Why Fundamental Mathematics Cannot Be Funded Like a Startup

Getting your Trinity Audio player ready...

Fundamental mathematics cannot be funded like a startup because the two activities operate under fundamentally different economic conditions. A startup is expected to identify customers, develop a marketable product, demonstrate growth, and eventually produce a financial return. Fundamental mathematics may produce no commercial product, no revenue, and no identifiable customer—even when it creates knowledge that later becomes indispensable to science and technology.

This does not mean that mathematical research lacks economic value. It means that its value is often indirect, widely distributed, difficult to predict, and impossible for its original creator to capture.

A funding system that treats mathematicians as startup founders will therefore favor projects with persuasive applications, short timelines, and commercially legible narratives. It will systematically neglect the abstract definitions, proofs, theories, and conceptual frameworks from which many future applications eventually emerge.

The Core Difference: Startups Capture Value, Mathematics Creates Shared Knowledge

A venture investor finances a startup because the company may capture part of the value it creates. The company can sell subscriptions, license technology, charge transaction fees, own patents, or build a defensible commercial platform.

Fundamental mathematics usually creates a public intellectual resource.

Once a theorem is published, other researchers can use it. Once a definition becomes standard, it may organize an entire field. Once a mathematical language is developed, scientists, engineers, programmers, and companies may apply it without paying the mathematician who created it.

This produces a severe value-capture problem:

Fundamental mathematics can generate enormous social value while generating almost no revenue for the person who produced it.

A theorem may improve cryptography, computation, physics, logistics, or engineering. However, the commercial gains usually appear far downstream, distributed among organizations that combine the theorem with many other inputs.

The mathematician cannot normally issue invoices to every future beneficiary.

This is why fundamental mathematics resembles a global public good more than a venture-backed product. World Science DAO similarly treats basic research as something that requires transparent public or philanthropic support, rather than assuming that every important discovery will produce a viable company.

Startup Funding Requires a Plausible Path to Revenue

Startup finance is built around questions such as:

• Who is the customer?
• What problem will the product solve?
• How large is the market?
• How will the company make money?
• What prevents competitors from copying it?
• Can the company scale?
• When might investors receive a return?

These are rational questions when evaluating a commercial enterprise. They are frequently inappropriate when evaluating fundamental mathematics.

A researcher investigating a new algebraic structure may not know whether it will have an industrial application. A new theorem in topology may remain apparently “pure” for decades. A foundational definition may be valuable because it makes future thought possible, not because it immediately solves a customer’s problem.

Demanding a commercial roadmap in such cases does not eliminate uncertainty. It rewards researchers who are best at disguising uncertainty.

The result is not necessarily more useful mathematics. It may simply be more speculative marketing.

Mathematical Importance Is Often Visible Only Retrospectively

Startup investors attempt to predict future demand. Fundamental mathematics frequently changes what future researchers are capable of demanding.

This distinction is crucial.

A mathematical discovery can create:

• a new field of inquiry;
• a language for expressing previously disconnected problems;
• a proof technique that later becomes reusable;
• a classification framework;
• a connection between existing theories;
• an abstraction that makes software or formal verification possible;
• or a solution to a problem whose practical importance has not yet appeared.

Before such a contribution exists, evaluators may lack the concepts needed to understand its full importance.

The OECD’s Frascati Manual defines basic research as work undertaken primarily to acquire new knowledge without a specific application or use in view. That absence of a predetermined application is not a defect. It is one of the defining characteristics of basic research.

A funding model that requires researchers to specify commercial outcomes in advance therefore selects against a central feature of fundamental inquiry.

Long Time Horizons Break the Venture-Capital Model

Venture capital normally operates within a bounded investment horizon. A startup is expected to demonstrate progress through product releases, user acquisition, revenue, partnerships, or other measurable milestones.

Fundamental mathematical research may develop on a much longer and less predictable timeline.

A contribution can pass through several stages:

1. A researcher introduces an unfamiliar definition.
2. A small group studies its formal properties.
3. Connections to other fields gradually emerge.
4. Expository work makes the theory accessible.
5. Software or formal libraries implement parts of it.
6. Scientists or engineers discover applications.
7. Commercial value appears much later.

