Stanford’s Wu Tsai Neurosciences Institute and Institute for Human-Centered Artificial Intelligence have awarded grants to five projects combining neuroscience and AI. The work ranges from studying how infants learn to using brain activity and language models to explore psychotherapy and smell.
Stanford HAI Watch analysis
What happened
The awards are the second round of a collaborative programme launched in 2022. Stanford says this year’s projects aim both to advance neuroscience and to build AI informed by human experience.
The projects include using AI to connect depression-related brain scan patterns across EEG and fMRI; modelling infant learning using brain scans and video; and building a foundation model to study how neuronal cell types develop. Other teams will examine how therapeutic dialogue relates to brain activity, and how brain signals and language models might help map people’s experience of smell.
Why it matters
These are not five bids to make a chatbot sound vaguely brainy. They put AI methods to work on concrete scientific questions, while also asking what human learning, perception and communication could teach AI researchers.
The projects could help researchers investigate patterns that are difficult to study with one data source or discipline alone. Their aims are research questions, not established clinical results: the announcement describes proposed work, not demonstrated improvements to diagnosis or treatment.
Our read
The most compelling part is the two-way ambition. AI is being used to study the brain, but human perception and learning are also being treated as useful material for building better AI. That is a more interesting exchange than simply asking whether a model can pass another test.
What to watch
- Whether combining EEG and fMRI helps researchers distinguish depression subtypes.
- What infant learning reveals about making AI learn more efficiently.
- Whether the psychotherapy and smell projects find useful links between experience and brain activity.
- Which results can be replicated and translated into practical tools.
Discussion spark: Which of these research directions is most likely to produce useful results first: learning from infants, linking brain signals to experience, or applying AI to clinical neuroscience?
Sources and evidence
- Source update (11 November 2026, 00:00 UTC)
Independent WittyWires tracker for public updates about Stanford HAI. Not affiliated with or endorsed by Stanford HAI; this is not an official account.