IBM and University of Michigan to Develop an AI that can Interact Naturally with Humans

Early this year, IBM and the University of Michigan (U-M) launched a $4.5 million partnership to develop an AI that can interact with humans more naturally and effectively.

Dubbed as “Project Sapphire,” IBM and U-M will attempt to develop in the next couple of years an AI academic adviser for undergraduate engineering and computer science majors at the university.

“By partnering with the University of Michigan, we have an enormous opportunity to apply AI technologies in new ways and transform human-machine communication,” David Nahamoo, IBM Fellow and Chief Technologist for Conversational Systems of IBM Watson, said in a statement.

“Natural conversations bring in so many different aspects of human intelligence –

knowledge, context, goals and emotion, for instance. In many ways, to build a versatile conversational system is a grand challenge for artificial intelligence,” Satinder Singh Baveja, Professor of Computer Science and Engineering and Director of U-M Artificial Intelligence Lab, said.

According to IBM and U-M, this AI academic adviser will not give hard-coded responses and scripted replies to students. Rather, this academic adviser will apply probabilistic and statistical methods of reasoning in understanding context and conditions. This AI adviser will be trained using recorded real-life teacher-student conversations.

The Project Sapphire proponents, IBM and the U-M Artificial Intelligence Lab, stressed that the AI adviser is not meant to replace human academic advisers.

One of the researchers of Project Sapphire – Emily Mower Provost, Assistant Professor of Computer Science and Engineering at the university – will study the emotional cues of students in order to enable the AI adviser to recognize when students need a human adviser even if they do not verbally ask for one.

Project Sapphire is another project of IBM Watson. IBM describes Watson as “a technology platform that uses natural language processing and machine learning to reveal insights from large amounts of unstructured data.”

The resulting AI of Project Sapphire, according to IBM and the U-M Artificial Intelligence Lab, could be “embedded into cognitive systems across many industries to improve how they learn and codify human expertise, understand a user’s intent and context, and deliver appropriate responses that direct conversations towards a stated goal.”