Previously a skeptic, James Cameron is now a Stability board member. Plus: fake Keir Starmer and Brian Sewell. We review a weird week in AI.
When society finally collapses in a mire of chatbots, wildfires, rising sea levels and autonomous drones armed with machine guns, it’s a wonder whether the surviving dregs of humanity will look back at 2024 as a tipping point. Certainly, this week (commencing 23rd September) has seen some fascinating developments in AI, the technology that may one day destroy us all.
First came the news that Piranha II: The Spawning filmmaker James Cameron has joined the board of directors at Stability AI, the company behind Stable Diffusion. If you’re unfamiliar, Stable Diffusion takes user prompts and, ripping off (sorry, synthesising) existing images made by real people, creates a new, faintly creepy looking picture.
Stability AI boss Prem Akkaraju said that Cameron’s decision to join the company was a “monumental statement,” and indeed it is, given that the director dismissed the use of AI to generate art a little over a year ago (a “disembodied mind” can’t write movingly about “love, about, lying, about fear, about mortality,” he said last July).
Cameron is also a committed environmentalist – so much so that he made Avatar 2 using solar power, served up veggie food to his cast and crew, and was put on the Time100 Climate list in 2023. It’s currently hard to say how Cameron squares his green credentials with a tech sector that requires not-insignificant amounts of energy to create a single creepy image. Two weeks before Cameron made his announcement, Oracle co-founder, billionaire Larry Ellison, revealed that his company would need to build “three modular nuclear reactors” to power its AI data centres. The power that AI demands is such that even he described the situation as “bizarre”.
Of his business move, Cameron told CNBC, “We sit in a unique nexus between big tech and gen AI on the one hand and the CGI-based workflows associated with movie and television visual effects. That’s a unique trade space to be in, because we all think the next big leap forward is to find a common platform between those two fields.”
Not that Cameron’s the only major Hollywood entity to throw its doors open to AI. Last week came the news that Lionsgate had signed a deal with a company called Runway, which will allow the latter to train its systems on the studio’s library of movies. Lionsgate boss Michael Burns said that the deal would mean it’ll see “capital-efficient content creation opportunities” in the future. In non-business speak, the studio is betting that it’ll eventually be able to make movies or TV shows by typing in a few command prompts, resulting in mass entertainment that won’t require the employ of pesky things like writers, directors or fussy, capricious actors.
If you’re wondering what the immediate future of AI-generated media might look like, you only have to look at another bit of news this week.
The London Standard – the relaunched, weekly edition of the 200 year-old Evening Standard newspaper – came out on Thursday the 26th September. On its cover is an image of an imperious-looking Keir Starmer – but it’s AI-generated, and so it looks about as much like the British Prime Minister as a shop dummy hurriedly dressed in a suit and wig. (The cover’s designed to highlight an interview with Sir Keir about his government’s decision to invest £10bn in a ‘hyperscale’ AI data centre in the North East of England.)
Elsewhere in the paper is a review of Vincent van Gogh art exhibition written by the respected art critic, Brian Sewell. Sewell passed away in 2015, but that means little to the Standard, which has used an AI chatbot to generate a fake write-up in the critic’s famously acerbic style.
Paul Kanereck, the publication’s ‘interim chief executive’ said of the relaunch in a statement:
“The London Standard is a bold and disruptive new publication. The first edition has multiple features on AI and London’s central role in this tech revolution. It includes an experimental AI review by our legendary critic Brian Sewell, and his estate are delighted.”
Elsewhere, Kanereck said that the Sewell column is “a one-off intended to provoke discussion about AI and journalism” – so at least he admits it’s little more than a publicity stunt.
The whole situation might seem blackly comic were it not for the backstory: the Standard was closed down last week and its 150 staff – 70 of them in editorial roles – were let go. That its relaunched publication is partly AI-generated must be a source of delight for the people who lost their jobs.
Key to AI’s seemingly relentless march is the optimism – bordering on euphoria – among its creators and investors. At its annual Connect conference this week, Meta boss Mark Zuckerberg gleefully showed off Orion, a pair of fancy glasses crammed with expensive tech, and announced that its Meta AI chatbot will soon be able to communicate with a host of celebrity voices. Dame Judi Dench, John Cena, Awkwafina and Keegan-Michael Key are among the famous names who’ve had their voices synthesised for Meta’s AI-powered assistant.
If you’ve ever wanted to have your emails read to you by M out of GoldenEye then this could be read as good news. If you’ve ever written your own episodes of 90s sitcom As Time Goes By, you could have robo Dench read them out to you, which is even better news.
Equally upbeat about the future is Sam Altman, CEO of OpenAI. In a post on his blog, titled The Intelligence Age, Altman writes this week that the coming epoch will be one of “massive prosperity” and that ‘superintelligence’ will be with us in “a few thousand days.”
There are certainly useful applications for AI in certain fields. Assuming it works, Project Guacamaya, an effort which uses AI to help save the Amazon rainforest, can only be a Good Thing. But certainly in the world of art and entertainment, there’s the sense that tech companies have decided what’s good for the rest of us and we’ll just have to shut up and accept whatever emerges. And if legislators, racing to keep up with progress, try to come up with some form of pact to help protect ordinary people from some of the darker aspects of AI, tech companies like Apple and Meta are so massive they can just ignore them entirely.
One big player in entertainment, at least, is flying in the face of tech’s ‘join us or die’ attitude. Speaking on behalf of Nintendo, legendary game designer Shigeru Miyamoto told The New York Times this week, “It might seem like we are just going the opposite direction for the sake of going in the opposite direction, but it really is trying to find what makes Nintendo special. There is a lot of talk about AI, for example. When that happens, everyone starts to go in the same direction, but that is where Nintendo would rather go in a different direction.”
It’s a good point: if everyone suddenly decides it’s the height of fashion to churn out automatically-generated slop, it might make sound business sense to offer customers the opposite as an alternative. I know which I’d choose. Of course, Miyamoto could change his tune in a year or so, much like James Cameron has.
The AI hype train is moving at a startling pace, and an increasing number of famous people, it seems, are opting to climb aboard. Whether AI is an NFT-like bubble, the fast-track to a work-free utopia, or a railroad to environmental and societal collapse, we won’t know for sure until a Hunter Killer flies out of its factory and mows down its first human victim.
