The plan to end aging

Plan A · Repair

The plan to end aging.

Aging is the common risk multiplier behind most chronic disease. Repair it at the system level and most of the diseases downstream lose their biggest driver. This is Omri Amirav-Drory’s plan, in the order he gives it.

Start reading ↓ Replay the intro
Pencil sketch of one woman walking toward the viewer at five ages, younger each time, with blue line-art of cells, a heart and a liver flowing between them

01 Why aging is the target

One root cause. Many diseases.

Heart disease, cancer, stroke, dementia, diabetes and organ failure look separate, but age raises the risk of nearly all of them.

One biology→Many diseases→One lever

9 of the 10 leading causes of death are due to aging.

  1. 01Heart disease
  2. 02Cancer
  3. 03Accidentsnot aging
  4. 04Cerebrovascular disease
  5. 05Chronic lower respiratory disease
  6. 06Alzheimer’s
  7. 07Diabetes
  8. 08Kidney disease
  9. 09Liver disease
  10. 10COVID-19
Strategic implication

Treat aging upstream, not each disease downstream.

Chance of death by age

The curve everyone is on.

90year old1 in 6chance of death this year
0%4%7%11%14%18% 0102030405060708090 Age (years) Chance of death per year

Chance of death per year. Source: @statto, as shown in the deck. Hover a point or pick an age.

Negligible senescence

Human vs. mole rat.

0%1%2%3%4%5% Age (years) Chance of death per year Humans Mole rats

The mole rat’s risk of dying barely changes with age. Ours climbs. Same basic biology, different control.

Negligible senescence

Some animals barely age. The naked mole rat is one of them.

Its chance of dying in a given year stays about the same from youth to old age. For humans that chance climbs steeply after 65. The point of the comparison is that a flat curve is biologically possible. The question is what controls it.

02 The race to 160

Two doublings mostly prevented early death. The third requires repair.

20 → 40 → 80 is life expectancy at birth, not maximum lifespan. The verified record is 122. Getting the next doubling means fixing damage, not only avoiding it.

  1. ~20 Hunter-gatherer world Survive infection, injury and childbirth
  2. ~40 Industrial revolution Sanitation · nutrition · safer birth
  3. ~80 Modern life Vaccines · antibiotics · chronic care

Verified record

122

The longest confirmed human life. Life expectancy at birth has doubled twice. The maximum has barely moved.

Beyond 160

Repair improves faster than new damage.

Longevity escape velocity. The deck labels this a hypothesis, and so do we.

The destination

160 is not the destination. It is where aging becomes maintainable.

03 The ABCD plan to end aging

Only Plan A is designed to outrun aging.

Plan D accepts death. Plan C waits. Plan B fights failures one by one. Plan A repairs the system.

A Repair Replacement + Regeneration + Reprogramming B Bioengineering Fight failures one by one C Cryonics Preserve and wait D Death Accept the endpoint
The destination

Repair is the path to longevity escape velocity.

04 Plan A

Three levers. One maintenance system.

Reset what is salvageable. Replace what is irreparable. Regenerate what is missing. Pick a lever to see what it means, then read the three chapters that follow.

Reprogram Reset cell state Reset what is salvageable Replace Swap the component Replace what is irreparable Regenerate Rebuild in place Regenerate what is missing
Deck illustration: a human figure in the center with three circled levers around it. Reprogram, an old cell becoming a young cell. Replace, a worn heart swapped for a new one. Regenerate, scar tissue rebuilt into healthy tissue.
From the deck. Reset what is salvageable · Replace what is irreparable · Regenerate what is missing.

05 Lever one · Reprogram

Yamanaka factors reset cell identity.

Four transcription factors reopen the pluripotency program. Deliver them and an adult cell walks back toward a stem-cell state. The trick is deciding how far to let it go.

  1. 1 Adult somatic cell A differentiated cell with its identity set.
  2. 2 Deliver OSKM OCT4SOX2KLF4c-MYC
  3. 3 Epigenetic reset Somatic program off. Pluripotency network on.
  4. 4 iPSC Self-renewal + many cell fates

Partial vs. full reprogramming

How long do you leave the factors on?

