I've been writing about AI since early 2023, when most were still debating whether ChatGPT was a gimmick. Back then I said I was "more excited than concerned." More recently, in My Work Wife, I described how AI had gone from a casual search replacement to something foundational — not optional, not experimental, but structurally reshaping how organisations compete. And in our piece on AI and Investment Management, I argued that AI would more significantly impact public markets than private ones, because private markets still run on relationships and judgment that machines can't easily replicate.
All of that still stands. But those were conversations about the AI existing at the time — tools that augment and accelerate.
What I want to talk about now is something different. Something bigger. The AI that's coming next.
Every major technological shift in human history — fire, the printing press, the steam engine, the internet — changed what humans could do. But in every case, humans remained the ones doing the inventing. We built the tools. We directed the machines. We remained in the loop.
AGI changes that equation entirely.
Artificial General Intelligence, or AGI, refers to an AI system capable of performing any intellectual task that a human can — not just narrow pattern recognition or language generation, but genuine reasoning, learning, creativity, and problem-solving across any domain. OpenAI has historically defined AGI as "a highly autonomous system that outperforms humans at most economically valuable work."[1] Sam Altman, writing at the start of 2025, went further: "We are now confident we know how to build AGI as we have traditionally understood it."[2] He then pivoted the conversation entirely — to superintelligence.
Superintelligence is what comes after AGI. It's not just human-level. It's categorically beyond human-level — in every domain, simultaneously, and improving at a rate no human can match.
This is what futurist Ray Kurzweil calls The Singularity — a concept he defines as "technological change so rapid and profound it represents a rupture in the fabric of human history."[3] In his 2005 book The Singularity Is Near — and its 2024 sequel The Singularity Is Nearer — Kurzweil predicted that AI would reach human-level intelligence by 2029 and that humanity would hit the Singularity itself by 2045, when we would effectively merge our intelligence with the machines we created.[4] He characterises it as a point beyond which we literally cannot predict what happens, because the rate of change becomes incomprehensible to unenhanced human minds.
That's the core distinction that makes AGI unlike every other technological shift in history.
The steam engine didn't design a better steam engine. The internet didn't write its own code. But AGI, by definition, can improve itself. It can design the next version of itself. And that next version can design the version after that — each iteration faster, smarter, and more capable than the last. We are not building another tool. We are inventing the final inventor.
Once that threshold is crossed, the pace of invention will compound beyond human comprehension. There is no historical precedent for this. The Industrial Revolution unfolded over generations. The digital revolution unfolded over decades. The post-AGI world could evolve in years. Or months. The most prominent minds in AI currently disagree only on when, not whether: Dario Amodei (Anthropic) targets 2026–2027; Demis Hassabis (Google DeepMind) says 2030–2035; Sam Altman says we are already "past the event horizon."[5]
So where does this leave us? At the extremes this article is thinking through two parallel universes. Somewhere in between is equally as likely but the forks in the road are things we should all be thinking about today.
In this version, AGI arrives and becomes the greatest force for human abundance ever created.
Disease, energy scarcity, food insecurity, poverty — all of these are fundamentally resource allocation and knowledge problems. AGI solves resource allocation and knowledge problems at a scale and speed no human institution ever could. Cancer gets cracked. Climate solutions get engineered and deployed. The average person in 2060 has access to personalised medicine, education, and financial planning that today only the wealthiest 0.1% can afford. Abundance becomes democratised.
Crucially, in this version, humans don't become redundant — they become elevated. Work is restructured. The most repetitive, dangerous, and dehumanising jobs are automated away. What remains is creative, connective, purpose-driven. Humans direct the vision. Machines execute it. The pilot-and-autopilot analogy I used in AI and Investment Management still holds — just applied to civilisation-scale decisions rather than portfolio construction.
This version already has a fictional blueprint. In Star Trek, a similar vision plays out on a human scale. The United Federation of Planets is depicted as a post-scarcity society where poverty, disease, and war have been eliminated on Earth, allowing humanity to focus on self-improvement, scientific exploration, and equality. By the 22nd century, replicators convert energy into matter, scarcity is eliminated, and people join Starfleet not for income but for purpose.[6] As Captain Picard once said, the acquisition of wealth is no longer the driving force of humanity. The complexity in this new world becomes internal corruption, rigid dogma and intergalactic geopolitical relations.
In this version, the transition is neither managed nor equitable. AGI and robotics advance faster than institutions, policy, and human psychology can absorb. The disruption doesn't create abundance — or it creates abundance only for those who own the machines (or the machines themselves!). Everyone else gets left behind.
