It was an idea that defied logic: an online encyclopedia that anyone could edit.

You didn’t need to have a PhD or even use your real name – you just needed an internet connection. Against all odds, it worked. Today, billions of people use Wikipedia every month, and studies show it’s about as accurate as a traditional encyclopedia.

But how? How did Wikipedia not just turn into yet another online cesspool, filled with falsehoods, partisanship and AI slop? Wikipedia founder Jimmy Wales just wrote a book called The Seven Rules of Trust, where he explains how he was able to build that rarest of things: a trustworthy source of information on the internet. In an era when trust in institutions is collapsing, Wales thinks he’s found a blueprint – not just for the web, but for everything else too.

Mentioned:

The Seven Rules of Trust by Jimmy Wales and Dan Gardner

A False Wikipedia ‘Biography’ by John Seigenthaler (USA Today)

Machines Like Us is produced by Mitchell Stuart. Our theme song is by Chris Kelly. Video editing by Emily Graves. Our executive producer is James Milward. Special thanks to Angela Pacienza and the team at The Globe and Mail.

Photo Illustration: The Globe and Mail/Brendan McDermid/Reuters

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Over the last couple of years, massive AI investment has largely kept the stock market afloat. Case in point: the so-called Magnificent 7 – tech companies like NVIDIA, Meta, and Microsoft – now account for more than a third of the S&P 500’s value. (Which means they likely represent a significant share of your investment portfolio or pension fund, too.)

There’s little doubt we’re living through an AI economy. But many economists worry there may be trouble ahead. They see companies like OpenAI – valued at half a trillion dollars while losing billions every month – and fear the AI sector looks a lot like a bubble. Because right now, venture capitalists aren’t investing in sound business plans. They’re betting that one day, one of these companies will build artificial general intelligence.

Gary Marcus is skeptical. He’s a professor emeritus at NYU, a bestselling author, and the founder of two AI companies – one of which was acquired by Uber. For more than two decades, he’s been arguing that large language models (LLMs) – the technology underpinning ChatGPT, Claude, and Gemini – just aren’t that good.

Marcus believes that if we’re going to build artificial general intelligence, we need to ditch LLMs and go back to the drawing board. (He thinks something called “neurosymbolic AI” could be the way forward.)

But if Marcus is right – if AI is a bubble and it’s about to pop – what happens to the economy then?

Mentioned:

The GenAI Divide: State of AI in Business 2025, by Project Nanda (MIT)

MIT study finds AI can already replace 11.7% of U.S. workforce, by MacKenzie Sigalos (CNBC)

The Algebraic Mind, by Gary Marcus

We found what you’re asking ChatGPT about health. A doctor scored its answers, by Geoffrey A. Fowler (The Washington Post)

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

AI art is everywhere now. According to the music streaming platform Deezer, 18 per cent of the songs being uploaded to the site are AI-generated. Some of this stuff is genuinely cool and original – the kind of work that makes you rethink what art is, or what it could become.

But there are also songs that sound like Drake, cartoons that look like The Simpsons, and stories that read like Game of Thrones. In other words, AI-generated work that’s clearly riffing on – or outright mimicking – other people’s art. Art that, in most of the world, is protected by copyright law. Which raises an obvious question: how is any of this legal?

The AI companies claim they’re allowed to train their models on this work without paying for it, thanks to the “fair use” exception in American copyright law. But Ed Newton Rex has a different view: he says it’s theft.

Newton Rex is a classical music composer who spent the better part of a decade building an AI music generator for a company called Stability AI. But when he realized the company – and most of the AI industry – didn’t intend to license the work they were training their models on, he quit. He has been on a mission to get the industry to fairly compensate creators ever since. I invited him on the show to explain why he believes this is theft at an industrial scale – and what it means for the human experience when most of our art isn’t made by humans anymore, but by machines.

Mentioned:

Copyright and Artificial Intelligence: Generative AI Training, by the United States Copyright Office

A.I. Is Coming for Culture, by Josha Rothman (The New Yorker)

Machines Like Us is produced by Mitchell Stuart. Our theme song is by Chris Kelly. Host direction by Athena Karkanis. Video editing by Emily Graves. Our executive producer is James Milward. Special thanks to Angela Pacienza and the team at The Globe and Mail. Media sourced from BBC News.

Support for Machines Like Us is provided by CIFAR and the Max Bell School of Public Policy at McGill University.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.