Faculty and students engaged in transformative AI activities at the USC Iovine and Young Academy aim to innovate the design of AI, the data used to support AI, and AI applications in society. Students engaged with transformative AI activities develop project-driven competencies in the following areas: 

Machine intelligence models and applications

Social analytics and interactions of societal processes with computing

Machine and human intelligence interaction

Computational interfaces for human performance

Machine intelligence and complex societal challenges

The design of Transformative AI starts with identifying inclusive societal transformations through the coalescing of diverse teams of stakeholders within a societal challenge space. These teams then craft AI systems that can integrate with diverse embodied human intelligence to cultivate augmented intelligence for societal impact.   Augmented intelligence requires a deep understanding of expert human performance within a complex challenge (e.g .surgery or product design) and the potential for expanding that performance through custom applications of machine intelligence. In turn, augmented human performance requires data-based modeling of augmented performance within a complex challenge and the development of capture and feedback interfaces that can facilitate the collaboration of humans and computational agents within that challenge. Augmented intelligence can apply to individual performance but has its most significant potential in situations where diverse human teams interact with computational agents to address complex, multifaceted challenges. 

Transformative AI thus flips the paradigm of first developing powerful AI systems and then contemplating the relation of these systems to society. Transformative AI also transforms the way we acquire and analyze data for training and interacting with data within augmented intelligence scenarios.  Diverse design groups can establish high-dimensional data sets that represent multiple perspectives and customizable interaction structures that are informed by the needs and preferences of each stakeholder. Transformative AI is also highly aligned with the Challenge-Based Learning methodology of IYA where intersectional technology, business, and design systems emerge from engaging in complex challenges. 

The Academy’s course, “Introduction to Machine Intelligence,” is a foundational course for Academy students interested in developing transformative AI projects and related expertise. The course introduces students to different machine learning models, as well as the data structures that underlie these models and the potential of these models to interact with human performance so as to advance augmented intelligence. The course assumes existing basic knowledge of probability and statistics. Such basic knowledge can also be gained by taking the Academy course “Introduction to Data Innovation and Analytics” or other related courses. 

Students interested in transformative AI should also complete 

At least one Academy course on interface development from this list: Designing Interactive Data Systems, Dev 1, Dev 2, Information Design, XR Design, Introduction to Physical Computing, XR Computing

An additional Academy CBL experience or Academy course that engages the design of human and machine intelligence interaction

Having completed these requirements students can enroll in the 400/500 level Academy courses in

“Transformative AI”: This course introduces modern deep learning techniques, those that are commonly referred to as AI, and how the rapid progression and growth of AI applications in society is impacting every aspect of our lives. The course will be grounded in a semester-long project to build AI solutions across a variety of modalities and will be enriched by discussions of current events around AI. Current events will explore domestic and international policy movements as well as current trends in technology investment, social acceptance of transformative AI applications, and more. 

“Augmented Intelligence”: This course introduces current machine learning techniques and analyzes the processes that each of the techniques uses for codifying and classifying human activity. It then explores how each type of computational codification can synergize with the human intelligence involved in that activity to generate augmented intelligence for societal impact. 

Data analytics for machine intelligence 

Machine intelligence and complex societal challenges


Setting transformative AI and augmented human intelligence as the overarching goals of the future of machine intelligence overcomes some of the current dualities of machine vs. human and can partially mitigate concerns about unintended consequences on society (such as the replacement of the human workforce by automation agents). The approach also provides human-centric design specs for the development of machine intelligence thus setting the stage for disruptive approaches to the future of computing. Technology companies as well as a wide range of businesses will be looking to hire developers and managers of systems where human expertise and machine intelligence interact and enhance each other. 

ACAD-220 Introduction to Data Innovation and Analytics

ACAD-230 Introduction to Machine Intelligence

ACAD-422 Cybersecurity, Privacy, and Trust in Connected Healthcare

ACAD 499 Guided Research in Transformative AI

Thanassis Rikakis

Michael Crowley

Michael Crowley, Professor of Teaching, Transformative AI

Alexander Titus, Assistant Research Professor

Aaron Siegel, Assistant Professor of Teaching

Patrick Dent, Associate Dean of Academic Affairs; Professor of Teaching

The Academy's Transformative AI intersectional area introduces students to different machine learning models as well as the data structures that underlie these models.

Design Strategy

Business of Innovation

Extended Reality

Interactive Technologies

Product Innovation

Health Innovation

B.S. Arts, Technology and the Business of Innovation

The Academy's Augmented Intelligence intersectional area introduces students to different machine learning models as well as the data structures that underlie these models.

Transformative AI

Subscribe to our Newsletter