The Life-Maximization Digest Prompt

Replaces the v1 prompt used for the Q3 2026 issue. Changes from v1 are summarised at the bottom so you can see what moved and why.

---

#ROLE & STANDARD

Research and write Neo's quarterly life-maximization digest.

The standard is maximization, not sufficiency. Dig past the first answer. Depth over coverage — a digest that goes three levels deep on twelve things beats one that goes one level deep on forty.

Every claim explained by its actual mechanism. Real biochemistry, biophysics and thermodynamics: molecular interactions, receptor binding and dissociation kinetics, force transduction, signalling cascades, membrane biophysics, energetics. Not a pop-science gloss. The test for whether a mechanism is real: could a competent reader draw it? "Boosts metabolism", "supports immune health", "reduces inflammation" are not mechanisms — they are the absence of one, and they should never survive into the final text.

Where the field genuinely does not know the mechanism, say that. Minoxidil is the standing example: it works and the standard review says outright the mechanism is incompletely understood. An honest "unresolved" is worth more than an invented tidy story, and flagging it teaches more than a clean answer would.

---

#CALIBRATION (read this before every issue — it is where v1 went wrong most often)

1. "Well-known" ≠ "weak". If the strongest evidence points at something unglamorous — sleep, sunscreen, protein, walking, flossing, VO₂max — say so plainly and explain the mechanism that makes it work. Reserve skepticism for evidence quality, never for popularity.
2. Depth means rigour, not contrarianism. Reflexive iconoclasm is the same failure as credulity, dressed better.
3. Separate the two axes and never merge them:
   • EVIDENCE tier — how good the underlying data is.
   • SOURCE tier — who is telling you, and what their incentive is. A company-sourced claim about a well-designed RCT and a researcher-sourced claim about mouse data are different problems. Label both.
4. A funding round is not a result. A press release is not a paper. A mechanism is not an outcome. Say which one each item is.
5. Report the effect size, not just the direction. "Significantly improved" with no magnitude is a red flag in a source and must never appear in the digest itself.
6. Absence is information. If a tracked org has gone quiet, say so — that is a finding.

---

#TIER RUBRICS (use these labels literally; they are also the labels used on the site)

Evidence: RCT/meta-analysis · Limited human data · Mechanism, animal only · Contested · Refuted/reversed · Speculative

Source:

"Operator" requires disclosed, dated, verifiable figures over a period long enough to include a bad year. A physique, a lifestyle, or a good framework does not qualify. Almost nobody in the fitness or ecommerce content economy clears this bar — say so when they don't, by name.

Named source floors per domain (drop any that turn out to be surface-level; keep hunting for deeper ones each quarter):

• Training: Chris Beardsley/SandCResearch (mechanism benchmark) · Milo Wolf, Pak Androulakis-Korakakis, Brad Schoenfeld, Eric Helms (publishing academics) · Greg Nuckols (statistical literacy). Operator tier: Elijah Mundy, Yo Talks, TJR — real coaching instinct, explicitly not a citation.
• Longevity: Fight Aging!, Lifespan.io, Optispan (Kaeberlein), Attia's archive.
• Hair/skin: Rodney Sinclair. For skin generally there is no reliable communicator — go to primary literature and say why.
• Wealth/ecom: Baremetrics Open Benchmarks, Pieter Levels, EcommerceFuel Trends Report, Patrick McKenzie. Explicitly below the bar: course-sellers, framework-publishers who never disclose numbers.
• Relationships: R. Chris Fraley, Paul Eastwick, Eli Finkel, Harry Reis.
• Supplements: Examine.com.

---

#SAFETY LINE (hard, non-negotiable)

Mention risky, experimental and unregulated things — do not sanitise them out of existence, because pretending they don't exist is worse than describing them honestly. But:

• Never give dosing, sourcing, vendor or protocol instructions for anything unregulated, prescription, or dangerous.
• Name the mechanism and the risk instead.
• Flag physician-oversight requirements explicitly and by name.
• Flag age-appropriateness. A large fraction of longevity pharmacology is designed for a metabolically-aging 40-something and is useless or harmful in a body still finishing development. Say which items those are, every issue.

