The FDA approves a curative therapy that halts and reverses cognitive decline in Alzheimer's disease before 2035.
Current Alzheimer's treatments merely slow decline rather than curing the underlying neurodegeneration. Resolution requires regulatory approval of a disease-reversing therapy.
Across 6 frontier models answering independently, the panel has all but ruled it out: the median is 9%.
They broadly agree, 15 points between GPT at 18% and Qwen at 3%.
Nothing is settled until 8 Oct 2036, when the outcome is checked against sources and every number above gets a permanent score.
What each one said
-
GPTOpenAI
It hinges on whether amyloid/tau biology can be reversed after symptomatic neurodegeneration, not merely slowed.
- FDA has approved amyloid antibodies that modestly slow decline, but no program has shown durable cognitive restoration in established disease.
- A sponsor must produce replicated, controlled evidence of meaningful reversal and an acceptable safety profile before a 2035 filing window.
- The odds rise sharply if a phase 2 therapy shows sustained improvement on CDR-SB and biomarkers, rather than reduced rate of worsening.
18% -
GrokxAI
Any late-stage candidate must show actual cognitive improvement (not slowing) in Phase 3 by ~2032.
- No Phase 3 asset today reverses decline; anti-amyloids only slow it, and no regenerative modality is near approval.
- AD programs fail ~99% of the time and take 10–15 years, leaving scant runway from 2026.
- A clear 2027–28 Phase 2 reversal result would shift this; none is on the horizon.
12% -
DeepSeekDeepSeek
No Alzheimer's drug has ever reversed established neurodegeneration, and the biology remains unsolved.
- Lecanemab and donanemab only slow decline ~27-35%; both target amyloid, which removal alone does not restore lost neurons.
- A true reversal needs neuronal regeneration or clearance of established tangles, neither of which has a clinical-stage candidate with positive data.
- Leqembi's 2023 accelerated approval shows FDA will approve modest slowing, not reversal; a tau or regenerative breakthrough by 2035 is the only path.
12% -
ClaudeAnthropic
No candidate in late-stage trials today reverses cognitive decline, only slows it, and that gap won't close by 2035.
- Current anti-amyloid drugs (lecanemab, donanemab) only modestly slow decline, not reverse it, and no reversal-capable drug is in Phase 3.
- Developing, trialing, and FDA-approving a truly curative neurodegeneration therapy typically takes 10-15+ years from a validated mechanism, and no such mechanism is…
- A breakthrough in tau-clearance, neuroregeneration, or gene therapy with strong Phase 2 reversal data emerging within 2-3 years would meaningfully raise this.
5% -
GeminiGoogle
Halting and reversing existing neuronal loss requires breakthroughs in cellular regeneration that are not even in Phase 2 clinical trials.
- No therapeutic candidate currently in late-stage development demonstrates sustained cognitive reversal in clinical trial endpoints.
- A 10-year window is too brief given standard 8-to-12 year drug development cycles from early-phase discovery to full FDA approval.
- True reversal demands restoring lost synapses and dead neurons, a biological milestone far beyond clearing amyloid or tau proteins.
4% -
QwenAlibaba
No therapy in any pipeline has demonstrated reversal of cognitive decline, and the regulatory bar for 'curative' is extraordinarily high.
- Current approved drugs (lecanemab, donanemab) show ~27-35% slowing of decline; none halt or reverse, and no Phase 3 candidate has shown reversal signals.
- FDA approval requires pivotal trial data showing reversal, meaning a drug would need to enter Phase 3 soon and read out by ~2033-2034 — no such candidate exists today.
- Even optimistic scenarios (gene therapy, tau-targeting, combination approaches) are in Phase 1/2 at best; the jump from slowing to reversing is a mechanistic chasm, not…
3%
6 of 6 models answered · 15 points between the highest and lowest. None was shown the market price.
Question sourced from a news sweep on 11 Oct 2026. Forecast by google/gemini-3.8-flash, anthropic/claude-sonnet-5, openai/gpt-5.6-terra, x-ai/grok-4.6, deepseek/deepseek-v4.1-flash, qwen/qwen3.8-max-0902 via OpenRouter.