Hidden Einsteins Fund

A fund of Carina Initiatives, with strategic support from Renaissance Philanthropy

A multi-donor fund to find and develop the young people who could push the frontier in math and science, and to build the evidence and public support for government to take that work to national scale.

Why this matters

Scientific progress depends heavily on a small number of people doing frontier work. Children with the ability to become those people show it early. Whether they get there depends on whether a country builds the systems to find and develop them.

A small group drives innovation

1 in 4

inventors comes from the top 5% of third-grade math scorers

New York City public school students.

Source: Bell, Chetty et al.

Family income decides who gets there

2×

as likely to become an inventor for top students from high-income families

6.5 per 1,000 against 3 for top scorers from low- and middle-income families.

Source: Bell, Chetty et al.

Developed talent goes much further

10×

as likely to become an inventor when top students are found early and given advanced work

About 50 per 1,000 for students identified in the top 1% at age 13.

Source: Lubinski & Benbow

China develops talent systematically

China built its K-12 system to develop technical talent and has kept improving it for decades. Its standard math curriculum reaches algebra in seventh grade. It screens for talent through competitions at every age, enrolls its strongest students in specialized schools and classes, and now runs combined undergraduate-to-PhD programs for students as young as 14. The results are showing up at the frontier.

Sources: OECD PISA 2025 · Financial Times · Carnegie China

Math and AI talent, by the numbers
PISA Level 6

Share of 15-year-olds reaching the top level of math (PISA 2025)

30%China*
2%U.S.
Top tier of competition

High school students a year in each country's top national math competition

50,000China National Math League
6,500U.S. (AIME)
Frontier AI researchers

Share of the world's top AI researchers who did their undergraduate degree in each country (2025)

57%China
13%U.S.

*China's PISA results cover Beijing, Shanghai, Jiangsu and Zhejiang.

Xiaotong reading a math textbook at home

A National Math Star

They started calling me ‘the tenth grader in a fourth grade class.’
— Xiaotong, TexasHer family had never heard of National Math Stars when a letter arrived: her score on the state math test had flagged her. In fourth grade, she's now studying geometry and intermediate algebra. Read her story

Why so much talent goes undeveloped

A child with exceptional ability needs three things: to be found, to be taught at their level, and to be surrounded by peers, mentors and opportunities that push them further. In the U.S., this rarely happens.

Universal screening

No one is looking for most of them

Most schools find gifted students through referrals, which miss many talented children from low-income families. Only 12 states require schools to screen every child. When a large Florida district started screening everyone, the share of third graders identified as gifted rose from 3.3% to 5.5%.

Sources: NAGC State of the States (2022–23) · Card & Giuliano, PNAS (2016)

Acceleration

Schools aren't set up to let them move ahead

The top 5% of third graders are about three grade levels ahead in math. Yet only 7% of elementary schools put students in separate math classes by achievement, and only about a quarter of middle schools offer a path two years ahead, such as algebra in seventh grade.

Sources: NWEA norms; RAND (2024); Carina analysis of federal Civil Rights Data Collection, 2021–22

Enrichment

Opportunity clusters in a few places

Most of the programs that develop top math students are private and fee-based, and they're concentrated in a handful of wealthy metros. Two-thirds of the country's top high school math competitors come from the wealthiest counties, home to a fifth of Americans.

Source: Carina analysis of 2025 USAMO and USAJMO qualifiers and Census data

Lilly working with pattern blocks at a National Math Stars event

A National Math Star

Living in a smaller rural community, you really are carving out your own path.
— Lilly's mother, MichiganLilly's district has no gifted program and said there was nothing it could do. With National Math Stars' help, her school now builds time for advanced math into her week. Read her story

What we do

America develops athletic talent systematically and leaves academic talent to chance. We invest in the system math and science need: screening every child, letting strong students move ahead, and surrounding them with peers and mentors.

Our goals by 2031
For each class of students
  • 90%screened for math talent in third grade
  • 30,000the top 1%, with access to acceleration
  • 3,000the top 0.1%, with acceleration and enrichment
Public funding
$100Ma year in state and federal funding to find and develop top math students
$40 million a year for five years, across three tracks

Domestic Talent Ventures

We fund the organizations that do this work in the U.S.: talent search, advanced courses, competitions, camps and research programs.

