rationale:

"None by 2040" (40%)

The most likely outcome remains that none of these questions is answered by an AI-originated result before 2040, though the probability sits below 50% given the extreme pace of AI capability growth. To qualify, a general-purpose AI system must produce both the central idea and the main result, and biologists must come to accept it. While general systems recently scaled from 5% to 100% on FrontierMath Tier 4 within 18 months, biology operates under very different constraints. Wet-lab rounds still take 1–5 weeks, independent replication is slow—as of October 2026, no AI-originated biology claim has been independently replicated Early progress on accelerating drug development with Claude—and several of these broad questions may never yield a single, consensus answer. Across the cluster, the shared bottleneck is not merely generating breakthroughs but reaching scientific consensus on multi-causal biological questions, making the timelines for accepted answers slower than capability growth alone might suggest. Finally, there is a strong risk of human preemption, as human-led research teams remain highly active across all these domains.

Anesthesia (18%) and Non-coding Genome (16%)

If an AI does answer one of these first, the mechanism of general anesthesia and the function of the non-coding genome are the strongest candidates. Anesthesia benefits from fast experimental feedback loops with controllable, measurable transitions. Recent human-led work using cortical assembloids and macaque studies https://www.cell.com/cell-reports/fulltext/S2211-1247%2826%2900126-9 Stem cell-derived brain models offer a new window into ... demonstrates that tractable perturbation systems exist, offering a promising setting for an AI agent to identify an accepted unifying mechanism. The non-coding genome is similarly attractive because it is highly scalable and data-heavy. Tools like the AlphaGenome Atlas (Sept 2026) already predict variant effects AlphaGenome Atlas: in silico mutagenesis of the entire ... AlphaGenome Atlas: in silico mutagenesis of the entire ..., but they count only as tools for a general agent here. The non-coding genome trails slightly behind anesthesia because it faces a higher risk of definitional disputes over what constitutes biological "function" versus junk DNA Non-coding DNA's Alpha Moment, whereas anesthesia has clearer mechanistic readouts.

Memory (11%) and Alzheimer's (9%)

How the brain stores long-term memories and the primary causal driver of Alzheimer's pose stiffer challenges. Memory storage has concrete mechanistic footholds, such as KIBRA–PKMζ modeling PKMζ-KIBRA interactions, molecular turnover, and memory and preserved synaptic architecture in hibernation https://www.science.org/doi/10.1126/science.aee7004, but translating these local findings into a generalized, brain-wide causal account is a massive leap. Alzheimer's research currently attracts immense funding and AI-hypothesis generation, such as Biomni-AD's target prioritization Mount Sinai Scientists Recognized With Global AI Prize for ... and Kosmos's flippase-gene findings Kosmos: An AI Scientist for Autonomous Discovery. However, resolution requires identifying the primary causal driver. Since Alzheimer's is likely multifactorial Refining, not abandoning, the amyloid hypothesis, establishing a single primary cause requires lengthy clinical validation, making a clear-cut AI-originated answer before 2040 much less likely.

Aging (6%)

The upstream molecular cause of aging is the least probable outcome. Like Alzheimer's, validation requires multi-year lifespan studies, severely capping the speed of iterative AI-led research. Furthermore, the premise of a single "upstream molecular cause" is deeply contested; recent cross-mammal studies and consensus at ARDD 2026 point toward modular architectures and multi-causal theories Universal transcriptomic hallmarks of mammalian ageing ... The evolution of aging research: from theories to ... ARDD 2026 Daily Recap. While general agents are generating early leads—such as Co-Scientist proposing metabolic regulation mechanisms https://deepmind.google/blog/opening-new-paths-in-aging-research/ and genetic factors https://deepmind.google/blog/fast-tracking-genetic-leads-to-reverse-cellular-aging/—the biological reality likely precludes a single upstream answer, rendering the question uniquely difficult to resolve as specified.