AI fears ‘being overhyped’ by Silicon Valley
A STANFORD University professor who taught the entrepreneur behind CHATGPT has warned US tech giants are “hyping up fears” over artificial intelligence (AI) to suffocate potential rivals with regulation.
Andrew Ng, a former Google executive, said AI labs were using concerns around the nascent technology to raise funds so they can “fight phantoms that they conjure themselves”.
Despite concerns the technology could pose an existential threat to humanity, Mr Ng said some of the fears about AI were “overhyped”. The biggest US labs were using these concerns to push for rules that would make it harder for challengers to emerge, he said.
Mr Ng taught Sam Altman when he was studying at Stanford, before the entrepreneur dropped out in 2005.
After working for various start-ups and a career as a technology investor, Mr Altman founded Openai in 2015, which a year ago launched CHATGPT.
CHATGPT can generate text in human-like English from simple prompts, with uses such as writing emails or summarising reports.
Openai has worked on adding more capabilities to the tool, such as image and audio recognition. The sudden popularity of these AI tools and their new abilities have prompted debate among scientific experts over the dangers of the technology.
Some have warned humans could lose control of powerful AI bots, which could threaten mankind’s future.
However Mr Ng, who helped set up Google’s Brain division in 2011 and now leads online learning company Coursera, said: “I feel like being fearful that this piece of software which predicts the next word at a time could lead to human extinction is a massive leap that I just struggle to understand.”
Mr Ng said the “vast majority” of his peers were not worried about AI safety.