BBC Pidgin of Tuesday, 30 May 2023

Source: BBC

More warnings come from sabi pipo say AI fit finish human beings

AI na di short form of Artificial Intelligence AI na di short form of Artificial Intelligence

Artificial intelligence fit lead to di extinction of humanity, experts - including di heads of OpenAI and Google Deepmind – don sama warning.

Plenti pipo don support one statement dem publish on di webpage of di Centre for AI Safety.

"Mitigating di risk of extinction from AI suppose be global priority alongside oda societal-scale risks like pandemics and nuclear war," e tok.

But odas say dem don too overblow di fears.

Sam Altman, Chief Executive of ChatGPT-maker OpenAI, Demis Hassabis, Chief Executive of Google DeepMind and Dario Amodei of Anthropic don support di statement.

Dr Geoffrey Hinton, wey issue one earlier warning about risks from super-intelligent AI, don also support di call.

Yoshua Bengio, professor of computer science for di University of Montreal, also sign am.

Dem dey often describe Dr Hinton, Prof Bengio and NYU Professor Yann LeCunn as di "godfathers of AI" for dia groundbreaking work for di field - wey dem jointly use win di 2018 Turing Award, wey recognise outstanding contributions for computer science.

But Prof LeCunn, wey also dey work for Meta, don tok say dem dey overblow dis apocalyptic warnings.

Many oda experts similarly believe say fears say AI go wipe out humanity dey unrealistic, and na distraction from issues like bias in systems wey don already be problem.

Arvind Narayanan, one computer scientist for Princeton University, don previously tell BBC say disaster scenarios wey dey like science fiction dey unrealistic: "Current AI no dey anywia near capable enough for dis risks to materialise. As a result, na distracted attention away from di near-term harms of AI".

Dem beg for pause

Media coverage of di supposed "existential" threat from AI don go up since March 2023 wen experts, including Tesla boss Elon Musk, wey sign one open letter dey beg dem to stop di development of di next generation of AI technology

Dat letter ask if make we "develop non-human minds wey fit eventually outnumber, outsmart, make us out of date and replace us".

Di new campaign however get one very short statement, dem design to "open up discussion".