'Conscious AI' may already exist as expert receives backlash over terrifying warning

dailystar.co.uk:

artificial intelligence expert has been criticised for saying that AI used for finance, research or medicine could have already developed a consciousness.OpenAI co-founder Ilya Sutskever claimed on Twitter that: "It may be that today's largest neural networks are slightly conscious."Sutskever did not mention any specific developments but was likely referring to mega-scale neural networks, which is the language processing system built by OpenAI for translations, missing words and question answering.The expert was also unclear what "slightly conscious" means for the manmade machines, because the concept of consciousness in artificial intelligence is still considered a controversial idea.While AI has performed impressive tasks like driving cars, flying aircraft and creating an artificial voice or face, many believe machines having conscious thought is just "hype".AI researcher Toby Walsh at UNSW Sidney responded to Sutskever's claims, saying: "Every time such speculative comments get an airing, it takes months of effort to get the conversation back to more realistic opportunities and threats posed by AI."Thomas G Dietterich, an AI expert at Oregon State University in the US, said that he hasn't seen any evidence of consciousness in AI and suggested Stuskever could be "trolling" those on the social media platform.For the latest breaking news and stories from across the globe from the Daily Star, sign up for our newsletter by clicking here.He tweeted: "If consciousness is the ability to reflect upon and model themselves, I haven't seen any such capability in today's nets.

But perhaps if I were more conscious myself, I'd recognise that you are just trolling."The exact consciousness has been debated by scientists and.