Meta’s chief artificial intelligence scientist Yann LeCun is no stranger to attracting controversy on Twitter.

He’s had his fair share of online feuds over the years, in which he’s stridently defended the benefits of artificial intelligence (AI).

But something he posted on Wednesday night hit a nerve for one of Australia’s leading AI researchers.

“I’m all in favour of technological advances benefiting everyone. But first, that’s a goal for politicians and democracy to achieve,” Mr LeCun tweeted.

“The mere possibility of unequal distribution is not a sufficient reason to stop the progress of science and the development of technology.”

Rebecca Johnson, who’s an expert in the ethics of AI at the University of Sydney, said the comments signalled a big problem in the AI world right now.

She feels too many people in positions of power at tech companies don’t think it’s their responsibility to develop ethical AI that best serves the needs of society.

Yann LeCun has attracted controvery online over the years, for continuing to defend the benefits of artificial intelligence. ()

“People like Yann LeCunn divorce themselves from accountability and say it’s up to the politicians and got nothing to do with me,” she said.

“These companies say ‘Oh yeah, we believe in ethics’. OK, well, then enforce it at all levels of your company, including your chief scientist and chief CEO.

“We have got to move away from these individualistic rock star perceptions of, ‘You can’t touch me’.”

The rise of generative AI

Generative AI – which uses algorithms to create text, images, and audio – has exploded this year.

The ChatGPT model has taken the world by storm, with people using it to write a resume, explain complex concepts, make a diet plan or even get relationship or psychological advice.

This rise has raised fears of students cheating on assignments, with five Australian states banning the technology.

Earlier this month a group of AI experts even called for a six-month pause on developments, citing “risks to society”.

Google and Microsoft have launched AI chatbots, with Google CEO Sundar Pichai describing AI as being good at “synthesising insights for questions where there’s no one right answer”.

Rebecca Johnson has been concerned with what she’s found in the evaluation of AI. ()

What is the right answer?

The ethics of how these algorithms are built, and who deems them a success, is what Ms Johnson, a PhD researcher, is concerned about.

Put simply, evaluation metrics are used to determine things like whether an AI program has a bias against women, if it’s accurate or if it’s disseminating hate speech.

Source link