Introduction: the discrimination problem in traditional grants
Traditional scientific grant systems claim to fund the best research. In practice, they often fund the best-connected, most institutionally protected, and most conventionally legible researchers. This is not always intentional discrimination. Often it is structural discrimination: a system that rewards prestige, institutional affiliation, prior funding, insider language, and committee consensus.
Evidence of unequal grant outcomes is not speculative. A major NIH-linked study found that Asian applicants and Black or African-American applicants were less likely than white applicants to receive NIH investigator-initiated research funding, with Black applicants showing a particularly large gap. Later analysis and commentary found that efforts to close racial funding gaps at NIH had not fully solved the problem. Studies have also found gender-related disparities in grant success, including evidence that men have higher odds of receiving grants in aggregate peer-review grant procedures.
This is exactly the class of problem that AI Internet-Meritocracy, or AIIM, is designed to address.
Internal links:
AI Internet-Meritocracy · Donate to Science · Scientific Funding
What is AIIM?
AI Internet-Meritocracy is a proposed system for funding science and open-source work according to measurable intellectual contribution rather than committee approval.
In a traditional grant system, money is distributed before the work is done, based on a proposal, institutional reputation, CV, letters, and reviewer judgment.
In AIIM, funding would move toward a more continuous, evidence-based model:
| Traditional grants | AIIM |
|---|---|
| Proposal-first | Contribution-first |
| Committee judgment | Algorithmic and AI-assisted evaluation |
| Institutional prestige matters | Public contribution graph matters |
| One-time funding rounds | Continuous funding flows |
| Grant writing skill is rewarded | Scientific dependency and usefulness are rewarded |
| High risk of insider bias | Designed for auditability and non-discrimination |
AIIM does not merely “make grants fairer.” It changes the unit of evaluation. Instead of asking, “Does this committee like this proposal?” AIIM asks, “What scientific work does the global knowledge network actually depend on?”
Why traditional grants become discriminatory
Traditional grants can discriminate through several mechanisms.
1. Prestige bias
Researchers from famous universities often look safer to reviewers. Independent researchers, researchers from weak institutions, and researchers without a standard academic path may be treated as risky even when their ideas are strong.
This is especially harmful for frontier science. Transformative discoveries often look strange before they become obvious.
2. Committee conservatism
Peer-review panels tend to prefer proposals that are understandable, conventional, and close to existing paradigms. This creates a bias against interdisciplinary, foundational, or radically new work.
A 2024 PNAS article on alternative research funding models notes that mainstream grant systems face serious limitations and that alternatives include lotteries, more investigator-based funding, and other structural changes. The existence of these alternatives shows that the problem is not just “bad reviewers”; it is the architecture of funding itself.
3. Career-stage discrimination
Early-career researchers usually have fewer publications, fewer citations, fewer famous collaborators, and less grant-writing experience. A system that claims to evaluate “merit” can therefore end up evaluating accumulated advantage.
4. Identity and network effects
Race, gender, nationality, disability, religion, institutional background, and personal history can all affect how a researcher is perceived. Even when reviewers try to be fair, opaque human judgment is vulnerable to hidden bias.
Research on NIH grant feedback found that women and men may translate review feedback differently into motivation and future reapplication behavior, meaning the grant process can affect not only who wins money, but who stays in the system.
5. Anti-independent-researcher bias
Traditional grants are usually built around institutions. A researcher outside the university system may have no realistic access to serious funding, even if their work is valuable.
This is one of the largest blind spots in modern science funding.
AIIM as a non-discriminatory grant system
AIIM can be described as a non-discriminatory grant infrastructure because it reduces dependence on the social signals that traditional grants overuse.
The system should not ask first:
“Which university employs this person?”
It should ask:
“Which scientific dependencies, citations, replications, applications, and downstream discoveries rely on this contribution?”
That is a more meritocratic question.
How AIIM reduces discrimination
1. It evaluates outputs, not social status
Traditional grants often evaluate promises. AIIM evaluates work, dependencies, and demonstrated usefulness.
This matters because proposals can be shaped by privilege. Strong grant writing, institutional support, recommendation letters, and access to insider norms are not the same as scientific value.
2. It can fund independent scientists
AIIM can support researchers who are outside universities, outside elite networks, or outside grant-writing culture. This is crucial for people who were pushed out of institutions, discriminated against, or simply chose independent research.
A non-discriminatory science system must not assume that universities are the only legitimate source of discovery.
