ChatGPT users have been left surprised by the artificial intelligence bot responding to their queries in Welsh.
The innovative technology is impressive, but still not without flaws, as users found when it began to translate their English questions and its responses into Cymraeg.
The Financial Times was the first to report the issue, which users have been complaining about for some time.
It appears to only occur when using the voice function, in which users speak their requests out loud rather than typing them.
The company told the FT it was aware of a problem with its voice transcription system, Whisper.
The Welsh Government recently agreed a new ‘data partnership’ with OpenAI
Only last month, the Welsh Government celebrated a new ‘data partnership’ with OpenAI, ChatGPT’s parent, in order to help it understand the language and improve the usefulness of AI for Welsh speakers.
AI technology continues to develop at pace.
Only a few weeks ago, a recent study discovered that people find ChatGPT-produced jokes funnier than those written by humans.
The participants in the research by the University of Southern California were unaware who wrote the jokes before saying which they preferred.
They voted resoundingly in favour of the AI jokes – with 75 percent saying they were funnier than the human ones.
A recent study discovered that people find ChatGPT-produced jokes funnier than human ones
Meanwhile, a leading US medical body is encouraging doctors to use it to free up their time.
An American Academy of Family Physician study looked at how well the AI model can interpret and summarise complicated medical studies, which doctors are encouraged to read in order to stay up to date on the latest research and treatment developments in their field.
They found the chatbot was accurate 98 percent of the time – giving physicians rapid and accurate summaries of studies in a range of specialities from cardiology and neurology to psychiatry and public health.
By Daily Mail Online, August 16, 2024
