We live in a rapidly changing world – one in which new technologies like artificial intelligence (AI) are changing the way humans and machines live, work, and think together. Underlying this paradigm shift is a vast dependence upon troves of data - easier now more than ever to collect, yet their analysis still requires careful and bold theoretical reflection.

The Center for Theoretical Behavioral Sciences at UC Irvine is engaged in the pursuit of robust theories to transform the social, behavioral and information sciences to meet this challenge. As we address and advance our understanding of AI’s impact on how we understand and operate in our minds, bodies, societies, and larger environments, we need new questions. We need new theories.

CTBS is the site of this theory building. The center explores the delicate balance between data and theory focusing in particular on the interplay of consciousness and computational processes. Center collaborators and audiences are experts in computational and mathematically oriented research. They are social scientists poised to crack the next developments in artificial intelligences through new theories of intelligence, consciousness, and computation.

A rich history

CTBS carries on the rich history of the Institute for Mathematical Behavioral Sciences (IMBS) which for three decades fostered a fruitful and symbiotic relationship between mathematics and the behavioral and social sciences. IMBS laid the groundwork for new mathematical approaches to psychology, evolutionary systems, decision making, language, social contracts and social constructs. Important work in genetic algorithms, crowdsourcing, and probability supported the development of machine learning algorithms driving today’s AI.  During the 30 years of its operation, IMBS focused on deriving descriptions of complex mental functions from simple mathematical principles called axioms as envisioned by the fathers of mathematical psychology. More recently, however, axiomatic approaches have been blended with computational modeling of mental functions. But for these computational models to be considered explainable, which means that a model can be explained in a scientific paper or in a scientific talk, these models must contain a small number of abstract principles that describe what is being computed and how. Providing a function that maps inputs to outputs may not be sufficient. The small size of human working memory, which can hold and manipulate only a few items/concepts at a time, imposes severe constraints on what can be considered explainable. Artificial memory would seem to remove those constraints. But if AI is intended to interact with natural intelligence, that is, with us humans, the theories will have to be expressible with only a few mathematical principles. This makes it clear that the new Center for Theoretical Behavioral Sciences will be a natural extension of IMBS.

Topical expertise to meet the challenges of tomorrow

CTBS spans a range of topics. On one end, researchers study cognitive functions of a single mind: perception, memory, thinking, decision making, learning, concept formation. The next thrust involves interactions between two people - conversations, cooperation, competition - including the formation of models of the other person's mind (theory of mind). Still another thrust involves groups of people like those analyzed in sociology (economic sociology, population processes, social networks, religions, human-AI systems) and political science (democratization, elections, voting). Finally, even larger groups, like those that are analyzed in economics, are also explored – such as micro and macroeconomics, political economy, public economics, and urban economics. And many of these behavioral functions span more than one level. For example, problem-solving and games refer to individual people, as well as to groups of people (the wisdom of the crowd). These levels now also entail computational agents, whether we ascribe intelligence to them or not. Their increasing ubiquity means that soon they will no longer be separable from human interactions.

A brilliant future

The paradigm shifts CTBS seeks to inspire are not mere philosophical exercises. Natural and artificial intelligences now interact daily—from computer vision and human motor response in “self”-driving vehicles to the artificial brain meeting the natural brain in the act of reading and writing. Society as a whole is proceeding on a pragmatic or practical level with the interdigitation of natural and machine intelligences. CTBS provides the space for theoretical reflection and refinement to understand and transform that interface.