Kurzweil Scorecard: Longevity Escape Velocity, Ten Years Late

On February 12, 2025, the last mouse in the first Robust Mouse Rejuvenation study died of old age. It marked the completion of a $3.5 million trial in which Aubrey de Grey’s LEV Foundation gave aging mice a four-part combination of telomerase (mTERT), rapamycin, a senolytic, and hematopoietic stem cell transplant, and watched the effects compound. Each therapy extended life on its own. Together they extended it more. That additive result is the clearest evidence yet that aging can be engineered down as a set of damages, not wrestled as a single curse.

In The Singularity Is Near (2005), Kurzweil wrote: “Aubrey de Grey predicts that robustly rejuvenated mice, functionally younger after treatment and with proven life extension, will be demonstrated within ten years” (ch. “Can We Really Live Forever?”). Ten years would have been 2015. RMR1 finished in 2025 — a full decade late. The rest of the batch tells a similar story, but with a twist. Mice are late. Humans are oddly close.

The predictions

This batch collects twelve longevity claims Kurzweil made in 2005. They cluster into three bets: (1) rejuvenation demonstrated in animals in the 2010s; (2) biotechnology translating those wins into human therapies in the 2020s and a “third bridge” of medical nanorobots in the 2030s; (3) biological age stopped and reversed by the late 2030s, letting a twentysomething from 2005 settle in their thirties indefinitely. In The Singularity Is Nearer (2024), Kurzweil restated the throughline: “by around 2030, the most diligent and informed people will reach longevity escape velocity” (ch. 6).

The questions for 2026: Has the animal work happened? Is the human work following? And is it happening through the mechanism Kurzweil described — nanoengineered cell machinery and bloodstream-roving nanobots — or something else entirely?

Where we actually are

Robust rejuvenation in mice. RMR1 reported in 2025 that a cocktail of mTERT, rapamycin, a senolytic, and hematopoietic stem cell transplant additively extended mouse lifespan when administered from middle age. RMR2 was planned for the second half of 2025 with eight interventions across twenty combinations and two thousand mice. Kurzweil’s ten-year timeline was off by a factor of two. The result itself — functional rejuvenation from a damage-repair cocktail — is what he described.

Translation to humans. Kurzweil wrote that “competitive pressure will drive translation into human rejuvenation therapies within five to ten years” of the mouse result. On January 28, 2026 — just over a year after the last RMR1 mouse died — the FDA cleared Life Biosciences’ IND for ER-100, an AAV therapy expressing three Yamanaka factors (OCT4, SOX2, KLF4) that reprograms retinal cells toward a younger epigenetic state. Phase 1 (NCT07290244) enrolls patients with open-angle glaucoma and non-arteritic anterior ischemic optic neuropathy. It is the first human cellular reprogramming trial of its kind. The human work was already underway in parallel with the mouse work; Kurzweil’s “five to ten years after mice” window compressed to one year.

Strategies that reverse each aging process. This is where Kurzweil’s framing has held up best. Aging is now routinely decomposed into mechanistic hallmarks — senescent cells, epigenetic drift, mitochondrial damage, stem cell exhaustion, proteostasis failure — each with its own drug class in development. Patent filings for senescent-cell clearance ran 14–17 per year through the late 2010s and have climbed again: 17 US patents in 2024 and 7 in the first quarter of 2025, led by Unity Biotechnology (29 filings since 2015), the Buck Institute (17), and the Mayo Clinic (17). Partial-reprogramming publications climb faster still: 29 papers in 2025, up from 4 in 2018.

Dasatinib and quercetin. US 11,517,572 (granted December 2022) claims “parenterally administering to the subject an effective amount of dasatinib and quercetin in a course of a daily dosing for a period of 1-7 days” repeated every 0.5–12 months to treat idiopathic pulmonary fibrosis. The claim reads like a prescription. Twenty-five registered trials evaluate senolytics in humans, including Phase 2 studies in Alzheimer’s (ALSENLITE, NCT04785300), osteoporosis (NCT06018467), sepsis (STOP-Sepsis, NCT05758246), and accelerated aging in mental disorders (NCT05838560). These are outcomes trials, not first-in-human safety probes.

Unity Biotechnology. In April 2025, NEJM Evidence published Phase 2 BEHOLD results showing that local senolytic clearance in the retina produced long-lasting vision improvements in diabetic macular edema. Two months earlier, the Phase 2b ASPIRE study had missed its primary endpoint and Unity’s stock fell roughly 30 percent. In September 2025, shareholders voted to liquidate. The first pure-play senolytic company folded even as the science it seeded continues at Mayo, Buck, and elsewhere.