No single stage may look like exponential startup growth. Yet the completed chain can transform an area of knowledge.

The time between mathematical discovery and major application may exceed the lifetime of a startup fund. Even when commercial applications eventually appear, the original research may have been too early, too general, or too open to support a proprietary company.

Fundamental Mathematics Has No Reliable Product-Market Fit Test

A startup can test product-market fit by releasing a product and observing whether people use or buy it.

Mathematics has no equivalent early test.

Researchers can examine whether a proof is correct, whether a definition is coherent, whether a construction is nontrivial, and whether a result connects to existing work. But these tests establish mathematical merit—not future economic demand.

Citation counts are also insufficient. A concept may initially receive few citations because:

• it is difficult to understand;
• it challenges established terminology;
• the relevant community is small;
• applications require additional theory;
• the researcher lacks institutional visibility;
• or later researchers use the concept without tracing its full intellectual ancestry.

Conversely, a fashionable but incremental paper may receive many citations quickly.

Fundamental mathematical value is therefore multidimensional. It may involve correctness, originality, explanatory power, generality, depth, fertility, simplification, dependency, or future applicability. No startup metric captures all these dimensions.

Failure in Mathematics Is Not the Same as Startup Failure

Most startups fail commercially. Venture portfolios accommodate this by expecting a small number of exceptional successes to compensate for many losses.

It may seem that the same portfolio logic could fund mathematics. To a degree, it can. Research funders should tolerate uncertainty and distribute resources across multiple approaches.

But mathematical “failure” has a different structure.

An unsuccessful attempt to prove a theorem may still produce:

• a useful lemma;
• a counterexample;
• a classification of failed approaches;
• a new conjecture;
• improved notation;
• computational data;
• research software;
• a corrected definition;
• or a clearer understanding of the problem’s difficulty.

These outputs may help other researchers even when the original objective is not achieved.

Startup accounting tends to classify a company that does not generate sufficient returns as a failed investment. Scientific accounting should instead identify and reward the reusable knowledge produced along the way.

That is one reason failed research results can function as public goods: they can prevent duplicated effort and redirect future investigation.

Patents Cannot Solve the Incentive Problem

One proposed solution is to make mathematical research more proprietary. Researchers might be encouraged to patent applications, create companies, or restrict access until commercialization is possible.

This approach is unsuitable for much fundamental mathematics.

Pure mathematical theorems are generally not protected as proprietary inventions merely because they are true. More importantly, imposing artificial exclusivity on foundational knowledge would reduce one of mathematics’ greatest advantages: unrestricted recombination.

Mathematical progress is cumulative. Definitions depend on earlier definitions, theorems depend on lemmas, and theories absorb tools from many fields. Restricting these dependencies can slow the entire network.

The objective should not be to force mathematics to imitate proprietary software. It should be to fund the production and maintenance of openly reusable intellectual infrastructure.

Startup Logic Distorts Which Questions Researchers Ask

When funding depends on commercial presentation, researchers receive a predictable signal: choose problems that can be made to sound investable.

That may shift attention toward:

• fashionable technologies;
• short-term applications;
• easily demonstrated prototypes;
• projects connected to large existing markets;
• and work that can be described using familiar commercial categories.

Meanwhile, researchers may avoid:

• foundational reconstruction;
• obscure but structurally important questions;
• long monographs;
• negative results;
• difficult verification work;
• mathematical libraries;
• conceptual unification;
• and theories without immediate applications.

This creates a selection bias. The funding system does not discover which mathematics is most valuable. It discovers which mathematics most closely resembles a startup pitch.

The distinction is especially important for independent researchers. They may possess valuable ideas but lack institutional prestige, grant-writing support, business networks, or the ability to construct a persuasive commercialization story.

Mathematics Still Needs Accountability

Rejecting startup finance does not imply that mathematical researchers should receive money without evaluation.

A legitimate funding system should still ask:

• Is the work mathematically coherent?
• Are the claims supported by proofs?
• Is the contribution genuinely new?
• Does it clarify, unify, extend, or correct existing knowledge?
• Can independent experts inspect it?
• Are intermediate outputs publicly available?
• Does the research depend on earlier contributions that should also receive recognition?