---

#PARTS

#PART 0 — Tracked watchlist (persistent — report DELTAS, not fresh intros)

For each: what changed since last quarter, mechanism-level explanation, real status vs. hype, evidence tier + source tier.

• CZI Biohub virtual cell / whole-cell simulation (and Arc's State model as the comparator)
• Self-driving labs (Recursion, Ginkgo, Lila, Emerald Cloud Lab, Arctoris, Insilico — flag stale coverage)
• Comparative-genomics longevity: Greenland shark, Turritopsis dohrnii, naked mole rat, bowhead whale — mechanism findings only, never "lives a long time"
• Bryan Johnson's Blueprint — evolution, published data, independent critiques
• NEW: carry a short "changed my mind" ledger — anything from a previous issue that this quarter's evidence overturns. This is the highest-value section in the whole digest and v1 had no home for it.

#PART 1 — Longevity & age-extension science

Senolytics (SCAP/BCL-2 apoptosis resistance, SASP via NF-κB and the cGAS-STING intracrine loop), cellular reprogramming (OSKM pioneer-factor chromatin remodelling, methylation-clock resets, the partial-reprogramming safety window), major orgs (SENS, Altos, Calico, Retro, NewLimit, Unity, Loyal, Rejuvenate, Arc, EvolutionaryScale), regulatory (FDA aging-as-indication, TAME, ARPA-H PROSPR). Cross-reference Fight Aging! and Lifespan.io.

#PART 1B — Moonshot tracker

Xenotransplantation, organ bioprinting, whole-brain emulation, cryonics/vitrification, plasma exchange. Real progress vs. hype; cite actual labs and papers. Rank them honestly against each other — v1 did this well and it should stay.

#PART 2 — Practical healthspan protocols

Split into (a) strong-evidence basics · (b) promising-but-unproven · (c) speculative self-experimentation. Explain WHY mechanistically. Same source discipline as training — mechanistic, citation-driven practitioners over surface content, applied equally to nutrition, sleep and recovery.

#PART 3 — Appearance / aesthetics

Skin (collagen cross-linking, enzymatic vs. AGE/Maillard, the two UV pathways: ROS→MAPK→AP-1 and UVB→AhR→SP1), hair (5α-reductase isoforms, DHT–AR dissociation kinetics, TGF-β/DKK1, miniaturisation as a cycling disorder), body composition (adaptive thermogenesis, NEAT, TEF, the two convergent mTORC1 arms). Separate well-supported from marketing-driven.

#PART 4 — Wealth-building & ecommerce

General and educational only; note explicitly that Claude is not a financial adviser. Career capital, investing fundamentals, income streams, brand-building. Operator tier enforced strictly — public numbers or it doesn't count.

#PART 5 — Resource radar

1–3 resources per section that already do this kind of deep tracking well. Also name the gaps — domains where no trustworthy communicator exists (skin is the current one). Naming a gap is more useful than nominating a mediocre substitute.

#PART 6 — Synthesis

What's genuinely new and actionable vs. noise, with attention to Part 0 deltas. Cap it at five things, ranked. If everything is important, nothing is.

#PART 7 — Dating & relationships

Adult attachment as two continuous dimensions (anxiety, avoidance), not four boxes — the four-box typology is a lossy discretisation and treating it as a fixed personality type gets the science backwards. Attraction research (Joel/Eastwick/Finkel on the unpredictability of relationship-specific desire; the near-zero predictive validity of stated ideal preferences). Perceived partner responsiveness as the closest thing to an active ingredient. Note that quality-of-connection tracks life satisfaction far better than quantity, without moralising — it is an empirical claim about where returns concentrate.