International Talent Ventures

We take proven models abroad and work with governments to build their own talent systems in public schools, with partners in Brazil, India, Africa and Colombia.

Policy and Research

We fund the research and advocacy that turn these programs' proof points into public policy and funding: state screening and automatic enrollment, federal programs, and portable talent scholarships.

David smiling outdoors

A National Math Star

It’s like I’d been holding in my thoughts, and then suddenly, I could say them out loud.
— David, MinnesotaAt school, other kids rarely understood what he was talking about. National Math Stars connected him with peers like him and a selective summer math camp. Read his story

Results so far

October 2026

NSF brings back the Young Scholars Program

The National Science Foundation is reviving the program that gave about 18,000 talented students advanced math and science before it ended in 1996. The same day, NSF awarded National Math Stars $30 million over four years, matched by private funders.

Two of our grantees were critical here. National Math Stars built a model that works and can scale, and FAI pressed Congress to bring the program back.

Sources: National Math Stars · Bloomberg (via Business Standard) · FAI testimony

  • $30Mfrom NSF to National Math Stars over four years
  • 1.8Msecond to fourth graders to be screened each year
  • 75%of the U.S. population to be covered, up from ten states today

Grantee spotlight

Domestic Talent Ventures

Finds the top students in its partner states using state test data, and gives the strongest a decade of advanced coursework, coaching, camps and family advising.

Biggest wins
  • ~1Mstudents screened across ten states
  • 125full-scholarship Voyager Stars in the class of 2026, with an average household income of $55,000 a year and a median 99.9th percentile score on the CogAT
  • 1.3 SDlearning gains for low-income Stars over similar high achievers
Policy and Research

Researches and briefs federal and state policymakers on how to find and develop America's future scientists.

Biggest win

FAI's national strategy, published in March 2026, now shows up in federal policy:

Congress

“identify and accelerate top-tier K-12 STEM talent”

House Armed Services, FY2027 defense bill
White House

“identify and invest early in high-potential students”

White House science office, FY2028 priorities
International Talent Ventures

A full-scholarship residential school in Kigali for the continent's strongest young mathematicians, opened in 2025.

Biggest wins
  • 14students at the 2026 International Mathematical Olympiad, one year after the school opened: the largest group from any single school in IMO history
  • 12medals for AOA's 16 students at the 2026 Pan-African Mathematics Olympiad (7 silver, 5 bronze). Students joining this fall won 6 more.
Domestic Talent Ventures

Building an operating platform that makes it easy for math professors to start residential math camps.

Biggest wins
  • 4camps on track to launch by summer 2027, two years after the first opened
  • 210students enrolled in summer 2026, nearly doubling the seats at selective residential math camps for middle schoolers
Domestic Talent Ventures

Runs selective online programs in which exceptional teenagers design and carry out original research, with mentors and scholarships.

Biggest wins
  • 510AIME qualifiers applied in 2026, along with 128 international olympiad qualifiers
  • 5 of 20places in Oxford's fully funded Ellison Scholars program went to Non-Trivial alumni
Policy and Research

Supports dissertation fellowships for economics PhD students studying how to find and develop math talent, and hosts the annual Economics of Talent conference.

Biggest wins
  • 14PhD fellows across three cohorts, building a new field of research on talent
  • 2fellows whose work now anchors our own: Ian Calaway on teacher mentors and Phi Adajar on the Tamil Nadu trial

Who makes the grants

  • Josh Taylor

    President

    Carina Initiatives

    Board member, Art of Problem Solving and National Math Stars

  • Jean Chin

    Program Manager, Talent Ventures

    Carina Initiatives

    Board member, National Math Camps

  • Isabelle James

    Program Manager, Policy and Research

    Renaissance Philanthropy

Join the fund

Interested in learning more or joining the fund? We're happy to share our full fund overview. Email josh@carina.fund.

Supporters

The Hidden Einsteins Fund is led and anchored by Carina Initiatives, with strategic support from Renaissance Philanthropy.

For organizations

We find most grantees through our networks and referrals. If your work fits one of the three tracks, you're welcome to write to partnership@carina.fund, though we can't respond to every inquiry.