Controversy about AI can wipe out humanity

According to New Scientist , more than a third of AI researchers, through a recent survey, believe that future decisions from AI can cause a global catastrophe as bad as nuclear war. The survey was conducted with 327 researchers with recent work in this area.

Meanwhile, Mr. Marcus Hutter’s comments were made in a recent article published in the journal AI . The main body of the report listed Marcus Hutter’s relationship with DeepMind and the Australian National University in Canberra, as well as how he used AI to predict the future of humanity.

AI then came to the conclusion that it is very likely that in the future there will be a “war” between AI and humans. At first, it seems innocuous, but it will be “death grave” if humanity loses.

After Motherboard published an article titled “Researcher at Google’s DeepMind Says AI Will Kill Humanity”, DeepMind decided to distance himself from the controversy and asked Motherboard not to mention the company. .

In a statement sent to Motherboard , DeepMind said the AI ​​engine used had a “serious error” and was being removed. Mr. Hutter’s contributions are for reference only.

” DeepMind was not involved in this, asking the author of the article to edit the content to clearly reflect the incident to the public ,” a DeepMind spokesperson wrote in an email to Motherboard .

The spokesperson said that while DeepMind is not directly involved in any of the above, the company is committed to investing in fighting the bad sides of AI and thinking “deeply about safety, ethics, and ethics.” ethics and broader social impacts of AI and the research and development of safe, effective AI models that are consistent with human values”.

DeepMind declined to comment on whether it agreed with the conclusions of the paper co-authored by Hutter.

Michael Cohen, one of the co-authors of the article, also asked Motherboard to edit his article. However, it is Motherboard ‘s editorial policy not to correct the title unless it is factually incorrect.