In the last week, Reach, home of the Daily Mirror and the Express, has begun to use bots to produce things like listicles for local news sites, such as 7 things to do with visitors to show off Newport. Where else is generative AI being used in journalism today?
An early adopter, Associated Press (AP), started automatically generating summaries of earnings reports in 2014. Within a year they were generating around 750 a month, ten times more than AP reporters could manage.
Microsoft tried to replace journalists with artificial intelligence to curate stories on the MSN and Edge homepages, but it didn’t work too well. Aberrant results included Bigfoot sightings, mermaid captures and a story about Grimes suing Elon Musk.
CNET were caught using machine-written stories that were displeasingly unpredictable and inaccurate and, according to Futurism, felt like an attempt to replace entry-level writers. They were obliged to publish lengthy corrections to some, even though they claimed they were conducting an experiment.
Following a round of redundancies and in collaboration with OpenAI, BuzzFeed now use AI to enhance quizzes and personalise some content.
At The Next Web, their late robot colleague Satoshi Nakaboto produced automated reports on cryptocurrencies, while AI is now being used to produce images for the TNW newsletter.
Where else do you suspect generative AI is being slid into journalistic output, whether blatantly or surreptitiously, based on what you are seeing, hearing or reading?
Main image: euronews.next, copyright Canva