Challenges: Trivial, Grand, and Whitehead
Introducing the Protocol Institute Challenges program
One of the problems I’ve been thinking the hardest about is how to catalyze and incentivize fresh efforts around protocol research problems and ideas, both old and new.
Too often, emerging fields witness a rush of what we might call repackaged creative energy, where people see an opportunity to wrap their favorite stalled old problems or ideas in shiny new packaging, to try and give them a new lease on life, with no substantive changes in the framing of the problem or methods of attack. This is rarely a matter of bad faith or cynical pursuit of funding opportunities. It is simply genuinely hard to get past cosmetic connections and resonances between the old and new, let go of your favorite questions, and really see through new eyes.
Protocol Studies is particularly prone to this syndrome since you don’t have to squint too hard to see almost anything as a protocol problem. This is one reason we spend so much time and energy refining our protocol-pilling techniques. It’s not enough to get you interested in protocols. That’s easy. We have to get you reflexively and natively seeing through and thinking in protocols before we can really tap into your talents. We have to get you literate in Protocolese. As Helena Rong from our 2025 cohort put it, we have to try and make Protocolese everybody’s second language.
How can we get people to truly transpose familiar challenges to an actually generative protocol-thinking register that unlocks lines of thought that were not accessible before? And how can we get people to see entirely new challenges that were invisible before the protocol lens became available?
This is a hurdle for our third Protocol Symposium in September – the first edition with an open call for proposals. Abstracts for talks and workshops are due Sunday June 14, by the way.
I’ve been wondering how to induce a good mix of programming that reframes familiar old problems in weird new ways as protocol problems, as well as programming that tackles new ideas and problems that fundamentally aren’t visible at all without the protocol frame. And we’d like to try and catalyze this kind of activity in a sustainable way that can fuel both particular events like this, and ongoing work.
We have the beginnings of a solution prototyped. We hope it will inspire at least a few novel proposals for the Symposium – our new Challenges program.
The Challenges Program
I’d like to introduce (drumroll!) the Protocol Institute Challenges program.
The linked page is currently a list of a dozen or so challenges inspired by open threads in our archives, ranging from trivial to grand to the grandest kind on our scale – Whitehead challenges. These challenges are rated in difficulty using a 9-point Fibonacci sequence scale (1, 2, 3, 5, 8, 13, 21, 34, 55) which will be familiar to those who have used the planning poker model in software development.
Anyone can indicate their interest in seeing a challenge tackled by clicking the 👀 emoji, which increments the collectively assessed value of the problem (distinct from its difficulty) using a cunning quadratic weighting scheme. Votes from logged-in members will count for more.
Currently only PI members can pose challenges, but anyone can get a problem onto the list by persuading a member to post it on their behalf. If you’re an alum of any Summer of Protocols program since 2023, you’re eligible for membership. Just apply using the link on the top right (use whatever email you used for your Summer of Protocols participation so we can correctly detect and approve your membership request). If you’re new to PI, and would like to become a member, first participate in any upcoming programming, such as the Symposium, or one of the SIGs.
Perhaps the most important feature is that – unlike traditional challenge programs – Protocol Institute Challenges are generally going to be open-ended and inexhaustible by default, which means there will likely not be a single dispositive “solution” that solves a challenge once and for all.
This logic is inherited from Frank Chimero’s classic essay, Only Openings, which is a core reference for Protocol Studies, and also the focus of the very first posted challenge – the Chimero Protocol challenge. For fans of James Carse’s finite and infinite game model, we’re trying to build our challenge program around infinite game principles. The goal is not to win, but to continue playing.
At the moment, this is a beta program, and we have not yet attached any funding to it. We’ll begin exploring funding models and seeking sponsors shortly. There is a prize purse for the Symposium though, so if one of the challenges interests you, propose a talk on it for the Symposium.
This design for our Challenges program is inspired by many existing models, and the logic is explained below.
Program Logic
The design of our challenges program draws from three precedents: Big challenge programs, small challenge programs, and crowdfunding programs.
We’ve also added tweaks of our own, in particular to accommodate the nature of humanities and social science research, which don’t quite fit the challenge program models designed for STEM research.
Big Challenge Programs
The first category of precedents includes mechanisms like X-prizes and Grand Challenges, which offer big rewards, running to millions of dollars, for obviously difficult and important problems. The challenges tend to have a sort of on-the-nose quality to them. These can also take the form of singular challenges like Kennedy’s moonshot challenge.
Self-driving cars famously emerged from this kind of mechanism. DARPA pioneered Grand Challenges that lend themselves to precise formulation in terms of a real-world engineering challenge, which is also characteristic of X-prizes.
The nice thing about big challenge programs is that they have a built-in charismatic appeal that speaks to a broad audience, and typically offer ways for almost anyone to get involved, from amateur to expert level. They are also typically self-promoting, since they offer some amount of spectacle and drama.
The downside of big challenge programs is that most profound and radical breakthroughs tend to start in small, unremarkable ways on the margins of mainstream attention. It is obvious that a moon base would be an interesting accomplishment, and it is also obvious that it is a hard problem. It was less obvious early on that mRNA science would solve a pandemic. Many of the major discoveries in the history of science – x-rays, penicillin – would not have yielded to grand challenge type mechanisms.
Big challenge programs also tend to rely too much on big money as an incentive, and not enough on curiosity, unorthodox perspectives, and idiosyncratic personal visions.
