I was disappointed to read that the Guardian has followed other organisations in preventing OpenAI from harvesting its content for training (The Guardian blocks ChatGPT owner OpenAI from trawling its content, 1 September). While I understand the importance of protecting intellectual property, what worries me is the quality of information that will be left for training artificial intelligence. Whether we like it or not, generative AI will augment, complement and in some instances replace how news content is created, as well as other aspects central to our lives.
There is a need for a wider debate, which should include commercial arrangements around intellectual property. However, I for one would prefer that generative AIs are trained on information that can be both trusted and respected.
Praveen Naidoo
Barnstaple, Devon
I feel the Guardian has shot itself in the foot by banning ChatGPT from scraping its website. In doing so, it has put short-term commercial considerations ahead of its tradition of campaigning for truth and justice.
If ChatGPT is denied access to such outlets as the Guardian, the Washington Post and the New York Times, and is force-fed a diet of the Daily Mail, the Sun and the Daily Express, it is not hard to predict the outcome. Future generations will absorb the bile and prejudices of the rightwing press of our time. The work of the Murdochs, Beaverbrooks and Rothermeres will have been done for them.
AI is now an unstoppable force, and we should be concentrating on its education by ensuring the information it is fed is accurate and balanced.
Paul Sawbridge
Ealing, London
On one level, I applaud the Guardian for blocking large language model trawlers from sucking up its content. But surely it’s better that these systems are trained on honest content? I’m sure the purveyors of fake news are not blocking them. And if generative AI content overwhelms factual text on the internet, its usefulness may drown under a tide of effluent.
Ian Watson
Auckland, New Zealand
READ SOURCE