AI haters gonna hate hate hate
Here's what Cate Blanchett, A24 and a £10m grant have to do with it
Reality TV watchers are the loudest when they hate someone. Hate drives the chats. People often say they hate a particular show on Bravo (Boring! Fire everyone!), but it is still the most-watched show in the whole franchise. It's a strange thing.
This week, we've seen a lot of chatter on AI backlash, coming especially after the announcement of the A24 x DeepMind colab. I read deals and press releases so you can form an informed opinion beyond the keyboard warriors.
Today's briefing includes:
New AI deals you should know about, within tech and the creative industries
What we SHOULD be angry about when it comes to AI
AI in film festivals, and why that Sam Altman film was cancelled by Amazon.
Understanding the “AI race" and the current so-called backlash
A £10m funding pot in case you want to be the change
How AI affects your security and steps to safeguard yourself against hackers
THE CASE FOR THE BACKLASH (?)
Around 40% of voters want AI banned from most industries, protests have scuppered nearly $100bn of US data-centre projects, and warring AI megadonors just poured tens of millions into a Manhattan congressional race. The Economist’s headline reads as the backlash is only getting started, but if you read it through, you'll see they're not actually in favour of the backlash and halting AI. And they're very concerned about the US losing to China in the AI race. More on this later.
CATE BLANCHETT LAUNCHED A TOOL TO PROTECT YOUR IDENTITY FROM AI, A registry where you allow OR prohibit AI use of your name, face, voice, likeness and movement. “Your identity is your IP in the age of AI.” Which is clearly needed, if it works, since nearly 1000 actors signed an open letter against AI voice cloning of CHILD ACTORS, and they alleged that a "major studio has responded with an attitude of take it or leave it”
Talent manager, take note: if your client uploads anything directly to a platform, assume the terms of service are licensing it to train a model.
AI STOCKS TUMBLE AS THE MONEY GETS NERVOUS. Tech stocks fell as corporate customers rethink their AI spending; Apple is raising prices on Macs and iPads as AI demand drives up chip costs; and OpenAI is leaning toward delaying its IPO to next year. Money, not industry feeling, seems to be the only thing that could properly stop or rejig the AI world. Who's surprised?
THE LOUDEST ONLINE HATE: indie-film darling A24 teamed up with Google DeepMind, and the comments became a bloodbath: “Do you know your fanbase?. I'm not saying it's good or bad, but it is, as far as the press release, a research partnership. Google gets no access to A24’s library or data, and there’s no mandate for any filmmaker to use anything. A24’s defence: “We’d rather have a seat at the table than on the sidelines… we want to dictate what tools get built for artists rather than having tools handed to them.” Worth remembering, Disney licensed characters to OpenAI, Lionsgate is in bed with Runway, and Netflix bought Ben Affleck’s AI startup.
MEET THE REAL ANTI AI WARRIORS
Music critic Ted Gioia thinks AI supporters are shrinking toward “less than ten per cent of the populace — roughly the size of a crazy cult,” and that “you should expect the tech backlash to escalate further.” He says this ends unhappily for Silicon Valley. There are some absolutely baller moves in this piece on how people are ” poisoning" AI, with artists deliberately lacing their tracks with data designed to corrupt any model that scrapes them.
That is not surprising, considering Google argues YouTube’s terms of service hand it a broad license to train AI on anything uploaded straight to the platform, which means the indie artists who used YouTube to “level the playing field” are getting scraped, while labels and publishers often negotiate their own deals. A lawsuit claims Google’s Lyria 3 model was trained on YouTube songs with no payment to the artists.
on the other hand..
MOISES: THE BRAZILIAN AI THE MUSIC INDUSTRY WELCOMED. music.ai Born in Brazil in 2019, now $40m raised and 70 million users seem to be welcomed by the industry. Why might that be? It is trained only on licensed content, “tools, not replacements” as a stated value, ethical partners only.
BIG TECH (AND BIG MONEY) DOES NOT CARE
Amazon quietly dumped Luca Guadagnino’s nearly-finished film about Sam Altman’s 2023 firing. Some may wonder how much Amazon, a tech power and major player in the current A.I. arms race, was interested in poking a bear like Altman, who wields enormous power and has many key allies. Earlier this year, Amazon and OpenAI announced a massive strategic partnership. The deal included a $50 billion investment in OpenAI by Amazon. You may remember that we shared Ronan Farrow's profile on Sam Altman before.
