“Do you not think we're actually, maybe, the crash test dummies?” That was the stonker Chris Stokel-Walker recently skewered an academic with at a conference.
She had tried to defend the use of ‘synthetic data’. Now she was stumped. To her mind, taking imperfect humans out of the equation was equivalent to shoving a mannequin into a soon to be obliterated car. It was ethical and, frankly, more efficient.
But the academic had failed to take into account the hundreds of millions of people who have already tested the faulty and unfinished products which are large-language models (LLMs). “Maybe we're currently in the middle of the car crash?” Stokel-Walker warned.
This interaction, amongst supposedly sensible and highly-intelligent people, demonstrates how the rise of generative AI has subverted people’s view of technology. Caution is cast aside because AI can change the world. And seemingly minor blips like ‘hallucinations’ will be rectified in the future. Get stuck in now or miss out, is the mantra.
Rightfully, Stoke-Walker, author of the forthcoming How AI Ate The World, is worried about “ordinary people”, the people who don’t read newsletters like this and don’t know that Geoff Hinton, often described as the godfather of generative AI, actually thinks those ‘hallucinations’ are a feature, rather than a bug, of LLMs.
They also don’t know that Meta’s chief AI guru Yann LeCun has proposed an entirely different approach to OpenAI to reach artificial general intelligence — a concept nobody can really explain. And they definitely don’t know that the Italian data regulator has raised privacy concerns about ChatGPT.
Stokel-Walker isn’t a doomer. And he’s not necessarily anti-AI. He just wants to bring the debate back down to the ground, take the hype out and think about the societal and economic impacts of generative AI. It’s first order principles.
Much like my own warning about the culture of perpetual-Beta seeping out of the video game sector into the wider world, he’s worried about how generative AI is already being readily embraced, without many questions being asked.
“The average person thinks this is kind of an intelligence. And it knows what it's talking about when it's not, it's just a pattern matching machine. And not always a good one at that,” Stokel-Walker told me.
And is the rise of generative AI, something which exploded into the zeitgeist from end of 2022 and the start of 2023, really a ‘revolution’?
“I think it is a little bit of Column A and a little bit of Column B: I do think that there is kind of like a fissure and a break from what's gone before,” Stokel-Walker explained. “And this is something that is quite significantly different. And I think that's in part because more people know about it.”
At the vanguard of this revolution is Sam Altman, the former president of start-up accelerator Y-Combinator turned CEO of OpenAI who successfully held off a coup by posting love heart emojis on Twitter. The Great Man Theory of history would put Altman centre-stage. But how important is Altman to this story for Stokel-Walker?
“When you think about gen AI for most people you think of ChatGPT,” he said. “It’s very history to fit into Grand Men Theories when you talk about tech, and that’s a big issue because women are often overlooked, but it did shock me that when he was deposed by the then OpenAI Board, that every single employee basically rolled behind him.”
The author even produced a documentary on the failed ousting for the BBC. ”If OpenAI is nothing without Sam Altman, then the current generative AI revolution is nothing without OpenAI. So that means his position is really powerful,” Stokel-Walker argued.
His new book, published by Canbury Press, is effectively split into two halves: the first section is conserned with the history of generative AI, mapping out the biggest technological developments since the 1950s, while the second half focuses on the issues.
The worries include the potential environmental impact of the race for more compute (something which I’ve written extensively about), AI-powered disinformation and the technology’s current and potential impact on work. We are already seeing publishers advertise for ‘AI assisted’ reporting roles in the media industry.
At each stage of the book we’re given a case study of how generative AI is impacting the world, some positive and some negative. The main achievement of the title is that Stokel-Walker makes it crystal clear that generative AI is out there right now and people are experimenting with it, modding it and changing the world around them.
“AI is eating the world. Arguably, it already has. But if it’s not to chew us all up – and spit us out, broken, battered and beaten – then we need to act,” he concludes.
🤔 Other tech and media news I’ve found interesting
Anglofuturism is here, folks. UK-based self-driving vehicle start-up Wayve has raised a cool $1 billion or £837 million in a Series C round led by Nvidia. The company was spun-out of Cambridge University and founded by Alex Kendall in 2017.
TikTok fights back. The ByteDance-owned platform has sued the US Government after the business was ordered to divest from the American-arm of the company amid privacy and national security concerns (and the backdrop of the US election, of course). You could see TikTok’s reaction as inevitable: if it yields in the US, it could set off a chain reaction, one which would probably hit Europe next.
Iger shakes it up. After fending off Nelson Peltz’s activist challenge, Disney CEO Bob Iger has announced some major changes at Disney. Amongst other initiatives, the streaming arm of the business will be following in Netflix’s footsteps by cracking down on password sharing. His team have also ordered Marvel to slow-down on its output, limiting the action hero franchise to no more than three movies per year and two shows.
BuzzFeed clings on. In a bid to stay listed on NASDAQ, BuzzFeed has enacted a stock split. The short story of this interesting equity management move is that there are now fewer Class A shares, inflating the company's share price in the process. It has to be above $1 to remain on NASDAQ. It went up to $1.70, then dropped to $1.69.
The Times ups its audio. The UK-title has launched a new show, The Royals, hosted by Roya Nikkhah and Kate Mansey. The News Corp-owned outlet will be seeking to expand its portfolio of podcasts across the year. At last count, the combined Times titles had more than 575,000 digital subscribers.
Latest Political Press Box. Talking of podcasts, the latest episodes of The Political Press Box have dropped, including long-form interviews with pollster and fellow Substacker Rob Ford, The Independent’s newest hire David Maddox and Tortoise Political Editor Cat Neilan.
TMTG latest. There’s always something going on at Trump Media it seems. The NASDAQ-listed business has now fired its auditors. TMTG shares are trading around $47 per stock.
Some hope for local news. The latest Pew research shows that US adults do value local news media. But only 15% of Americans said that they have paid for such content.
🎥 Video essays
📖 Essays
How disinformation is forcing a paradigm shift in media theory
Welcome to the age of electronic cottages and information elites
Operation Southside: Inside the UK media’s plan to reconcile with Labour
📧 Contact
For high-praise, tips or gripes, please contact the editor at iansilvera@gmail.com or via @ianjsilvera. Follow on LinkedIn here.
186 can be found here
185 can be found here
184 can be found here
183 can be found here
182 can be found here