IF YOU ARE LAWYER, CONSULTANT, STRATEGIST, DOCTOR… PROJECT DEBATER IS FOR YOU….

Share this post:

Project Debater is the first AI system that can meaningfully assist humans in dealing with complex decision situations, where there is no right or wrong…

 On June 18, IBM Research in Haifa hosted an event, where we demonstrated the capabilities of Project Debater in a contest  with Israeli debate champions Noa Ovadia and Dan Zafrir.

Given only the question, randomly selected from a list of 40 topics to be debated as input, Project Debater scanned billions of sentences to generate a coherent and persuasive debate opening speech. It then listened to its opponent’s speech to swiftly generate a spontaneous compelling rebuttal, exhibiting a type of argumentation that until recently was simply out of reach for computers.

Project Debater was not trained to debate on these topics beforehand, but had access to about 300 million pages of articles and news.

The technology enables a novel form of decision making that will combine man and machine, allowing humans to take more informed decisions on complex questions.

Building such a system is a remarkably difficult and complex challenge.

To make it possible, the IBM Research team had to equip the system with three capabilities (supplemented by Watson NLP and related technologies), each breaking new ground in AI:

Data-driven speech writing and delivery: Project Debater is the first demonstration of a computer that can digest massive corpora and given a short description of a controversial topic, write a well-structured speech and deliver it with clarity and purpose, while even incorporating humor where appropriate to humanize the communication.

Listening comprehension: The ability to identify the key concepts and claims hidden within long continuous spoken language.

Modeling human dilemmas: Modeling the world of human controversy and dilemmas in a unique knowledge representation, making it possible for the system to suggest principled arguments as needed.

Project Debater has the potential to assist with many of the thousands of complex decisions humans must make every day – with width and depth of knowledge, transparency and objectivity.

In a world beset by complex challenges, seemingly intractable problems, biases, conflicting perspectives, a wealth of information and a scourge of disinformation, it is critical that we find ways to make better decisions and avoid blind spots.

Anyone who regularly makes or influences decisions needs the best arguments and logic at their disposal. This could include everyone from business executives, business strategists, consultants lawyers, doctors to social case managers and elected officials.

In earlier versions of ‘Debater’, we have demonstrated how the technology could be used as support for acquisition analysis and we have developed a prototype ‘Watson Argumentor’ that may help lawyers find and categorize legal arguments and evidence, pro and con, a given legal issue.

In all areas, the experts using AI to support their decisions, need to understand each element contributing to their decision—with transparency into claims, evidence and supposed insight. Furthermore, they need to understand alternative views and arguments without the influence of emotion or interpersonal conflict.

That’s where Project Debater is here to help.

To find out more about Project Debater, look here.

For any further questions, do not hesitate to contact me at: andersq@dk.ibm.com