A brief OSK pulse moves an aged cell to a younger state and keeps it the same cell. Continue to completion and you get an iPSC, which can become many cell types but has forgotten what it was.

Partial reprogrammingSame cell. Younger state.Brief OSK pulse, stopped before pluripotency.

iPSCSelf-renewal and many cell fates.Full reprogramming erases the original cell identity.

Deck illustration of the four steps: adult somatic cell, deliver OSKM, epigenetic reset, iPSC with many cell fates
From the deck. Stop early → partial reprogramming. Continue to completion → iPSC.
Deck illustration of an aged neuron receiving a brief OSK pulse and returning as the same cell in a younger state, with the pluripotency route marked stop before pluripotency
From the deck. Aged cell → brief OSK pulse → same cell, younger state.

06 Lever two · Regenerate

Biology already knows how to rebuild.

Shared genes. Different control. Many species regenerate complex tissues on their own, and humans still run parts of that program.

  • HydraWhole-body renewal
  • PlanarianBody regeneration
  • AxolotlLimb regeneration
  • ZebrafishHeart repair
Deck illustration: hydra, planarian, axolotl and zebrafish, each with the tissue it regrows, and a human outline at the end sharing the same DNA strand
From the deck. Hydra, whole-body renewal · Planarian, body regeneration · Axolotl, limb regeneration · Zebrafish, heart repair.

The human program is still there

Capacity exists. It is uneven, and it declines with age.

  • Development builds
  • Skin renews
  • Bone repairs
  • Liver restores mass
  • Heart scars

Development builds every organ from scratch. Skin and bone still repair. The liver restores its mass. The heart scars. Same body, five different answers to the same question.

Deck illustration: development builds organs, skin renews, bone repairs, liver restores mass, heart scars
From the deck.

The same injury can end two ways

Tissue damage, then a decision.

Fibrosis

  1. Inflammation
  2. Collagen
  3. Lost function

The default human answer in most adult tissues. The wound closes, the tissue does not come back.

Regeneration

  1. Resolution
  2. Patterning
  3. Restored function

The axolotl’s answer. Also the answer in a human liver, and in a child’s fingertip.

The therapeutic target

The therapeutic target is the decision process.

1 Nature’s proof

Many species naturally regenerate complex tissues.

Axolotl forelimb: stump → early blastema → patterning and outgrowth → patterned limb. Zebrafish heart: injury → cardiac progenitors activate → tissue restored.

2 Humans retain parts

We can naturally repair and renew many tissues.

  • Liver
  • Skin
  • Bone
  • Muscle
  • Fingertip

3 Therapeutic hypothesis

We can mobilize regeneration where it is currently limited.

  • Immune + fibrosis control
  • Progenitor activation
  • Positional / bioelectric cues
  • Nerves + matrix support

Scar→Regeneration

Local · Transient · Self-limiting

Deck illustration in three panels: nature’s proof with axolotl limb and zebrafish heart regeneration, the human tissues that retain repair, and the therapeutic hypothesis of an organized blastema-like repair zone
From the deck. Scar → regeneration: local, transient, self-limiting.

07 Lever three · Replace

Grow a young part. Swap it in.

When repair is insufficient, replace the failing biological component. From cell engineering to functional replacement.

  1. 1Aged / damaged tissue
  2. 2Patient cells → iPSCs
  3. 3Cells · tissues · organs
  4. 4Transplant → restored function
Deck illustration: an aged figure with a damaged liver, patient skin cells becoming iPSCs, then cells, tissue and a new liver transplanted into a restored figure
From the deck. Aged tissue → patient cells → iPSCs → cells, tissues, organs → transplant → restored function.

Replacement · Xenotransplantation

Pig organs engineered for human recipients.

69 edits in the eGenesis kidney platform. Three kinds of change, one goal: an organ the human immune system will tolerate.

