Forum

Reporting on AI - What annoys you the most?

  • Christian Simon profile
    Christian Simon
    16 August 2018
  • 3 comments
    Total votes: 3

This is a question to all, but specifically to the people in this forum actually building or researching "AI". When you read journalistic coverage of "AI", what are the most common mistakes you come across, what do you think is most often mischaracterized and what do Journalists get wrong most of the time? I am not talking about special interest magazines or university publications, but the reporting in standard, general interest newspapers, TV-stations or news websites. 

I am especially interested in the topic of Hype: Do you think Journalists fall for baseless hype too quickly? Or do they still underestimate the effect "AI" will have on our everyday lives? If you could change the coverage of the topic any way you wanted, what would it look like? 

Tags: 
journalismQuestioncoverageHype

Comments

Log in to post comments
Bjoern Juretzki's picture
Bjoern Juretzki,
20/08/2018 10:58

I find the media coverage of AI a bit worrying. Most importantly, I have the impression journalists fail to fully understand what AI is and realistically can and cannot do. There is really a lack of understanding of how advanced the technology currently is and how limited its applications often are.

In this context, I can highly recommend this Guardian article on the media discourse on AI being unhinged: https://www.theguardian.com/technology/2018/jul/25/ai-artificial-intelligence-social-media-bots-wrong

Total votes: 1
Mona Sloane's picture
Mona Sloane,
17/08/2018 16:02

I agree with Miika, intelligibility and transparency are key for socially just technology, not just AI, and well-researched and critical journalism is key for achieving that. In that sense, it plays a key role for society, but that is old news. As for AI reporting: I find that generally, technocratic framings and a tendency to fall for the hype are problematic, as is just repeating what companies put out through their press releases. More investigative journalism on AI innovation is needed that looks at the people, not just the technology, over a longer term. And conversations with researchers need to be had more, not just with corporate spokespeople or politicians. Perhaps opinion pieces would be good for this. Cathy O'Neil (author of 'Weapons of Maths Destruction') writes terrific opinion pieces for Bloomberg. There also are very good tech newsletters out there that could be drawn from, for example MIT Tech Review. And then most AI/tech researchers share their research on Twitter and are actually having discussions about it there. Tuning into that can help giving explainable AI more substance. 

Total votes: 3
Miika Blinn's picture
Miika Blinn,
17/08/2018 09:19

Would be nice if coverage focused more on explainable AI: What possible today? What solutions is are reserchers working on? I think this an area too much neglected in the public debate, though a vital one: If you want to promote trust in the systems, explainablity/transparency of the black-box-decisions is key. It’s ethically highly questionable if AI decisions, that have significant impact on people’s life are not comprehensible and verifiable (e.g. by external experts in an audit). 

Total votes: 3