professor bengio, welcome to democracy now!we talk about an issue i think most people cannot begin to comprehend. if you could start off by talking about why you have signed this letter warning of extinction of humanity, but talk about what ai is first. >> thank you for having me. thank you for talking about this issue that requires more awareness. the reason i signed this, like geoff, i changed my mind in the last humans. what triggered this for me is interacting with chatgpt and seeing how far we had moved much faster than i anticipated. i used to think that reaching human level intelligence with machines could take anymore decades. if not centuries. the progress of science seemed to be, well, slow. as researchers, we tend to focus on what does not work. but right now, we have machines that pass what is called the current test, which means they can converse with us and it could easily as as being humans. that was supposed to be a milestone for human level intelligence. i think they are still missing a few things but that kind

professor yoshua bengio said that he worries about bad actors misusing the tech.h academics and the companies building ai agree that regulation is the best way to try to maximise its benefits and minimise the risks. chatgpt, for example, was unheard of this time last year. yann lecun is confident that ai will be a useful tool rather than a master, and he heralded a new renaissance for humanity. chatgpt couldn't have put it better. zoe kleinman, bbc news. a renaissance maybe but the bbc has been hearing from one uk worker who has already lost his job to artificial intelligence — along with his colleagues. dean meadowcroft and his team of marketing copywriters were asked by their bosses to trial working with the ai app chatgpt before christmas. four months later, the entire team were laid off. dean says he didn't see it coming. funnily enough, in a dark irony, i didn't, quite at first. i thought there was a lot of doomsday prophesying online. and i thought, like everybody else, i'd seen everything from ai—generated novels and poems to ai—generated content and i look

ai pioneer yoshua bengio recently warned of the potential dangers of the misuse of the technology.latforms for defence. it's working on assisted project for us air force f—16 pilots, as well as autonomous aircraft like this, called the vbat, and small drones and quadcopters. we have deployed ai capabilities since 2018. they are currently used by the american special forces units, so unfortunately we cannot talk much about the application. contemporary drones�* ability to fire weapons is controlled by human operators. could ai technology one day lead to these machines making the decision to destroy targets by themselves? the us department of defence has outlined a very strict ethical principle around ai, and one of those is about having governance and responsible use of those mechanisms, and so i don't see a world where we don't have humans at least observing and interacting with these systems to make those decisions. putting visions of apocalyptic automota aside, these technologies could help the planet too. the hydrus diving robot from australian company advanced navigation can be