Mo Gawdat, centre, in Grand Central Station in Chasing Utopia. Photograph: Atlantic StudiosView image in fullscreenMo Gawdat, centre, in Grand Central Station in Chasing Utopia. Photograph: Atlantic StudiosReviewChasing Utopia review – renegade Google exec Mo Gawdat searches for ethical AI in alarming insider warningDelivering much information about the scale of what’s coming, documentary also follows Gawdat’s campaign to get the programs with empathyAnother day, another warning about AI; vis-a-vis the reality we all know, this has roughly the same reassuring effect as a plane fuselage ripping off mid-flight. Starting off with familiar criticisms, such as putting the world out of work and handing over power to tech barons, Alex Holmes and Lina Zilinskaite’s film blasts an concentrated stream of AI concerns in its 83-minute runtime. By the time it is talking about current efforts to create computers out of human brain cells, potentially integrable into our own craniums, and implying this might be a good thing, it is (ironically) hard to know how to process all of this.The Cassandra at the film’s centre is Mo Gawdat, former chief business officer at Google X, now a touring cautionary voice trying to get the world to listen about the perils of AI. Once overseeing advanced projects for the tech giants, his biggest moonshot lies ahead: to introduce a moral dimension into a tech race that looks increasingly like the frenzied season finale of late capitalism. He talks about feeling parental pride in watching Google’s AI-driven robotic arms learn to grasp objects, as children do. And he feels that humanity’s capacity for benevolence is exactly the training resource needed by neural networks in order to prevent the technology ushering in catastrophe.The parental angle is personal for Gawdat: he quit Google in the aftermath of the tragic death of his son after a botched appendix operation. So he has an evangelical urgency addressing AI’s current human shortcomings: how it’s enabling a kind of digital narcissism through hyper-optimised social media and porn, facilitating mass surveillance and automated warfare, and evolving in an exponential growth curve that may soon escape human control (Geoffrey Hinton chips in here). The tech bros – of course not interviewed here – don’t seem too bothered. The uncanny-valley affect of Mark Zuckerberg and Sam Altman suggest that alien superintelligence has been 3D-printing human avatars for some time now.Given how quickly AI has been shackled to the basest human impulses, Gawdat is frustratingly less specific about what enlightened AI would look like. Suffusing neural network training data in examples of human positivity and altruism, as he proposes, seems almost laughably naive. But then maybe it is not so airy-fairy; empathy may need to encompass digital entities that, for practical purposes, will be cognisant and sentient. A top Bhutanese lama agrees with him that the current agenda of “containing” AI and making sure it “serves” humanity contains too many old oppressive tendencies. It’s hard to know how seriously to take someone proposing the same answer as Ghostbusters II – positive vibes to banish the negative ectoplasm. But blockbuster times need blockbuster thinking, and the interviewees here supply plenty.Explore more on these topicsFilmDocumentary filmsAI (artificial intelligence)ComputingGooglereviewsShareReuse this content