I have been sitting with Calum Chace’s Artificial Intelligence and the Two Singularities for a while now, and it is one of those books that keeps surfacing in my thinking whenever someone brings up AI and immediately jumps to Skynet. Chace’s argument is more interesting, and useful.
He proposes that there isn’t only one singularity — but two. A technological singularity, where machines surpass human intelligence and the whole trajectory of the species bends around that fact. And an economic singularity, where AI’s deployment across every industry produces mass unemployment long before anything resembling super-intelligence shows up. Chace’s point, and the one I keep coming back to, is that the economic version is the more likely one — and the one few are prepared for.
Narrow AI vs. General AI
Chace draws a line between artificial narrow intelligence (ANI) — the type of AI that’s already prevalent, performing specific tasks — and artificial general intelligence (AGI), which would involve something closer to consciousness and subjective experience. We are not near AGI yet. What we regularly incorporate is ANI, which reorganizes how we work.
That distinction reframes the potential anxiety. People picture Terminator. What they should be picturing is in fact, call centres, e-commerce, and manufacturing automating due to faster speed, cheaper labour, and more consistent performance — no sentience required.
The “iceberg”
One of the ideas from the book that’s stuck with me is what Chace calls the “iceberg” phenomenon. A junior lawyer appears to be on thin ice — his entry-level job is exactly the kind of task AI can now complete. But he is actually standing on an iceberg where a mass of newly available work still stands beneath him (at least for now). While automation may be progressively replacing these jobs, too the transition is slower and stranger than the panic may suggest, and the displacement takes place from the bottom up.
What’s left for us
Chace lists lower-income service work, retail, manual labour — and, more provocatively, lawyers and journalists — as sectors already being reshaped. Not eliminated. Reshaped.
Chace offers seven factors that keep humans working even as AI penetrate the labour market:
- Partnership with AI rather than replacement by it
- Lower wages, as a trade-off for staying employed
- Work that requires empathy
- Goods and services rebranded as “artisanal” — made by human hands
- Entrepreneurship
- The arts
- Categories of work that don’t exist yet
Last one is honest because it is an admission that few knows what might be coming, including Chace. He quotes Kevin Kelly of Wired — “machines are for answers; humans are for questions” — as a kind of consolation. AI still doesn’t have common sense. It can’t ask the questions that streamline the problem. That belongs to us, at least for now.
He also draws a line between artistry and creativity. AI can produce “art.” But the human artist expresses something of personal significance, something that derived from lived experience. Whether that distinction holds up — whether audiences will care about the difference, might be a separate question. But Chace’s claim is that it does matter, and that it will keep mattering economically, even as the technical gap between AI output and that of human closes.
2035, and the Economy Nobody’s Designed Yet
Chace marks 2035 as a transitional year, when AI plays a much bigger role across transportation, manufacturing, retail, construction, health care, finance, education, and government — and unemployment climbs as a result. His response to that isn’t doom rather, a call to design “a new type of economy which we have not yet designed.” He suggests a few shapes it could take: a “Star Trek” economy of energy abundance, a version of socialism built around universal income, or collective ownership running on blockchain infrastructure that requires no governments. What his warning concerns is sequencing — if technological unemployment “arrives in a rush,” we get dystopia. But if we see it coming and plan for it, we are ahead.
Where my view diverges
Here’s where I depart from his argument slightly. I appreciate the optimism, and I think his diagnosis is largely right — the economic singularity is more plausible and more pressing than the technological one. I don’t think that’s controversial among workers in the field. But I have a harder time imagining the leap from “capitalism as it currently functions” to any of the alternatives he sketches. That’s not just a shift in policy. It’s a shift in mindset, and it requires governmental will that doesn’t currently exist anywhere I can point to.
Chace seems to think mass unemployment might be the forcing function — that people won’t embrace a shared or redesigned economy until they’re pushed into it. And we actually got a real-world test of that idea during the pandemic. Mass disruption to work, real-time experimentation with income support, entire sectors reorganizing overnight. Did any of that shift the underlying capitalist mindset? I am not sure. The disruption happened. The reimagining didn’t follow. So I’m left wondering whether Chace’s forcing-function theory actually holds, or whether we’re capable of absorbing enormous economic shocks without ever questioning the system that produced them.
Source: Chace, C. (2018). Artificial Intelligence and the two singularities. CRC Press, an imprint of the Taylor & Francis Group