Small Challenge Programs
The second category is mechanisms like Erdős challenges, which frame and pose problems whose significance and economic value may not be immediately obvious even to experts, but which direct serious attention and resources based on the intuitions and credibility of individuals whose judgment and taste are trusted by a broader community.
OpenAI’s recent solution of an Erdős problem made big news not because there are big prizes attached to Erdős problems, but because of the bragging rights attached to them. The savant mathematician Paul Erdős for whom the problems are named, and who originally posed them, mostly offered small personal bounties. The vast majority featured bounties in the $0–$500 range, and only 13 were in the $1000–10,000 range. Mathematicians vie to solve Erdős problems for the same reason they are proud of a low “Erdős number” (your degrees of separation from Erdős as measured by co-authorship links; I’m not a mathematician but my Erdős number is 4 via my late PhD advisor) – the chance to BIRG in the light of acknowledged genius.
The problem solved by GPT was worth only $500. A notable feature of Erdős problems is that the value of the prize often does not correlate at all with the difficulty of solving it.
This is a feature, not a bug. The low financial stakes and the lack of its correlation to practical value makes the contest about genuine curiosity and nerd-energy, rather than mercenary motives. Even an acknowledged genius like Erdős did not necessarily have perfectly calibrated mathematical instincts, and one of the reasons Erdős-type frontiers of discovery are so valuable is that small, apparently irrelevant and unimportant starting curiosities and questions can lead on to profound discoveries and advances. And seemingly big and important challenges might turn out to be meaningless once solved.
Some programs, like the Clay Millennium problems prizes, fall in between big and small challenge categories. There is big money attached (a million dollars), but most of the problems are not obviously important except to mathematicians working in the area (P=NP and the Navier-Stokes problems are the exceptions, since they both have recognized engineering significance). There can also be programs with institutional levels of gravitas, but with no prizes attached, such as Hilbert’s problems.
Small challenge programs have the benefit of reflecting the unique tastes and judgments of unusual individuals, tapping into loftier motivations, and directing attention to margins and weak signals.
The downside of small challenge programs is that they may not be able to marshal the resources required to work on the subset of challenges that require them, and can also descend into the solipsism of a taste-making elite.
Crowdfunding Programs
The third kind of mechanism is the youngest, and originated on web platforms like Kickstarter and GoFundMe. In their original form, however, they were not tailored to the particular needs of drawing attention and resources to research-like challenges. They were generic mechanisms, equally applicable to research and creative domains on the one hand, and to charitable causes and personal needs.
Two innovations in the blockchain sector build in features that make crowdfunding particularly intelligent, and better suited for challenge programs.
The first is quadratic funding: A mechanism for turning the wisdom of the crowds into a kind of mimetic collective intelligence that can function as a substitute for the taste and judgment of individual geniuses. It takes the syndication approach used in many angel-funding models to another level using mathematics rather than personal influence. In our beta design, we’ve taken some cues from Gitcoin, one of the pioneers of this approach.
The second is the idea of retroactive public goods funding (RPGF), pioneered by the Optimism RPGF model. In this model, funding is provided for work already done, with demonstrable value for the commons. We ourselves have been a beneficiary of this model, having won a grant in the first year of the Summer of Protocols program. The unique strength of retroactive models is that they incentivize work whose value is too illegible or obscure in the beginning, and only apparent to a lonely minority.
We haven’t yet built a retroactive element into our model, but we are carefully thinking the matter through.
The great advantage of crowdfunding models is that they achieve scale through aggregation, and creative insight through composition of diverse and pluralistic viewpoints. To the extent they work, you have less need to rely on big, institutional sources of funding, or the taste and judgment of individual geniuses.
The downside of crowdfunding models, as with all market-like models, is vulnerability to various kinds of complex capture phenomena, and the faddish and capricious tendencies of crowds.
Gaps in the Model
I hope it’s obvious from the description earlier, as well as evident from the live challenges page, how we’re approaching our program design problem. We think we’ve borrowed the most useful elements from big, small, and crowdfunded challenge program models. But the design is as yet incomplete, which is why it’s prominently tagged beta.
The biggest gap (besides plumbing to funding) is coverage of humanities and social science challenges. As it has emerged, the three-year-old field of Protocol Studies is firmly situated right in the infamous gap between the Two Cultures described by C. P. Snow, with Balrogs lurking in the chasm below. Our archives contain everything from technical specifications for speculative new protocols, to industrial engineering ideas, to avant-garde art and fiction.
Historically, the humanities have relied more on personal patronage and the tastes of institutional gatekeepers than on structured programs and competitive formats that rely on objective criteria to gauge progress against well-posed challenges. More recently, state patronage (with risks of being co-opted into propaganda) and aggregated micropatronage on web platforms (with risks of audience capture) have been the default.
One of our meta-challenges is to develop challenge mechanisms suited to the humanities and social sciences into our model that mitigate some of the problems with historic mechanisms.
The Symposium
The 2026 Protocol Symposium will be the first real test of our ability to curate exploration and investigation activities in an open-play mode. In previous years, we had the advantage of working with known people and materials. This year, everything is wide open. We have some early submissions in, but it’s too early to tell what sort of programming we’ll end up with.
If you like our mission, this is your chance to get in on the ground floor and help shape it. Check out our original call for abstracts last week, then head over to the call page and put in your talk and workshop proposals. That deadline, again, is Sunday June 14.
And of course, if you need inspiration (or have inspiration to share), check out the Challenges page.