AI IS EVERYWHERE
AN AI FILM PREMIERED AT TRIBECA, and a sculptor’s film became the first AI-assisted animation at Cannes. Márcia Mayer, ex-Pixar, thinks AI is a tool to cross thresholds you couldn’t before. They still had a crew of 45 people on the project. She quotes Björk: “You can’t blame the computer. If there’s no soul in the music, it’s because nobody put it there.”
FRESH FROM CANNES LIONS, a big question was on everyone's mind: how do you influence an AI model? Once Google serves an AI summary above the search results, a Fortune 500 brand can go commercially invisible overnight. So OpenAI pitched ads inside ChatGPT, Adobe paid $1.9bn for Semrush to make brands “visible” to models, and Profound hit a $1bn valuation just for tracking how ChatGPT, Claude and Gemini describe your brand.
LINKEDIN + ADOBE LAUNCH A GLOBAL AI-SKILLS INITIATIVE: Marketing job postings demanding AI literacy are up 113% year over year.
CANVA GROW 2.0 builds, launches and optimises your ads in one place.
TIKTOK’S SYMPHONY AGENT will shortlist creator videos about your brand for you.
WIMBLEDON launched an AI fan experience for the Championships.
🌍 ABOUT THAT AI RACE I MENTIONED
That first article on The Economist says that if America chokes its own AI, it could “cede the global frontier… to authoritarian China". So, less about AI and more about power.
The US ordered Anthropic to block non-Americans from its top models, Mythos and Fable. Within the fortnight, Tokyo’s Sakana launched “Fugu” advertising “frontier capability without the risk of export controls,” China’s 360 unveiled a rival, and the US allowed a partial release of the model it had just restricted.
A journalist at Wired, however, met with China's AI experts and guess what? THEY'RE ALSO FREAKING OUT! The sessions were nervous, circling things like recursive self-improvement (models rewriting their own code to get better, indefinitely) and humanoid robots. Chinese open-weight models are popular with US researchers, and both sides are stuck on the same unanswered question: how to keep these models open enough to push science forward without handing hackers a cyber-weapon?
Meanwhile,in London, the AI scene traces back to Demis Hassabis keeping DeepMind in the city after selling it to Google in 2014, but “the most advanced AI is built by a handful of American companies, on American soil, under American law, and what the rest of us are permitted to do with it can change on a Friday afternoon.” The one homegrown long shot (Cosine) is racing to train a “sovereign” model on $15m.
There’s still ONE thing missing from this conversation, in my humble opinion: and it's about WHO owns AI. Mariana Mazzucato argues that government funding was essential to develop the INTERNET and other key technologies: Where would Google be today without the state-funded investments in the internet, and without the US National Science Foundation (NSF) grant that funded the discovery of its own algorithm?
The question, then, is not whether A.I. will change the world. It will. The question is: Who will own and control that future?
Artificial intelligence was not created out of thin air. The data and language used by generative A.I. tools didn’t just pop into Sam Altman’s head. A.I. is built on our collective intelligence: our books, songs, artwork, journalism, computer code, scientific research, videos, conversations, images and ideas spanning generations. - source
WHAT WE SHOULD BE ANGRIER ABOUT
AI IS GETTING WOMEN WRONG. Of 133 AI systems studied, 44% showed gender bias; ask a model to finish a sentence starting with someone’s gender and about 1 in 5 responses come back sexist. Of 138 countries, only 24 even mention gender in their AI strategy.
ASSUME YOU WILL BE HACKED. AI has collapsed the time to exploit a known software bug from over 700 days in 2020 to about 44 in 2025. The realistic posture now is “assume you’ll be compromised, and limit the damage.” Here's a quick to-do list if you're freaking out
Use a password manager, and a unique password everywhere.
Two-factor on your email first — app or hardware key, not SMS.
Treat every email and DM as untrusted.
Update your devices the day the update drops.
Keep less cash in your day-to-day account, and hand less data to every form and free trial — the weak link is usually some third party you forgot you signed up for.
MAYBE YOU CAN BUILD THE CHANGE?
Innovate UK’s Next Wave: Breakthrough Wave 1 is putting up to £10m behind early-stage “createch” projects sitting at the intersection of the creative industries and frontier tech. UK micro-businesses and SMEs, with total project costs between £100k and £500k. Deadline: 11 August
WORK OUT WHAT YOUR AI HABIT COSTS THE PLANET: A free calculator that estimates the energy, carbon and water footprint of a single AI query.
Would love to hear your thoughts.
As usual, thanks for reading, liking and sharing