B both contained as objects within the hash of the figure’s NFT. Because the figure directly uses work from both A and B, each is included automatically as an author of the figure as well as the subsequent manuscript. NFTs for those publications have two copies, one held by each author. This continuation of authorship down the publication chain remains true no matter how long the chain becomes.

Subsequently, D publishes a new manuscript which cites the original written by A and B. As soon as the citing manuscript is published as an NFT, citation tokens are awarded to the authors of the cited manuscript, A and B, signaling that their work was used in a new project. If D went on to update their manuscript, no additional citation tokens would be provided, only one citation per chain of scientific products. Citation tokens will be fungible, so authors accumulate them in the same way they accumulate citations through Google Scholar or any other centralized platform today. Citation tokens will also be divisible and rewards will be weighted based on the type of publication. For instance, a peer-reviewed manuscript would earn more tokens per citation than a standalone figure or data product.

Incentivized Peer Review

Scientists are currently expected to act as peer reviewers on a purely voluntary basis. There are no requirements from journals that enforce a specific quantity or quality of peer review. Instead, the blockchain and DAO-centered journal model will offer additional tokens for peer reviewers. In the example provided, scientist C acts as the reviewer for the manuscript submitted by A and B. The review is initiated as a smart contract which enforces that upon completion of the review and acceptance by the editor: 1) the reviewed product will be republished as a new NFT with the reviewer’s signature proving that the review was done, and 2) the reviewer will be rewarded a separate NFT representing their association with the object. Upon citation of the reviewed object, citation tokens are also awarded to the reviewer, C, in order to incentivize a strong review. Of course, the review is less valuable than the work itself, so the rewards will be weighted appropriately.

Often peer reviewers wish to remain anonymous. In general, the blockchain model for intellectual property would be open and public; however, it is important to maintain the option for private review so that critiques can be honest and the reviewers do not face direct scrutiny. If a reviewer chooses to remain anonymous, their NFT will be held in a private wallet, and any citation tokens will accumulate through private transactions using zk-proof technology. Citation tokens will appear in the reviewers public wallet, so the fact that they acted as a reviewer on some scientific document that got cited would be public information. The same would be true if an individual wanted to publish an article anonymously but still get credit for the work.

Distributed Data Storage

Storage and sharing of scientific knowledge including data products, articles, software, etc., will be on a distributed database built on IPFS (e.g. filecoin). In essence, volunteers from the community rent storage space on their computers which are then used to store the complete archive of scientific information, and those volunteers are incentivised to do so. Because the network is decentralized, access to this information cannot be restricted from anyone or at any time. Protocols that are currently running decentralized storage networks use rational proof-of-storage algorithms to maintain that the data do in fact exist in the state in which they were sent. This distributed database will run in parallel to the publication blockchain discussed above where object hashes are stored. Therefore, when a scientist has a new idea to publish, the network will require them to upload the associated data, manuscript, etc. to this distributed database before they are rewarded their publication token. Thus, all published data will be free and publicly available. The hash for the published object will forever be attached to the publisher’s signature and will therefore be provably theirs. That author will then be attributed whenever that data or idea is cited in the future.

Distributed Funding Mechanisms

Funds for scientific research are generally distributed through government programs, either at the federal, state, or local level. There are some exceptions including privately funded grants as well as crowd-sourced funding; these are generally looked down upon because the science can proceed without peer review. An alternative would be to use smart contracts similar to Gitcoin’s funding mechanism. Here, both private and crowd-sourced funds could be matched to public funds within the context of a conventional peer review. Private donors would choose a proposal that suits them and submit funds to a smart contract. This contract would hold the community funds until a peer-review decision is made on that proposal. If the peer review is successful, those funds would be contributed to the funded grant; if not, those funds would be returned to the donor. In some cases it may be appropriate to use a government-funded matching pool and quadratic funding allocation based on the number of donors, as in Gitcoin’s funding model.

Conclusion

Digital ownership is not true ownership unless it is represented in a trustless way on a public blockchain. If not, digital objects can be copied and redistributed or worse, they can be transformed and distorted. Intellectual property is perhaps the most valuable form of ownership and is no exception. In the current academic system, ownership and scientific incentives are undervalued and taken advantage of by centralized and archaic publishing regimes. Blockchain - and its overhaul of the art market in particular - give insight for an academic future with more fair incentives, self-custody of intellectual property by the thinker themself, and generally a more distributed scientific field which levels the playing field for its participants. Significant work has already been done toward this effort. Proposals have already been laid out for smart-contract-based peer review. Organizations like Bloxberg, ResearchHub, and Planck Manuscripts are already working on implementing similar incentivization mechanisms to that described above. Research fundraising is being coordinated around decentralized groups like VitaDAO. Eventually, all the work that is being done now to secure financial infrastructure on Ethereum will also be used to secure intellectual property and academic infrastructure.