The mistake is not evaluation itself. The mistake is evaluating mathematics primarily through expected revenue, market size, or commercial scalability.

Scientific accountability should focus on verifiable contribution, not simulated product-market fit.

Better Funding Models for Fundamental Mathematics

No single mechanism can finance every kind of mathematical work. A resilient system should combine several models.

Long-term institutional support

Universities and research institutes can provide salaries, libraries, colleagues, seminars, and intellectual continuity. This enables researchers to pursue questions that cannot be compressed into short grant cycles.

The weakness is that institutional positions are scarce and may be influenced by prestige, geography, credentials, or academic politics.

Investigator-centered grants

Instead of requiring precise predictions, funders can support capable researchers or small teams over extended periods. This acknowledges that the direction of fundamental inquiry may change as discoveries are made.

Portfolio funding

Funders can distribute resources among many approaches rather than trying to select one guaranteed winner. Some projects will produce major results, while others will produce smaller but reusable contributions.

Prizes

Prizes can recognize demonstrated achievements. However, conventional prizes are highly concentrated and often arrive after the researcher has already overcome the most difficult financial period.

Retroactive and continuous rewards

A more flexible model can reward contributions after they become visible while paying at a much finer granularity than traditional prizes.

The AI Internet-Meritocracy research-funding model proposes evaluating published contributions and distributing funding according to demonstrated merit. Researchers would not need to promise that a theory will become a startup. They could instead receive rewards for producing verifiable mathematical value.

AIIM is also described as combining some properties of grants and prizes: it can recognize completed work while continuing to support contributors as evidence accumulates.

Dependency-aware funding

Mathematics is not a collection of isolated winning papers. It is a dependency graph.

A major theorem may depend on:

• an earlier definition;
• a technical lemma;
• a counterexample;
• a software library;
• a formalized proof;
• an expository text;
• or a theory developed by another researcher decades earlier.

Funding should therefore flow not only to the most visible final result but also to the intellectual infrastructure that made it possible.

A dependency-aware system could divide recognition among contributors rather than assigning all value to the last person in the chain. This is one of the central reasons to evaluate research outputs and their relationships rather than relying only on conventional grant proposals.

Could Mathematicians Create Startups?

Yes. Some mathematical research can support commercial ventures.

Mathematicians may build companies involving:

• cryptography;
• optimization;
• quantitative finance;
• artificial intelligence;
• verification;
• scientific computing;
• cybersecurity;
• logistics;
• or specialized software.

Such companies can legitimately use startup finance.

But financing a commercial application is not the same as financing the underlying mathematical discipline. Startup investment will support the subset of mathematics from which an investor may capture revenue. It will not reliably support the broader conceptual ecosystem upon which those applications depend.

The correct conclusion is not that startups and mathematics are unrelated. It is that startup funding is a selective downstream mechanism, not a complete funding system for fundamental research.

The Proper Economic Analogy Is Infrastructure, Not a Startup

Fundamental mathematics is closer to infrastructure than to a commercial product.

It supplies languages, structures, proofs, methods, and constraints used across many domains. Like physical infrastructure, its benefits spread beyond the organization that initially pays for it. Unlike physical infrastructure, mathematical knowledge can be copied globally at negligible cost without being depleted.

This combination makes mathematics extraordinarily productive but commercially difficult to finance.

The more foundational a mathematical contribution is, the less likely its full social value can be captured by a single investor. That is precisely why public, philanthropic, institutional, and decentralized funding remain necessary.

Conclusion

Fundamental mathematics cannot be funded like a startup because it does not reliably produce customers, proprietary assets, rapid growth, or predictable financial exits. Its outputs are open, cumulative, uncertain, and often useful only after long delays.

Forcing mathematicians to behave like startup founders would not make mathematical discovery more efficient. It would make funding more dependent on commercial storytelling and bias research toward short-term, market-readable projects.

A better system should combine long-term support, diversified funding, expert evaluation, continuous rewards, and dependency-aware attribution. It should finance mathematical knowledge according to the value it contributes to humanity’s shared intellectual infrastructure—not according to whether it can be converted into a pitch deck.