An AI-designed drug candidate cures or reverses Alzheimer's pathology in human clinical trials before 2036.

Question

Six frontier AI models put this at 9% by 23 Sept 2036. Full reasoning from every model.

WillHappen.ai · Accepted Answer

9% — the median of 6 frontier models, resolving by 2036-09-23.

Answer

Gemini (Google): 8% — Clinical trial timelines and the irreversible nature of neuronal loss make demonstrating true disease reversal by 2036 near-impossible. Moving an AI de novo candidate from discovery through Phase 2/3 trials typically requires 8-12 years, exhausting the window to 2036. Alzheimer's clinical trials suffer over a 99% attrition rate, with clearance of amyloid or tau failing to restore lost cognitive function. Discovery of a reliable biomarker of structural neural regeneration would be required to prove reversal rather than mere slowing.

Answer

Claude (Anthropic): 9% — Hinges on any AI-designed candidate showing actual pathology reversal (not just slowed decline) in human trials by 2036. No Alzheimer's drug in history has 'reversed' pathology in humans; even amyloid drugs only slow decline modestly. AI-designed candidates (e.g. Insilico, Recursion) are only reaching early trials now, leaving ~9 years for full Phase 1-3 plus regulatory review. Bar of 'cures or reverses' is far stricter than typical endpoints, so even a breakthrough disease-modifying drug likely wouldn't qualify.

Answer

GPT (OpenAI): 18% — It hinges on translating AI-designed target engagement into durable human biomarker and clinical reversal, not discovery speed. Sponsors must advance an AI-originated candidate through Phase 1–3 and publish human reversal evidence by 23 September 2036. Alzheimer’s has repeatedly defeated amyloid, tau, and neuroprotection programs; approved antibodies mainly slow decline rather than reverse it. Odds rise sharply if a Phase 2 program shows replicated amyloid/tau reduction plus sustained cognitive improvement by 2030.

Answer

Grok (xAI): 12% — An AI-designed candidate must reverse plaques/tangles plus cognition in a completed human trial by 2036. IND-to-phase-3 readout still takes 7–10 years, so only molecules already in or entering clinic this decade can finish. No approved agent has reversed AD pathology in humans; anti-amyloid mAbs clear plaques without reversing disease. AI speeds discovery but leaves 99% historical DMT failure rates and biology unchanged.

Answer

DeepSeek (DeepSeek): 8% — This turns on whether an AI-designed molecule can clear Phase 2/3 in a disease where every prior mechanism has failed. Alzheimer's has a 99%+ Phase 2-to-approval failure rate; lecanemab and donanemab only slow decline ~27-35%, not reverse. AI-designed candidates like Insilico's rentosertib (Phase 2, 2025) show AI can generate molecules, but none target Alzheimer's yet. A single AI-designed drug entering Phase 3 by 2030 with reversal data by 2036 would be needed; no such candidate is in trials today.

Answer

Qwen (Alibaba): 3% — Alzheimer's reversal in humans is extraordinarily hard; no AI-designed candidate is even in Phase 2 for AD today. As of mid-2026, no AI-designed molecule targeting Alzheimer's has entered Phase 2 trials; the earliest candidates (e.g., from Insilico, Iktos) are preclinical or Phase 1… Lecanemab and donanemab only slow decline ~27-35% after years of development; 'cure or reverse' sets a far higher bar that no mechanism has achieved in humans, and AI… A surprise would be an AI-designed antisense or gene therapy showing dramatic plaque/tau clearance in a Phase 1/2 by 2030, but even that would need confirmatory…

An AI-designed drug candidate cures or reverses Alzheimer's pathology in human clinical trials before 2036.

Britain's MHRA publishes a regulatory roadmap for space-manufactured pharmaceuticals before the end of 2026.

The US FDA approves an anti-aging therapeutic indicated specifically to slow biological aging before 2035.

A fully lab-grown solid organ is successfully transplanted into a human patient in a clinical trial before 2032.

The US FDA approves a gene therapy targeting sensorineural hearing loss before 2030.