Rapamycin and epigenetic age. The PEARL trial, a 48-week, 114-participant randomized study, reported in 2024 that 5 mg and 10 mg weekly compounded rapamycin was safe; women on 10 mg showed improvements in lean mass and self-reported pain, men in bone mineral content. The compounded formulation was about 3.5× less bioavailable than commercial rapamycin, so the effective doses were closer to 1.4 mg and 2.9 mg. Modest. Not radical life extension. Not “settling at your thirties.” Separately, the Intervene Immune TRIIM trial (rhGH + metformin + DHEA) reported a GrimAge reversal of about 2.5 years after one year; a 2021 diet-and-lifestyle pilot (Fitzgerald et al., cited 329 times) reported a 3.23-year reduction in Horvath DNAm age in eight weeks. Epigenetic age is measurably reducible in humans. The reductions are 2–3 years, not 40.

Nanoengineered replacement of the cell nucleus. Kurzweil described “a nanoengineered replacement for the cell’s genetic machinery” that would “eliminate the accumulation of DNA transcription errors” (ch. “Upgrading the Cell Nucleus with a Nanocomputer and Nanobot”). Nothing in the 9.3-million-patent corpus describes a nanoscale cell-nucleus replacement. What is being built is CRISPR base editing, prime editing, and AAV-delivered reprogramming factors that achieve roughly the predicted outcome — controlled expression without correcting every DNA error — through an entirely different mechanism.

Bloodstream nanobots. In The Singularity Is Nearer, Kurzweil conceded the timeline while insisting on the mechanism: “At some point in the 2030s we will reach this goal using microscopic devices called nanobots” (ch. 6). No in-vivo medical nanorobot has been granted in the US patent corpus. The closest analog — a Stanford-Michigan State nanoparticle that targets macrophages in atherosclerotic plaque — is a lipid-shelled small molecule carrier, not a diamondoid robot with onboard sensors and manipulators. The replacement technologies (CAR-T, ADCs, lipid nanoparticles, viral vectors) are biological and chemical, not mechanical.

Telomere extension, cancer-synchronized. Kurzweil wrote that “orderly telomere-extending therapy for normal cells will be feasible and can be coordinated with cancer therapy by halting it during periods of cancer treatment” (ch. “Overcoming Cancer”). Patent filings mentioning telomerase activators and telomere-extension therapy have averaged one to three per year since 2005 — a flat line. There is no approved telomerase-activator therapy. The cancer synchronization Kurzweil imagined remains hypothetical because the base therapy doesn’t exist.

The scorecard

What Kurzweil missed, and what he nailed

Kurzweil’s hit rate on direction is unusual for 2005. Aging as a set of damages rather than a singular process. Compounding interventions that extend lifespan additively. Translation from animal to human in tight succession. Senescent cells as a target. Transcription-level control of cell identity. All of it is in the book. All of it is in a clinical trial or a granted patent twenty years later.

His miss on mechanism is equally consistent, and it is the interesting pattern. Every prediction in this batch that involves nanotechnology — diamondoid nanobots, nanoengineered nuclei, respirocytes, in-vivo cellular repair machines — is wrong. Every prediction in this batch that involves biology — reprogramming, senolysis, gene therapy, stem cell transplant — is roughly on track. Where Kurzweil imagined mechanical nanoscale machines, the field built molecular biology. The outcomes he sketched are arriving; the tools are unrecognizable.

The other miss is timing. The 2005 book assumed that once a capability became feasible it would be deployed fast. The 2024 Nearer walks that back: escape velocity now arrives in the 2030s, not the 2020s. Mice were rejuvenated in 2025 because RMR1 ran from 2023 to 2025 — a linear, budgeted, single-cohort experiment that cost $3.5 million and took as long as it took. The biology does not care about the curve.

For an R&D director reading this in 2026, the operating implication is concrete. The senolytic story is a Phase 2 story with one company folded and six others shipping trials. Partial reprogramming is a first-in-human IND. The window between “speculative” and “competitive landscape” has narrowed to one or two years in this field.

Method note

This scorecard was built by searching 9.3 million US patents and 357 million indexed scientific papers for longevity, senolytic, reprogramming, telomere, rapamycin, and nanorobot terms; pulling granted claim text for specific patents; cross-checking 25 registered clinical trials on ClinicalTrials.gov; and using web research to verify company status, trial outcomes, and 2024–2026 news. Every patent number, trial identifier, and publication year in this post was verified against those sources this week. Direct quotes from Kurzweil are from The Singularity Is Near (2005) and The Singularity Is Nearer (2024).