Menu

Missions and Platforms

SQI researchers aim to understand intelligence — how brains produce it and how it can be replicated in artificial systems. We approach this as a single grand challenge requiring the organized, collaborative efforts of science, engineering, the humanities and beyond. 

To execute on its vision, we have established “Missions,” long-term collaborative projects rooted in foundational questions in and centered around a single domain of intelligence, and “Platforms,” software systems that enable Missions research in new directions.

Each Mission brings together a team of scientists and engineers to pose and answer foundational questions of human intelligence where current AI falls short, to build engineered systems as scientific hypotheses to advance these studies, and to execute tests of those systems to ensure that scientific progress is iteratively guided by human intelligence results and real-world AI engineering challenges.  

Platforms are software systems that enable Mission teams to pursue research in new directions, specifically through creation of benchmarking and testing interfaces that use data from the Missions to help refine and expand the work.

Launched Missions

  • Perceptual Intelligence

    The Perceptual Intelligence Mission brings together computer scientists, cognitive scientists, and neuroscientists with a shared goal of producing the first machine executable models of human visual intelligence that work computationally, cognitively, and neurally.

  • Language and Thought

    The Language and Thought Mission broadly aims to understand the relationship between language and human intelligence. Scientific goals include understanding how humans and machine learning models interpret and generate language and determining the role of language in the acquisition, representation, and use of knowledge across various domains of cognition. Engineering goals include building AI systems that learn language from human-scale datasets and understand language with human-like robustness and flexibility.

  • The Development of Intelligent Minds

    This research mission broadly aims to understand how children grasp new concepts from few examples, how children build upon layers of concepts to reach an understanding of the world and have the flexibility to solve an unbounded range of problems. Can we build AI that starts like a baby and learns like a child?

  • Embodied Intelligence

    This research mission broadly addresses how we perceive the world around us and integrate this information to plan and complete tasks. Scientific goals include research into how perception, planning, and action interface, how we learn efficiently from small data sets and the creation of behavioral benchmark tasks.

Platforms

  • Scaling Inference

    A great deal of enthusiasm, in both AI and in brain and cognitive sciences, is focused on building large neural network models. This team is pursuing an alternate scaling route for AI systems and for NI models, based on inference in probabilistic programs. Their AI-facing goal is to show that end-to-end explainable AI systems built using probabilistic programming can match and exceed the speed, robustness, and flexibility of human intelligence, using 100x-1,000x less computation than deep learning. Their NI-facing goal is to leverage new techniques for neural mapping of probabilistic programs to build and test these AI systems as computational models of perception and cognition.

  • Brain-Score

    Brain-Score website
    The Brain-Score platform aims to yield accurate, machine-executable computational models of how the brain gives rise to the mind. Its benchmarking system enables researchers to sense the alignment of their model(s) to currently dozens of neural and behavioral measurements, and it provides these models to experimentalists to prototype new experiments and make sense of biological data.

Past Missions

  • Collective Intelligence

    This mission approaches how we can optimize human-AI group decision making. How can we create super-intelligent groups that operate effectively in dynamic environments? How can highly polarized groups make decision the whole group will implement?

  • Brain Guided Intelligence Hardware

    This research mission looks at electrochemical synapses as building blocks to emulate and advance learning models. What can we learn from biological synapses to build better, more energy-efficient engineered hardware?  Science research goals include modeling the circuits that both underlie complex learned behaviors and begin to emulate and advance state of the art learning rules.