Algorithm probes how AIs reason

Quartz  explores an algorithm devised by CS Prof. Trevor Darrell, L&S CS undergraduate student Dong Huk Park, CS grad student Lisa Anne Hendricks, and postdoc Marcus Rohrbach, along with researchers in the Max Planck Institute for Informatics,  in an article titled “We don’t understand how AI make most decisions, so now algorithms are explaining themselves.” Engineers have developed deep learning systems that ‘work’ without necessarily knowing why they work or being able to show the logic behind a system’s decision.   The algorithm uses a “pointing and justification” system, to point to the data used to make a decision and justify why it was used that way.