This episode currently has no reviews.
Submit ReviewAI-hallucinations, like a case last month when Google's AI summaries suggested cats had visited the moon, are a serious issue for the rapidly expanding technology.
Now researchers from Oxford University have outlined a system for flagging unreliable or inaccurate text generated by large language models. So, how does it work?
Guest: Dr Jonathan Webb, ABC Science Editor
Brainteaser (answer in the audio): If you write out the numerals from 1 to 1000, which digit occurs the most times?For more of the latest science news, look for the science section on the ABC News website and sign up to a weekly email newsletter.
This episode currently has no reviews.
Submit ReviewThis episode could use a review! Have anything to say about it? Share your thoughts using the button below.
Submit Review