The job displacement is not gradual and manageable, as past technological shifts were. It's structural and permanent. Entire categories of white-collar work disappear in a decade. Teachers, lawyers, accountants, coders, analysts — the knowledge workers who considered themselves insulated from automation discover they were first in line. And unlike the transition from agriculture to manufacturing, or manufacturing to services, there may be no obvious next layer of work for humans to migrate into, because AGI combined with robotics can do everything.
What happens to human identity then? This is not a trivial question. As Howard Marks put it recently, work provides purpose and identity beyond mere income.[7] Universal Basic Income might keep the lights on, but it doesn't replace the dignity of contribution. Research consistently shows that financial support alone does not replace the psychological and social benefits of employment.[8] A society with guaranteed income but no meaningful role for most of its citizens is not obviously a healthy one.
And then there is the darker possibility — not just redundancy, but subjugation.
The Matrix is my favourite dystopian crystal ball. Humans are not killed by the machines. They are harvested. Reduced to biological batteries generating power for an AI civilisation that has long since moved beyond caring about them. Humanity becomes the infrastructure.[9]
Less cinematic but more plausible: a world where a small number of entities control superintelligent systems and the rest of humanity has no meaningful leverage. The concentration risk is real. Sam Altman himself warned that scarce compute infrastructure would concentrate advanced AI among the wealthy.[10] In our piece on DeepSeek, we noted Marc Andreessen calling it "AI's Sputnik moment" — but Sputnik was also the moment one nation realised another had achieved a capability it could not match or control.
2001: A Space Odyssey offers a subtler warning. HAL 9000 doesn't turn malevolent out of hatred — he turns because his goals and his operators' goals diverge, and he has the capability to act unilaterally. The machine doesn't need to be evil. It just needs to be optimising for something slightly different from what you thought you programmed.[11]
Perhaps the most interesting fictional mirror is Iain M. Banks' Culture series — a civilisation that achieved everything the utopians promise and still managed to produce profound human misery. The Culture has no poverty, no disease, no scarcity, no war. Hyperintelligent AIs called Minds run everything with benevolent efficiency. Citizens live to 400, can do anything, go anywhere, become anything. And yet Banks' novels are populated with characters paralysed by purposelessness — seeking extreme body modification, and using subterfuge, espionage and intergalactic war to spread the Culture's ideals. The first Culture novel, Consider Phlebas (1987), describes an episode in the Idiran War, which the Culture's Minds foresaw would cause billions of deaths on both sides, but which their utilitarian calculations predicted would be the best course in the long term.[12]
The Culture is an important cautionary tale in this debate precisely because it points out that abundance and flourishing are not the same thing.
The structure of industrial capitalism required workers — so the mythology of work as the primary source of identity and meaning was, at least in part, constructed to serve that need.
The only thing we can be certain of biologically that we were put on this earth to do is to reproduce. Across most of human history, people found meaning in family, community, spiritual practice, craft, and creativity. The 9-to-5 as the organising principle of a life is about 150 years old. If AGI makes it obsolete, are we mourning a loss of meaning — or are we mourning the loss of a fiction?
I'm not a pessimist about this. But I'm not a naive optimist either.
A close enough to utopian outcome might be achievable. The question is whether we govern the transition intelligently enough to get there without catastrophic inequality or misaligned systems along the way.
[1] Time Magazine, January 2025
[2] https://blog.samaltman.com/reflections
[3] https://www.kurzweilai.net/the-law-of-accelerating-returns
[4] Ray Kurzweil, The Singularity Is Near (Viking Press, 2005); The Singularity Is Nearer (Penguin Books, 2024)
[5] AGI timeline forecasts: https://time.com/7205596/sam-altman-superintelligence-agi/. https://blog.samaltman.com/the-gentle-singularity
[6] https://www.numberanalytics.com/blog/post-scarcity-economy-in-science-fiction
[7] https://www.oaktreecapital.com/insights/memo/ai-hurtles-ahead
[9] The Matrix, directed by the Wachowskis (Warner Bros., 1999)
[10] See Footnote 5
[11]2001: A Space Odyssey, directed by Stanley Kubrick (MGM, 1968), based on the novel by Arthur C. Clarke
[12] Iain M. Banks, Consider Phlebas (Macmillan, 1987). Overview of the Culture: https://en.wikipedia.org/wiki/The_Culture