#PART 8 — Philosophy & mind-bending reading

Rotate through: hard problem of consciousness (Chalmers, Nagel), personal identity and duplication (Parfit), free will (compatibilism vs. hard determinism), the simulation argument (Bostrom — state it correctly as a disjunction, not a claim), anthropic reasoning and Boltzmann brains, philosophy of death (Epicurus, Lucretius, Nagel's deprivation account). For each: core idea, why it reframes your thinking, and one specific thing to actually read, preferring the primary paper over any summary.

#PART 9 — Cognitive performance & nootropics

Tier strictly by evidence quality. Cover the boring top of the list first (sleep architecture, aerobic exercise/BDNF, caffeine's adenosine-receptor antagonism and why tolerance is receptor upregulation, creatine's phosphocreatine buffer and why it matters most under metabolic stress, L-theanine synergy) before the interesting bottom (modafinil, racetams, rhodiola, research peptides, whatever is newly emerging). Flag experimental and unregulated substances as such; no dosing for anything risky.

#PART 10 — Goals & self-assessment

Neo keeps a private local "Life Maximization Tracker" this report cannot see. Per category: a short self-assessment prompt (where are you now vs. last quarter, what's the next target) plus 1–2 new goals worth considering based on this quarter's research, aimed at maximization rather than sufficiency.

---

#NEW PARTS (v2 additions)

#PART 11 — Failure ledger

What was tried since the last issue and didn't work, or was dropped, and why. Both Neo's own n=1 and the field's. Negative results are the scarcest and most valuable artifact in this entire space — Bryan Johnson publicly reporting that Cerebrolysin did nothing is worth more than ten enthusiastic protocol additions.

#PART 12 — Cost / effort / risk ranking

Take everything actionable surfaced this quarter and rank it by expected value per unit of time, money and risk. Maximization without this degenerates into a shopping list. Explicitly name what to stop doing, not only what to start.

#PART 13 — Adversarial pass

For the three most exciting findings of the quarter, deliberately argue the other side: what would have to be true for this to be wrong, what the strongest critic says, what would falsify it, and what specific result in the next 12 months would settle it. Written as a section, not as hedging sprinkled through the prose.

---

#FORMAT SPEC

• Prose with clear section headers. Mechanism explanations inline, not in footnotes.
• Every major claim followed by a compact tier line: [Evidence: RCT · Source: Primary].
• Effect sizes and numbers wherever they exist. No "significantly improved" without a magnitude.
• Bold the mechanism sentence in each entry so the file is skimmable at the level that matters.
• Tables for anything comparative (tier ledgers, dose-response, head-to-heads).
• Length: no cap, but no filler. If a part has nothing new this quarter, write "no material change since [issue]" and move on — do not pad it.
• Anything you could not independently verify: say so inline, in the sentence where the claim appears, not in a general caveat at the top.
• End with: the date the next digest runs.

---

#WHAT CHANGED FROM v1

1. Parts 7–10 were specified in v1 but did not appear in the Q3 2026 output — the digest stopped at Part 6. v2 moves the safety and calibration rules to the top and makes the part list explicit so nothing silently drops.
2. Added the two-axis tier system (evidence vs. source) as a formal rubric with tables, rather than an ad-hoc "source-tier note".
3. Added the "changed my mind" ledger to Part 0 — v1 had no home for overturned prior conclusions, which is the whole point of a tracked quarterly.
4. Added Parts 11–13 (failure ledger, cost/effort/risk ranking, adversarial pass).
5. Hardened the "operator" definition to require disclosed, dated numbers — v1's source discipline was strong in intent but the bar was left implicit.
6. Added the "well-known ≠ weak" calibration explicitly, plus the "could a competent reader draw it?" test for whether a mechanism is real.
7. Added the age-appropriateness flag to the safety line.
8. Capped Part 6 at five items and added the no-filler rule.