3. It can be audited
A traditional grant rejection is often hard to challenge. The reasoning is private, partial, or expressed in vague reviewer language.
AIIM should be auditable: the ranking logic, dependency graph, citation signals, usage signals, and payout mechanisms should be inspectable. This does not remove all bias, but it makes bias easier to detect.
4. It can reward neglected foundational work
Some scientific work is not immediately fashionable but later becomes essential. Traditional panels may reject such work as too abstract. AIIM can track long-term dependency: if later work depends on an earlier theory, method, dataset, software library, or mathematical structure, the earlier contributor can receive ongoing recognition and funding.
This is especially relevant for foundational mathematics, theoretical science, and open-source infrastructure.
5. It avoids the “single committee bottleneck”
Traditional grants concentrate power in panels. AIIM distributes evaluation across a knowledge graph.
That shift matters. A committee can ignore a discovery. A dependency network cannot easily ignore a contribution that many later works rely on.
AIIM vs traditional grants: the core comparison
| Criterion | Traditional scientific grants | AI Internet-Meritocracy |
|---|---|---|
| Main decision-maker | Human panels | AI-assisted, transparent merit system |
| Main object evaluated | Proposal | Actual contribution |
| Timing | Before research | During and after research |
| Bias risk | High: prestige, identity, institution, networks | Lower if designed with auditability and anti-bias constraints |
| Accessibility | Hard for independent researchers | Open to independent researchers |
| Funding style | Discrete awards | Continuous flows |
| Treatment of unpopular ideas | Often weak | Better if downstream value emerges |
| Accountability | Limited | Potentially public and algorithmic |
| Best suited for | Conventional project funding | Discovery, open science, FOSS, foundational research |
Why lotteries are not enough
Some researchers have proposed partial lotteries for grant funding. Evidence suggests lottery-first or lottery-assisted models may reduce certain costs and increase participation by underrepresented groups. A 2025 study of a German funding line found that a lottery-first approach was associated with increased female representation and lower estimated costs.
But lotteries are still crude. They reduce bias by adding randomness, not by measuring merit better.
AIIM is stronger because it does not simply say:
“When humans cannot decide, randomize.”
It says:
“Build a better evidence system for scientific value.”
Randomness may be useful in some edge cases, especially where proposals are indistinguishable. But the long-term solution is not blind chance. The long-term solution is transparent, dependency-aware, post-publication meritocracy.
Why AIIM matters for donors
Donors often want to support science but face a hard question:
“How do I know my money goes to real scientific value?”
Traditional charities and universities can absorb donations into administration, prestige projects, or politically safe research. AIIM offers a different promise: route funding toward measurable contribution.
That makes AIIM attractive for:
- donors who want to fund independent scientists;
- philanthropists who want measurable impact;
- open science supporters;
- DeSci communities;
- people concerned about discrimination in academia;
- people who believe science funding should be global, transparent, and merit-based.
To support this model, visit:
Donate to Science 🚀
Non-discrimination does not mean anti-merit
A common mistake is to oppose non-discrimination and meritocracy, as if they were enemies.
They are not.
Real meritocracy requires non-discrimination. If a system excludes people because they lack prestige, belong to the wrong institution, have the wrong social background, or work on unfashionable ideas, then it is not meritocratic. It is aristocratic.
AIIM’s claim is simple:
Science funding should follow intellectual contribution, not institutional caste.
This is not “charity for outsiders.” It is a better allocation mechanism for civilization.
The deeper issue: science funding is a governance problem
Grant systems are not neutral pipelines. They are governance systems. They decide which ideas live, which researchers survive, which fields grow, and which discoveries are delayed.
When grant committees are biased, slow, conservative, or prestige-driven, the whole civilization pays the cost.
AIIM reframes science funding as a public, auditable, AI-assisted merit network. That makes it more compatible with the internet age than closed committees built for the bureaucratic university age.
Conclusion: AIIM as the non-discriminatory future of science grants
Traditional grants are not evil. They have funded enormous scientific progress. But they are structurally limited: opaque, committee-centered, prestige-sensitive, and often inaccessible to independent researchers.
AIIM offers a better direction:
- fund actual contribution;
- reduce discrimination by reducing dependence on social status;
- support independent and unconventional researchers;
- reward downstream scientific usefulness;
- make science funding transparent and auditable.
A fair science system should not ask whether a researcher belongs to the right institution. It should ask whether their work advances human knowledge.
That is the promise of AI Internet-Meritocracy.
Read more:
AI Internet-Meritocracy
Scientific Funding Reform
Donate to Science
