Skip to main content

Recent news from SQI: April 2026

From automating developmental neuroscience studies to facilitating LLM "brainstorming" sessions, this spring, the MIT Siegel Family Quest for Intelligence supported research across neuroscience, AI, and cognitive science that continues to push the boundaries of our understanding of both human and artificial intelligence.
  • Animated alien character
    Image generated for a video used in developmental neuroscience studies.
    Bianca Santi, Matthew Soza, Greta Tuckute, Aalok Sathe, Evelina Fedorenko, Halie Olson

A Faster Way to Create Research Videos for Children

Researchers including Bianca Santi ’25, postdoc Dr. Halie Olson and others found a way to utilize simple speech synthesis and animation tools to generate engaging, child-friendly audiovisual stimuli for use in developmental neuroscience studies. That content would usually take weeks to animate manually—but using this pipeline, the whole process takes minutes or even seconds. Read more about the paper and what it means for the future of developmental neuroscience research on our website.

 

Dr. Greta Tuckute: Bridging Neuroscience and AI using Language

As a PhD student at MIT, Dr. Greta Tuckute worked on language processing research under Prof. Ev Fedorenko. Now she's a Research Fellow at the Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard University, where she studies how language is implemented in biological and artificial systems.

We spoke with Greta about her experiences at MIT, creating human-inspired models, and learning about her own brain through machine research.

Read the profile here.

 

Collective Brainstorming with LLMs

Running complex reasoning tasks through LLMs can be costly, slow, and energy-intensive. DisCIPL, a new framework developed by Jacob Andreas in collaboration with Josh Tenenbaum, Vikash Mansingkha, and others with funding from the MIT Siegel Family Quest for Intelligence, seeks to address this issue by leveraging the strengths of smaller language models in a collective 'brainstorming' approach—delivering faster, cheaper, more accurate, and more energy-efficient responses. The team has recently published a paper, which was covered on MITNews.

 

Where Human Intelligence Meets AI

In January, SQI presented research updates and an overview of the Platforms during a workshop hosted by the Schwarzman College of Computing. The talks were recorded, along with other SCC workshops.

Join our events mailing list to learn about events like this and for invitations to the SQI Seminar Series, where we welcome speakers from MIT and beyond to discuss the latest research in human and artificial intelligence. Recordings of previous seminars can be found on our YouTube channel.

Join the events mailing list to be notified of upcoming talks and other events.