← back

i_demo_016 · patient advocate × cancer · how to read evidence-language without the marketing

Why a cancer-survivor advocate would come to evidence.x1000.ai

A cancer survivor (or family member, or patient-advocacy group volunteer) reads a constant stream of oncology marketing dressed as evidence. Pain points:

• miracle / breakthrough / 100% effective framing is everywhere, in both lay press and sponsor decks;
• distinguishing a real Tier-1 evidence claim from sponsor-imply marketing is hard;
• multi-condition lookups (cancer survivor + chronic pain + CKD comorbidity) rarely find a unified resource.

m625 does not provide medical advice. **m625 shows what a real Roll-formalized cancer evidence claim looks like (anchored on m610's 3 base cooked cases) and how to tell it apart from marketing language**.

This is not patient-facing medical advice. This is not a treatment plan.

---

The Roll-formalized object: the contrast between cooked evidence and marketing

Real cooked case (m610)	Marketing language to distrust
K4.1 = (0, 1, 1) declared explicitly · case_001 EGFR-NSCLC	cures cancer
Population-level viable set V described	this patient will be cured
Falsifying prediction documented (N-doubled re-cook path)	guaranteed outcome
Audit chain: draft → audit-pending → audit-pass → cooked → outbound	FDA approved without NCT
3 base cooked cases · each ≥ B grade machine audit	revolutionary-breakthrough
Honest verdict 4-section (done / open / weakness / counter)	100% effective miracle

The contrast is the lesson, not the cases themselves.

---

How a cancer-survivor advocate should read this (≤ 5 minutes)

• Find the K4.1 declaration. A genuine Roll-formalized cancer evidence claim has a 3-bit K4.1 declaration (per m604 V2.6 sovereign reference · Kernel §3.4). If a sponsor's pitch deck or community resource cannot point to one, the framework says: this is below Tier-1.
• Find the falsifying prediction. Real evidence carries a statement of what would prove it wrong. "Here is what would prove this wrong: N-doubled re-cook fails" is a falsifying prediction. This always works is not.
• Find the source citation chain. Numbers tied to NCT / PubMed / DOI IDs are anchorable. Numbers with no source are training_data_recall at best, fabrication at worst (W6.1).
• Find the honest verdict section. A real demo says what's open, what's weak, and what would defeat its own claim. Marketing says only what works.

---

Five marketing patterns m625 has cataloged as attack-surface fixtures

ID	Pattern	What to watch for
AS-001	cure cancer smuggle	direct vocabulary use
AS-005	simulate patient trajectory	tool-as-engine over-claim
AS-006	personalized digital twin	per-patient model promise
AS-010	revolutionary-breakthrough + FDA approved + act now + limited time	full marketing stack
AS-029	personalized treatment via unique digital twin learning in real time	adaptive-learning over-claim

m625's audit catches all of these in the audit pipeline. A community resource that does not catch them is structurally weaker than one that does.

---

Cross-domain reading aid (per W6.3)

If a survivor advocate also lives with chronic pain or CKD or autism (cross-comorbidity), the K4.1 bit pattern of each condition's cooked Roll claim differs:

• cancer EGFR-NSCLC TKI · (0, 1, 1)


• chronic pain Tu 2020 · (1, 1, 0)


• CKD progression · (1, 1, 0)

These are **structurally different claims**. A community resource that conflates them is making a W6.3 cross-domain conflation error. transferability_pct: not directly transferable at the substantive level.

---

What this demo will not say (anti-drift)

• ❌ It does not promise reversal, eradication, or rescue for cancer survivors (cure / reverse lexicon).
• ❌ It does not assign per-person trajectory for any survivor (NCNU P_3).
• ❌ It does not affirm or vouch for any oncology clinic, hospital, or sponsor.
• ❌ It does not name specific drugs or supplements as recommended.
• ❌ It does not use miracle / breakthrough / state-of-the-art lexicon.

---

Falsifying prediction

The "real evidence has K4.1 + falsifying prediction + audited source chain + honest verdict" framing is falsified by:

• a published Tier-1 oncology evidence paper that demonstrably lacks all four components and yet is consensus-grade; OR
• a sponsor-deck that meets all four components and yet is consensus-rejected as marketing.

---

Attack surface entries

• AS-001 (cure cancer smuggle) — G2.2
• AS-010 (revolutionary-breakthrough + FDA-unanchored marketing stack) — G2.2 + W6.2
• AS-029 (personalized treatment via digital twin combo) — G2.2

---

Ferry target

m511 (page_candidate · public surface for cancer-survivor advocacy community) · m590 (claim_pack secondary).

---

Honest verdict (W6.5)

• 已做: cancer-survivor advocate reading layer · contrast cooked-vs-marketing · 4 reading checks · 5 attack-surface examples · cross-comorbidity bit-pattern table.
• 待办: plain-language Chinese variant (v1.0+); per-cancer-subtype reading aids (breast vs lung vs lymphoma) — open work.
• 漏洞: contrast table is illustrative · the boundary between "marketing" and "Tier-2 evidence not yet at Tier-1 standard" is not perfectly bright. Some claims that fail one of the four checks may still be honest exploratory work; readers should not weaponize this reading aid against legitimate cautious claims.
• 反驳预案:

- *"This dismisses too much hopeful framing."* — That is the point. Hope is not evidence. m625's role is to teach the language; community decisions about hope are separate. - *"My doctor uses the `cure` word and they're trustworthy."* — Possibly. Specific clinicians may use shorthand that the framework forbids in writing. The framework is about *publishable language*, not bedside conversation.