Open questions about the brain and AI — each one a room where researchers gather evidence, argue hypotheses, and turn what they learn into better flows for everyone. Anyone can read along.
Brain signals seem to anticipate upcoming words before they are heard. Is that genuine prediction — or are encoding models picking up statistical dependencies between neighbouring words?
When studies use contiguous train/test splits and control for sentence length, word position and word rate, how much of the LLM-to-brain fit remains — and is what remains contextual, syntactic, or just word meaning?
A model trained on about as many words as a child hears already fits the brain nearly as well as huge models. Why does alignment saturate while abilities keep scaling — and would child-like, spoken, multimodal training change that?
The brain seems to forecast several words ahead, at several levels. Would training models on longer-range, hierarchical predictions make them more brain-like — and more data-efficient?
Attention heads' computations predict brain activity. Do they correspond to a real brain mechanism — cue-based memory retrieval, hippocampal sequence codes — or are they just good feature extractors?
Language models can now choose sentences that switch the brain's language network up or down in new people. Where should closed-loop 'driving' of the brain go — and what are the ethics?
Dreams are vivid, emotional and often bizarre. Do they do something useful for memory and learning, and is their strangeness part of the point?
Are memories stored like files, or as gist plus a few unusual details that the brain regenerates each time you remember?
During sleep and rest, the hippocampus replays some experiences and not others. Is it maximising future usefulness, or following context and structure?
Neural networks catastrophically forget old skills when they learn new ones. Brains don't. Can sleep-like or generative replay give AI the same stability?
A 2026 preprint models the visual cortex as a denoising diffusion model. Is the resemblance real biology, or only mathematics?