IBM has developed a natural language processing advance via its Project Debater effort called Key Point Analysis that aims to use artificial intelligence to sum up crowd-sourced arguments.
The technology, led by IBM Research, is being showcased on Bloomberg TV’s “That’s Debatable” show. The show aired Friday and featured a debate on wealth distribution with US Secretary of Labor Robert Reich, former Greece finance minister Yanis Varoufakis, former US Treasury Secretary Larry Summers and Manhattan Institute’s Allison Schrager.
Noam Slonim, lead researcher for IBM’s Project Debater effort, said the goal of Key Point Analysis is to “enable AI systems to manage the human language.” “There’s a significant opportunity for using national language processing,” he said.
Primers: What is AI? | What is machine learning? | What is deep learning? | What is artificial general intelligence?
In the debut episode of “That’s Debatable,” IBM used Key Point Analysis to distill points for a debate on wealth redistribution. About 3,500 arguments were submitted on whether it is time to redistribute wealth and IBM is collecting thoughts on future debates, IBM’s technology found that 56% of the arguments were for redistribution, but 44% were against. Of that 44% against redistribution, 15% cited the moral hazards of discouraging hard work. Without the AI distilling those arguments, it’s unlikely that point would have been surfaced in a debate.
The breakdown of submissions went like this:
- 3508 submissions
- 1600 usable arguments
- 20 final key points
- 2 coherent narratives for each side.
Slonim explained that IBM’s Key Point Analysis is designed to reveal key points, generate a bullet summary and create quantitative numbers to review and a narrative. “It puts short comments into a summary that can guide decisionmakers and establish a communication channel between decisionmakers and those impacted,” he said.
IBM Watson Chief Architect Dakshi Agrawal said that the company is working to commercialize the technology for use in the Watson portfolio. “To move this to enterprise it has to be customizable, usable by line of business executives and have the ability to simplify and take knowledge with NLP,” said Agrawal. “It also needs developer tooling to use your own training data.”
Use cases could involve a company soliciting arguments from employees on a key topic and then distilling arguments down. IBM’s technology estimates quality of an argument and then presents a set of them in narrative form. The overall goal is to use AI to solve real-world problems, said Agrawal.