My former partner Tim O’Reilly* has written another important essay, and in the lost tradition of blogging about other blogs, I wanted to think out loud about it today.
In “Human Pilled” O’Reilly considers the role of humanity in an economy increasingly dominated by our obsession with building ever-more-powerful AI. Tim is one of the most consistent and original synthesizers of tech’s impact on society, and his essay continues a narrative that stretches back decades, citing not only key cultural texts, but open source software, Web 2 principles, and his own lived experiences.
What have always loved about Tim is his ability to name things I’ve been struggling to evince. During the dot com bust, I had a sense that the Internet would roar back even bigger and better than its initial incarnation. Tim agreed, and he had a moniker that stuck: Web 2.0. As we prepared to launch an event of the same name, his talent for naming important tech trends continued – he called the rise of user generated contributions to the web the “architecture of participation,” for example.
In “Human Pilled,” Tim crystallizes his observations around AI into several memorable phrases, including “runaway magic” – a nod to the Sorcerer’s Apprentice scene in Fantasia. But his most sagacious observation has to do with “missing mechanisms,” which he defines as “the rules and incentives that decide who gets paid, who gets credit, and who gets to take part” in a technological shift such as AI.
In my own work, I’ve been calling these mechanisms “architectures of control,” a nod to Larry Lessig’s seminal insight that “code is law.” O’Reilly also cites Lessig, but Tim’s fulcrum is economics, not law: “We could be spending a lot more time designing mechanisms and affordances that make the economy more productive,” he writes. “We’re spending enormous amounts of money making AI more capable, but comparatively little effort figuring out the mechanisms that will turn those capabilities into opportunities for everyone else.”
That’s exactly the point behind much of my writing, from “Where’s My AI Shareware?”, which asks why AI hasn’t sparked a culture of sharing and openness, as well as “Data Is Non Rivalrous. Why Have We Enclosed It?” – a rumination sparked by economist Sven Beckert’s Capitalism: A Global History – and “Will AI Ever Become a True Platform? Not Without Us…,” which considers architectures of control inside the enterprise. Without incentives that nudge capital, companies, and consumers toward producing what Tim (and Nobel laureate economist Edmund Phelps) calls “human flourishing,” we’ll be living a heightened version of the world we already have – where power, compute, control, and capital are concentrated in the hands of the very few.
Tim argues that current AI models are built on a “colonialist-style appropriation of existing intellectual property. Access to models is sold through expensive APIs, and closed models come with terms of service that forbid others from learning from their outputs. Advertising may be creeping into answers, and memory locks your context inside one provider’s walls. Is that the game we ought to be playing?”
Absolutely not, he concludes. “I think the game we should be playing is human flourishing, and I think the game of human flourishing is creating new affordances for people to create and share with each other.”
If you follow Tim’s work, you’ll see he’s already deeply engaged in identifying and building such affordances and mechanisms, including the open memory protocol, which allows individuals to maintain their personal context between any number of AI agents and models. Back in 2018 I posited a related missing mechanism, which I titled “The Token Act.” The idea is simple at its core: Require all companies who’ve reached a certain scale to build machine-readable data portability into their platforms. Then sit back and watch as innovators, entrepreneurs, and capital find unlimited use cases for that data (here are a couple just to get the juices flowing.)
There’s still time to build a relationship with technology that looks markedly different – and better – than what we’re stuck with now. If you want to engage with that future as it gets built, sign up for Tim’s newsletter here. You won’t regret it!
—
*From 2004 to 2011 Tim and I ran the Web 2 Summit. I also consulted on O’Reilly’s Make Magazine.