What is the compelling question or challenge?
Enabling the next leap in advanced computing approaches, including artificial intelligence, through incorporation of neurodiverse modes of thinking.
What do we know now about this Big Idea and what are the key research questions we need to address?
A team of computer scientists, neuroscientists, and data scientists at Vanderbilt are developing artificial intelligence (AI) algorithms that "look at" geometric patterns in tests of fluid intelligence such as Raven's Progressive Matrices, deciding which missing shapes would be most likely to fit in. These AI algorithms currently perform about as well as a human 17-year-old would, and they're only getting smarter, thanks to a study of the way certain people on the autism spectrum see the world.
Inspired by the writings of perhaps the most famous person on the spectrum, Temple Grandin, our team has been working on code that emulates the kinds of image-based thinking that Grandin used to design complicated farm equipment. The result is a form of AI that allows researchers to study a model of human cognition, determine how it problem-solves and then tweak it to perform better.
Most humans think in a combination of lots of different things. We think in words, we think in pictures, we think in smells and feelings. What we see in some people with autism is that they’re very much on the visual side. Temple Grandin talks about how she thinks really strongly with images, and certain kinds of language-based thinking are a little more difficult for her.
Our team of computer scientists and neuropsychologists is working to turn that system of thinking into code. Computers are programmed to solve various different types of human cognitive tests (such as Raven’s Progressive Matrices) that normally involve human subjects visually looking at a series of patterns. The new AI approach works only with "pictures" and pictorial manipulations, not symbolically or mathematically as in traditional AI approaches.
Major questions to be addressed include:
- What are the varieties of ways in which humans think and approach problems?
- What are the differences in approaches to problem solving, reasoning, ideation, and discovery employed by neurodiverse and neurotypical individuals?
- How can these differences inform novel design of AI and other advanced computing approaches?
- How can understanding these differences, and the development of computational tools based on them, support and improve quality of life for all people?
Why does it matter? What scientific discoveries, innovations, and desired societal outcomes might result from investment in this area?
Advanced computing approaches such as artificial intelligence will ultimately be limited in the possible scientific discoveries, innovations, and societal outcomes so long as we limit the forms of intelligence upon which they are based. We know that the cognitive diversity of the human mind can be essential to discovery and innovation, in surprising ways that might not otherwise be imagined. It stands to reason that learning to harness computational cognitive diversity will lead to scientific discoveries, innovations, and societal outcomes that we cannot yet imagine.
AI can be improved by looking at intelligence that isn’t neurotypical. It can draw from a richness of variety in human intelligence, making it more creative and effective. At the same time, researchers can use artificial intelligence to develop tools to enable neurotypical people to benefit from otherwise inaccessible cognitive modes, as well as to develop more effective supports for neurodiverse people.
Finally, by acknowledging the power of neurodiversity in our computational methods, we advance the societal benefits of a strengths-based (as opposed to deficit-based) approach to neurodiversity more generally.
If we invest in this area, what would success look like?
Success in this area would include:
- A substantially broadened array of cognitive models used in advanced computing approaches such as artificial intelligence.
- An expanded understanding of the types of cognitive models most suitable to different types of problems.
- Increased ability of researchers specifically and society generally to harness the data revolution.
- Increased recognition of, and broadened participation of, neurodiverse individuals and neurodiversity inspired approaches in science and engineering.
- Increased discovery and innovation outcomes resulting from these approaches.
Why is this the right time to invest in this area?
Advanced computing approaches such as artificial intelligence are becoming an ever more present feature of society and the ways in which we live and work and learn. At the same time, there is a rapidly growing awareness that neurodiversity is not necessarily or only a question of deficits, but also can represent differences and unique strengths and capabilities. In many areas today, the maximum discovery and innovation and benefits to society are to be found in the convergence of the technological and the human, including especially approaches that leverage the power of difference and diversity.
References
Reference #1
Reference #2
Reference #3
Show MoreShow Less