Explainable AI – News, Research and Analysis – The Conversation

The use of numerous AI systems is risky, because they cannot explain how they reach their decisions. ShubhamDhage/Unsplash

December 3, 2024

Opening the black box: how ‘explainable AI’ can help us understand how algorithms work

David Martens, University of Antwerp and Sofie Goethals, University of Antwerp

AI systems can appear to be black boxes – often, even experts don’t know how systems reach their conclusions. The nascent field of “explainable AI” aims to address this problem.

Visualising wealth and poverty through AI. Authors

November 27, 2023

How AI ‘sees’ the world – what happened when we trained a deep learning model to identify poverty

Ola Hall, Lund University; Hamid Sarmadi, Halmstad University, and Thorsteinn Rögnvaldsson, Halmstad University

Researchers fed an advanced AI algorithm with satellite photographs to see if it could identify areas of poverty and it interpreted the data through abstract images.

Do you trust AI systems, like this driverless taxi, to behave the way you expect them to? AP Photo/Terry Chea

September 13, 2023

Why humans can’t trust AI: You don’t know how it works, what it’s going to do or whether it’ll serve your interests

Mark Bailey, National Intelligence University

People can trust each other because they understand how the human mind works, can predict people’s behavior, and assume that most people have a moral sense. None of these things are true of AI.

Counterfactuals are claims about what would happen, were something to occur in a different way. For instance, we can ask what the world would be like had the internet never been developed. Shutterstock

January 26, 2023

Philosophers have studied ‘counterfactuals’ for decades. Will they help us unlock the mysteries of AI?

Sam Baron, Australian Catholic University

AI models are increasingly being used to make important decision about people’s lives – just take Robodebt. Yet the complexity of these systems means we hardly understand them.

May 24, 2022

When self‑driving cars crash, who’s responsible? Courts and insurers need to know what’s inside the ‘black box’

Aaron J. Snoswell, Queensland University of Technology; Henry Fraser, Queensland University of Technology, and Rhyle Simcock, Queensland University of Technology

Our research on a recent Australian court case shows how experts and lawyers can overcome opaque AI technology. But regulators could make it even easier, by making AI companies document their systems.

Understanding how artificial intelligence algorithms solve problems like the Rubik’s Cube makes AI more useful. Roland Frisch via Wikimedia Commons

January 13, 2021

How explainable artificial intelligence can help humans innovate

Forest Agostinelli, University of South Carolina

AI algorithms can solve hard problems and learn incredible tasks, but they can’t explain how they do these things. If researchers can build explainable AI, it could lead to a flood of new knowledge.

A robot and human interacting. Tommy Ton, tontommy.com

February 26, 2020

People prefer robots to explain themselves – and a brief summary doesn’t cut it

Mark Edmonds, University of California, Los Angeles and Yixin Zhu, University of California, Los Angeles

Having robots and other AI systems tell people what the AIs are doing makes them more trustworthy. A study finds that how a robot explains itself matters.

Articles on Explainable AI

Displaying all articles