OpenAI will use FT media to train AI systems

Under the agreement, ChatGPT users will receive summaries and excerpts of The Financial Times content and links to articles 
 
 
 
OpenAI

Financial Times signed a contract with the ChatGPT project. OpenAI licenses its content. It is used to train artificial intelligence systems. 
 
FT will receive anonymous payments as part of the agreement, which is the final agreement between OpenAI and the publishers. 
 
In the program, ChatGPT users  receive links to articles as well as summaries and quotes from FT journalists in response to replies where appropriate. FT Group CEO John Ridding said it was "great" that AI companies were paying advertisers for their tools. 
 
The New York Times is suing OpenAI and its main investor, Microsoft, for using its content to train large-scale language models, technologies that power conversations, such as ChatGPT. 
 
 
 
“OpenAI understands the importance of transparency, accountability, and pricing—all of which are important to us,” Ridding said. “It is also clear that these products contain reliable markets for the benefit of consumers.” 
 OpenAI  signed a similar agreement with Germany's Axel Springer, which owns American news agency Associated Press, France's Le Monde, El PaĆ­s Prisa Media, and published the Bild document. Brad Lightcap, OpenAI's chief operating officer, said it's important for the company to "represent good media because these products are successful." 
 
“As with any transformative technology, there may be great advances and great problems, but what is impossible is turning back the clock,” he said. Chats like 
 
ChatGPT are at the forefront of advances in artificial intelligence generative word-of-mouth technology that can generate original text, images or audio from simple handwritten text. 
 
However,  models using these tools are trained on large amounts of data  from the Internet, including copyrighted text. 
 
 
 
Getty Images, owner of one of the world's largest photo collections, is suing the company behind the Stable Diffusion video in the UK on  similar grounds. Last year, the FT's editor, Roula Khalaf,  said in an open letter that her editorial team  would "seek to improve AI tools" and train journalists to use AI-generated "stories". However, with the help of technology, this will be kept transparent and the media will be written by humans, he said. 
 
Niamh Burns,  senior research analyst at Enders Analysis, said the FT's "high-quality" content will drive OpenAI and improve chatbot results. 
 
“The tools will be of real value for OpenAI in terms of deploying a chatbot that can provide real, high-quality answers to questions from users who need up-to-date information,” he said. “FT's products and strengths also mean they are at the forefront of any substitution effects that  AI products bring to consumers,” he said.

Comments




    Font Size
    +
    16
    -
    lines height
    +
    2
    -