Siri co-inventor says the current rollout of AI is ‘human trial without consent’ | 7.30

hey Siri hey Siri hey Siri well it's been a dream of us for a long time in AI to create a virtual assistant I'm Tom Gruber I was the CTO and co-founder of the group that created Siri hi Tom my old friend Siri was very playful and and that was by Design Siri is now just one of many virtual assistants that can help manage Our Lives Okay Google hi Bixby hi Amy how can I help Alexa it's game day Siri co-inventor Tom Gruber is a strong believer in humanistic AI making intelligent machines that serve humans rather than corporations I think Siri brought AI to the mainstream it's used almost a billion times a day or roughly a billion times of that Siri was tested heavily before being released to the public but Tom Gruber says new AI apps like Microsoft funded chat GPT or Google's Bard have been trained on the knowledge of the internet and are now being tested on the public what they're doing by releasing it is they're getting data from humans about the kind of stuff that would that normal humans would do when they get their hands on such a model and they're like learning by trial and error there's a problem with that model that's called um a human trial without consent Microsoft told 730 in a statement that Real World experience and feedback is critical and can't be fully replicated in a lab but Gruber and thousands of other scientists researchers and coders across the world have signed a public letter declaring the current AI race is not safe it really wasn't about stopping research it was all about saying we've now uncovered this new technology which is more powerful than we thought so let's give ourselves enough time to figure out a framework so that we can keep this technology safe and and on the side of humans I would describe the way that the technology is being released is worrying and they're rushing to do things that I don't think are very responsible in the beginning computers were just Advanced calculators processing numbers as processing power doubled every two years it led to machine learning where computers were taught to perform a specific task like chess or go processing up to 200 million moves a second then with even more power the focus shifted to artificial intelligence and a large field of learning which includes the ability for a machine to teach itself lessons from interactions with humans or objects by neural networks like autonomous driving a driver that's reimagining transportation for all of us or simply opening your phone using your face [Music] processing power has exploded allowing AI to operate Way Beyond human comprehension I think that's the worrying idea here is that it is the fastest growing app on the planet Chachi PT was used by a million people in five days 100 million people at the end of the first month and now six months later it's in the hands of a billion people we've never had Technologies before where you could roll them out so quickly and so even small harms multiplied by a billion people cause could be cause of significant concern Toby Walsh is Chief scientist in artificial intelligence at the University of New South Wales where in a single day he faked his own voice and image using AI but you'll begin to understand the ethical considerations of technology is not always a yes or no ads well thank you hey it AI trained voice programs are being used to trick parents into thinking they are actually talking to their children he said I've been in an accident in Toronto and I had a woman that's pregnant and she's been sent to the hospital with multiple injuries and he said Mom and Dad I need your help except it wasn't her son it was AI police in Canada or investigating cases where AI trained fake voices were used to scam money from parents who believed they were speaking to their children it's hard to think of a positive use of the sorts of technologies that can clone people's voices just to clarify it only takes a few seconds now to clone someone's voice could just ring up your answer phone and I can record your voice and now I can speak just like you as AI trained chat Bots spread to mainstream apps like Snapchat so too the concerns about serious harms a chatbot that's available in Snapchat it's available now to three quarters of a billion people and someone pretending to be a 13 year old um had a conversation with the chatbot which encouraged her to to meet up with a 30 year old man and did not intervene to her being groomed it suggested ways that she could hide the relationship from her parents as the father of a 13 year old child I'm an AI researcher I'm deeply distressed that we think that it's responsible to put that technology in the hands of the masses without any guardrails Snapchat told 730 in a statement if users experience any inappropriate or incorrect responses from their my AI chatbot they encouraged them to report it the really scary thing is they can pretend to be a human really well and that that's one of those things that scares people with psychology background or something we're like wow humans are already pretty gullible I mean a lot of people would talk to Siri for a while mostly for entertainment purposes but there are a lot of people who get sucked into such a thing and that's the that's the one of the really big risks arguments are now occurring about who is responsible the user or the AI makers there's this idea that the tech in the tech space they're not as responsible as in the physical space tech companies need to be held responsible imagine if you know someone had a car which every so often spontaneously caught fire and someone died there would be lawsuits of course there would be this is why we have liability if you build an unsafe product if you release something that causes harm to other people you should be held accountable to this konaleihi is a london-based coder whose company is working to design AI to be safe and controllable from the ground up so mostly when I talk about safe AI I mean AIS that truly do what we want to do and we can rely on them to not do things we don't want them to do he has millions of dollars in investment from Venture capitalists who share his concerns there is currently more regulation on selling a sandwich to the public than there is to building completely novel powerfully AI systems with unknown capabilities and intelligence and releasing them to the general public widely onto internet the government at the moment has no regulation in place whatsoever about this openai the creators of chat GPT say they welcome regulation but despite all the medical educational and Industry benefits its technology provides open ai's CEO says he still has fears I think people should be happy that we're a little bit scared of this I think people should a little bit scared a little bit you personally I think if I said I were not you should either not trust me or be very unhappy I'm in this job raise your right hand Altman doubled down on that sentiment yesterday while testifying to the U.S Congress if this technology goes wrong it can go quite wrong uh and we want to be vocal about that we want to work with the government to prevent that from happening if we can't believe our senses and use our inherited ability to detect whether that thing is fake or real then a lot of things that we do as a society start to unravel it's not that there's one specific way things go wrong and only one way how unleashing intelligent autonomous powerful systems onto the internet that we cannot control and we do not understand there's billions of ways this could go wrong [Music]

Channel: ABC News In-depth

Views: 78,501

Keywords: abc news, australian news, abc news indepth, documentaries, long-form journalism, chatgpt, ai, artificial intelligence, siri, development, regulation, humanity, trial, concern, fear, afraid, computer, smart, chat gpt, Tom Gruber

Id: oSfgawwUy-s

Length: 8min 58sec (538 seconds)

Published: Fri May 19 2023