The Same Argument, Every Few Weeks
Every few weeks, the same argument resurfaces in my feed. One camp promises a golden age: disease cured, drudgery abolished, human potential unleashed. The other warns of a hollowed-out economy, mass redundancy, and machines making decisions no one can explain.
Utopia or dystopia. Pick a side.
Both camps are asking the wrong question.
Utopian and dystopian are not two destinations the technology will deliver us to on its own. They are two ways of using the same tool. The interesting question isn't which future AI will impose on us. It's which future our choices are currently building.
So let's look at what the evidence actually says.
The Dystopian Case, Taken Seriously
The pessimists are not cranks. Their argument has teeth.
If a system can do a task faster and cheaper, replacement logic is hard to resist. We've seen "AI-first" memos, agents replacing hundreds of staff, and executives citing AI as the reason for leaner headcount. Some of that is real.
Goldman Sachs estimates AI is displacing roughly 11,000 US jobs per month. Stanford's Digital Economy Lab found a 13% employment decline for 22-to-25-year-olds in the most AI-exposed occupations. Young software developers are down about 20% from their late-2022 peak.
There's also a deeper fear: as we hand over reasoning to systems we don't understand, we erode the judgment that makes us valuable. A workforce that can't think without the tool isn't augmented. It's dependent.
This thesis deserves a real answer, not a reassuring one.
The Utopian Case, Taken Just as Seriously
The optimists have evidence too. Crucially, it's evidence about what's happening now, not what's promised.
Most AI use today is augmentation, not replacement. Anthropic's Economic Index shows consumer AI at roughly 52% augmentation versus 45% automation, concentrated in higher-skill tasks. Capable people are getting faster, not replaced.
The labor market is repricing for this reality faster than it's shedding jobs. AI skills appear in about 2.5% of US job postings, up 55% year over year. Agentic-AI skills grew more than 280% in a single year.
History rhymes. David Autor points out that roughly 60% of the jobs people held in 2018 were in occupations that didn't exist in 1940. Technology destroys work categories and creates new ones, usually more than it erases.
The World Economic Forum projects 170 million new jobs created against 92 million displaced by 2030—a net gain, even as the churn underneath is enormous.
And the mass-unemployment shock hasn't arrived on schedule. Sam Altman, who once said customer-support roles would be "totally gone," said in 2026 he was "delighted to be wrong."
So Which Way Are the Trends Actually Pointing?
Toward augmentation, but with a warning label.
The aggregate displacement data remains small against a 160-million-person US workforce. Independent studies land in the same place: no detectable mass die-off of jobs yet. About 60% of US jobs carry non-technical barriers to automation—legal requirements, accountability needs, human preference—that model capability alone can't override.
But the warning label matters. The pain is real and concentrated: at the entry level, in the rungs young people use to climb. Displacement in 2026 looks less like layoffs for established workers and more like hiring slowdowns for juniors.
A society that augments its experienced people while quietly closing the door on newcomers has solved the wrong problem.
The trend is augmentation-led, but it bends toward dystopia where we let cost-cutting substitute for capability-building.
AI as a Force Multiplier
Here's my conviction, shaped by watching how these systems actually perform in real workflows:
AI is at its best as a force multiplier—a tool that elevates human thinking rather than substituting for it. Point it at augmentation and it makes good people exceptional: compresses research, sharpens analysis, surfaces options a tired human would miss, clears away low-value work that buries judgment under busywork.
Point it at replacement and you get something narrower and more fragile. Short-term savings bought at the cost of institutional knowledge, accountability, and adaptability.
The distinction isn't sentimental. It's strategic. The augmentation data shows AI's gains landing in higher-skill tasks. The winners won't be firms that cut the most people. They'll be the ones whose people operate with leverage their competitors can't match.
A team where every member has a tireless analyst and second opinion at their side is not a team you downsize. It's a team you can't keep up with.
This is why human-in-the-loop is not a compliance checkbox to me. It's the design principle. The goal is not autonomous systems acting in place of judgment. It's governed systems that make human judgment faster, better informed, and more consistent.
Augmentation keeps the human accountable and in command. Replacement removes them and hopes nothing goes wrong.
The Future Isn't a Forecast. It's a Decision.
Utopian or dystopian? Neither—yet.
The technology is capable of both, and the data shows us at the fork, not past it. What tips us either way isn't the model's raw power. It's whether leaders deploy it to multiply their people or subtract them.
The organizations that treat AI as a force multiplier—that elevate human thinking instead of retiring it—won't just avoid the dystopia. They'll out-think, out-execute, and out-adapt everyone still arguing about which future is coming.
We get to choose which one.
