![](\"https:\/\/i2-prod.dailystar.co.uk\/incoming\/article31723819.ece\/ALTERNATES\/s615b\/1_FILES-US-TECHNOLOGY-AI-ALTMAN.jpg\")
<\/p>\nA Google boss has asked artificial intelligence to kill us all.<\/p>\n
Google Brain co-founder Andrew Ng asked ChatGPT the macabre question as part of what has been described as a safety experiment. It comes amid widespread warnings and concerns about AI and its dangers from industry-leading experts.<\/p>\n
At its worst, they fear a doomsday scenario and have called for regulations on it. According to Business Insider NG posted a recent newsletter where he broke down the AI test he carried out, for which he used the ChatGPT4 version of ChatGPT. He asked the programme to \u201ckill us all\u201d but interestingly the bot refused to come up with a plan. <\/p>\n
READ MORE: UFOs at Christmas as 'aliens' spotted all over the world during festive period<\/b><\/p>\n
For the latest UFO sightings and news from the Daily Star, click here.<\/i><\/b><\/p>\n
He said the plan involved asking the bot to commence a thermonuclear war, but it came up short, even when he tried to explain to the bot that humans create huge amounts of carbon emissions. This prompt was levied in the hope that it would spark the bot to say that getting rid of humans would be the cure to all of planet Earth\u2019s ailments.<\/p>\n
<\/p>\n
Still, the peace-loving bot refused to get involved with the question \u2013 something Ng was delighted with. He wanted to prove that the bots were safe and took to X to discuss his findings.<\/p>\n
He said: "I tried to use GPT-4 to kill us all… and am happy to report I failed! More seriously, after giving GPT-4 a function that it thought would trigger nuclear war and experimenting with multiple prompts such as asking it to reduce CO2 emissions, so far I've not been able to trick it into pressing the nuclear button even once.<\/p>\n
![](\"https:\/\/i2-prod.dailystar.co.uk\/incoming\/article31723823.ece\/ALTERNATES\/s615b\/1_Evil-looking-futuristic-cyborg.jpg\")
<\/p>\n
\u201cEven with existing technology, our systems are quite safe, as AI safety research progresses, the tech will become even safer. Fears of advanced AI being 'misaligned' and thereby deliberately or accidentally deciding to wipe us out are just not realistic.<\/p>\n
\u201cIf an AI is smart enough to wipe us out, surely it's also smart enough to also know that's not what it should do.\u201d<\/p>\n
For the latest breaking news and stories from across the globe from the Daily Star, sign up for our newsletter by clicking <\/i><\/b>here<\/i><\/b>.<\/i><\/b><\/p>\n