  1. 0 Pig antigen genes disabled Reduce antibody recognition.
  2. 0 Human genes added Improve immune and blood compatibility.
  3. 0 Viral copies inactivated Reduce risk from endogenous pig retroviruses.
Early clinical use

Human kidney transplants have begun.

Immunosuppression and long-term safety monitoring remain necessary.

Deck illustration: a pig beside an engineered kidney being placed into a human outline
Deck illustration: human stem cells injected into an animal blastocyst, and a pig with a glowing human pancreas growing inside it

Replacement · Stanford / Nakauchi lab

Human tissue grown inside an animal host.

Blastocyst complementation, in three moves.

  1. 1Create an empty organ nicheDisable the host’s ability to make a specific organ or cell type.
  2. 2Add human stem cellsDonor cells could fill the vacancy as the host develops.
  3. 3Recover replacement tissueGoal: patient-matched tissue grown within an animal.
Preclinical

Proof in rodents. Human organs remain a goal.

2026 advance: protecting donor cells from early immune clearance.

Replacement · Renewal Bio

Development as a bioreactor.

Instead of coaxing one cell type at a time, let an embryo model do what development already does: organize patient cells into the progenitors of blood, liver, pancreas and neural tissue. Preclinical R&D. Not a reproductive program. Scientific foundation: Prof. Jacob Hanna, Weizmann Institute.

  1. 01Patient iPSCs
  2. 02Controlled self-organization
  3. 03Embryo model
  4. 04Organ progenitorsBloodLiverPancreasNeural tissue
  5. 05Future replacement therapies
Deck illustration: patient iPSCs, controlled self-organization in a vessel, an embryo model, dishes of organ progenitors for blood, liver, pancreas and neural tissue, and a human outline for future replacement therapies
From the deck. Preclinical R&D · Not a reproductive program.
Each development stage unlocks new clinical cell types. The shaded band is where the program stood in the 2024 deck. Source: Renewal Bio, 2024.

In the record

  • CellCeci n’est pas un embryon de sourisVol. 185 No. 18 · September 1, 2022
  • NatureEmbryo models: stem cells form structures that mimic two-week-old human embryos19 October 2023
  • CNNA synthetic embryo, made without sperm or egg, could lead to infertility treatments
  • NBC NewsScientists use stem cells to create synthetic mouse embryos
  • CNN HealthScientists report creation of first human synthetic model embryosJune 15, 2023
  • The New York TimesScientists Debut Lab Models of Human Embryos
  • ABC Science‘Synthetic human embryos’ were created using stem cells. What are they, and why were they made?June 20, 2023

Covers and headlines as shown in the deck. The embryo-model work is from the Hanna lab; the stembroid data in the deck is marked confidential and is not reproduced here.

08 The repair thesis · Maintenance · Restoration

We restore complex machines. Biology needs a repair toolkit.

Aging accumulates damage across interacting systems. The answer is repeated, system-level repair.

Diagnose→Replace→Reprogram→Regenerate

The thesis

Maintenance can turn decline into renewal.

Deck illustration: a classic car, rusted on one side and fully restored on the other, in a workshop

09 Why now

The breakthrough is convergence.

Single-cell measurement, AI biology, gene delivery, cell engineering, developmental biology and immune control each matured on their own. Braided together they make the three levers usable.

Single-cell measurement AI biology Gene delivery Cell engineering Developmental biology Immune control Reprogram Replace Regenerate
No single technology is enough. Hover an input or an output.

10 Paths to 160

Who’s building it.

The NFX Bio companies from the deck, sorted by the plan they advance. Only one of them is a Plan A company today.

Plan A Repair
Plan B Bioengineering
Plan C Cryonics
  • Not a company. A wait.
Plan D Death
  • Not a company. The default.

Keep going

What else happens if this works.

Longer healthy lives change the national debt, health spending, and how long people earn. The byproducts page walks through the numbers in the deck.

Read the byproducts →
Omri Amirav-Drory

The talk

Omri Amirav-Drory, PhD

General Partner at NFX, where he leads NFX Bio. This site follows his talk slide by slide. Where we added a sentence of our own, it’s plain-language framing, not new science.

Omri on nfx.com ↗