WF-JZWMQ4
All Image Captions Produced are Violent
MIT Media Lab researchers create AI-powered "psychopath" named Norman by training a model on "dark corners" of Reddit.
- Date it happened
- 2018-04-02
- Organisation involved
- MIT Media Lab
- Product, system or model
- Norman is an AI system trained on the captions of violent and graphic content on the Reddit thread r/watchpeopledie. Its purpose was to provide answers to Rorschach test prompts that would demonstrat
Where this came from
- news.avclub.com
- iflscience.com
- money.cnn.com
- curiosity.com
- theverge.com
- extremetech.com
- reddit.com
- internetofbusiness.com
- hothardware.com
- analyticsindiamag.com
- businessinsider.com.au
- popularmechanics.com
- news.sky.com
- newsweek.com
- cnet.com
- independent.co.uk
- geek.com
- phys.org
- livescience.com
- zdnet.com
- motherboard.vice.com
- digitaltrends.com
- livemint.com
- inc.com
- rollingstone.com
- bbc.co.uk
Attribution
This incident was imported from AI Incident Database and is used under CC BY-SA 4.0. Our additions to it — the structured fields, the translation, the checks against other reports — are published under the same licence.
This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.