R&D: 70% federal in 1951, now ~30%
In 1951, 70% of US R&D was government-funded and 30% private. Today those figures are almost entirely reversed, with the private sector now dominant.
Snapshot · Moonshots with Peter Diamandis
Michael Kratsios on the New Golden Age of American Science | EP #276
Explore episode
Aug 4, 2026
Where this was said
The Wildest Idea in the Golden Age Report: An AI-Native Science Marketplace
At 45:19 · chapter starts 43:20
Peter saves what he calls his favorite idea for last — the section of the Golden Age Report that imagines a fully AI-native scientific marketplace. The vision: funders post bounties for breakthroughs, AI agents identify promising leads and 'hire' autonomous labs to run experiments, results are cryptographically signed and verified, and smart contracts automatically release payment as milestones are met. [1] — Peter Diamandis "The most radical proposal in the Golden Age report: a decentralized scientific marketplace where AI agents post bounties, run autonomous la…" 43:20 Prediction markets guide grantmakers toward the highest-value problems; reputation systems identify reliable agents; human experts remain essential only for judgment calls about which questions matter. The entire system operates at machine speed — continuously, 24/7 — replacing the institutional coordination overhead of committees, journals, and grant cycles with market incentives. Kratsios acknowledges this is blue-sky thinking but insists the technical building blocks now exist and that he genuinely believes it's achievable. Peter frames it as the mechanism that could make Elon Musk's 2036 post-capitalist prediction come true.
A Fully Autonomous AI-Native Science Marketplace: The Wildest Idea in the Golden Age Report
The most radical proposal in the Golden Age report: a decentralized scientific marketplace where AI agents post bounties, run autonomous lab experiments, verify results cryptographically, and release payments via smart contracts — operating at machine speed, 24/7, with human experts reserved only for judgment calls on which questions matter.
Similar snapshots
540M years ago: first animal vision
Animals first sensed light 540 million years ago, triggering an evolutionary acceleration known as the Cambrian explosion within 10 million years.
50% of cortex involved in vision
It is estimated that half of all cortical activity in the human brain is involved in visual function, underscoring vision's central role in intelligence.
Human 6-year-olds learn tens of thousands of object categories
Cognitive neuroscience literature shows that by age 6, humans can recognize tens of thousands of different object categories — far more data than early AI systems were trained on.
~3% of FDA drugs have unclear mechanisms
Only about 3% of FDA-approved drugs have genuinely unclear mechanisms of action, making an unknown mechanism a meaningful early red flag for any compound.
30–50% of phase 1 drugs fail to reach phase 2
Even compounds that clear preclinical testing often fail in humans: 30 to 50% of drugs entering phase 1 trials do not advance to phase 2, frequently because human behavior differs from animal models.
BPC-157 published work: >80% from one group
More than 80% of published BPC-157 research comes from a single academic group whose researchers have IP and commercial interests connected to the molecule, limiting independent replication.
~100 approved peptide drugs; ~150 in trials
Roughly 100 peptide drugs are already FDA-approved, about 150 more are in clinical trials, and 600–700 are in preclinical development, underscoring the legitimacy of peptide science broadly.
BPC-157 founder withheld screening method
The scientist who discovered BPC-157 refused to disclose the screening method used to identify the compound and has never fully published the parent protein sequence — what Peter Attia calls 'scientific trust me, bro.'
Brain barely distinguishes imagined vs real
The brain's visual, auditory, and motor cortices light up nearly identically whether an experience is real or imagined, making visualization a form of actual neural training.
