Sam Altman said it without hedging on the "Relentless" podcast: humanity has already crossed the threshold of the technological singularity. He framed it as an observable fact, not a distant hypothesis.
The technological singularity is that critical point where an artificial intelligence surpasses human cognitive capacity in general reasoning and triggers a self-improvement cycle that escapes direct human control. That's what Altman claims has already begun. He's not talking about a speculative horizon but a process with dates, products, and quarterly updates.
In his essay "The Gentle Singularity," Altman was explicit: the event isn't going to happen, it's happening. He predicts AI systems capable of generating genuinely novel insights in upcoming cycles, along with robots executing physical tasks with real autonomy shortly after.
The word "gentle" deserves more attention than it's received. It suggests that the species' most radical transformation could arrive without visible rupture, without clear losers. That rhetorical choice reveals as much as any technical prediction.
In "Stones Don't Lie" I explore how power concentration tends to funnel benefits rather than distribute them. That same process is repeating here at a civilizational scale. When one organization controls the development, deployment, and narrative of the most powerful technology ever created, the structure is already unbalanced before any robot completes its first autonomous task.
This connects to what I address in "The Generosity in the Doorway" regarding the job displacement caused by artificial intelligence. Those who build the infrastructure usually end up managing the responses to its consequences too. That doesn't seem like coincidence. It seems like implicit design.
Altman's statement adds something the speculation lacked: temporal urgency. Dates fit into a strategic plan, and that changes the conversation. While some dismiss the impact as a future technical detail, Altman declares it already underway and asks for acceptance of the fait accompli.
What if this concentration isn't a correctable accident but a property of the economic terrain itself? Training frontier models requires capital that only a handful of organizations can raise. The tendency of leaders turning into bottlenecks may not apply when the problem is structural from the start.
Still, the case of Ulsan confirms the historical constant. Hyundai workers stopped the assembly line when faced with the Atlas robot, because the threat had already reached their plant. The weavers of 1811 weren't throwing a tantrum against the machine: they were responding rationally to who accumulated the benefits while they bore the costs. Two hundred years later, the process repeats with unsettling precision.
Ancient stones show similar dynamics in civilizations that believed they were progressing. I'm not a professional ruins-scholar, but those records make one simple thing clear: transformative technologies have rarely distributed their fruits equitably. The pattern persists.
So what about those novel insights Altman expects in the coming cycles? Who verifies that they're genuinely new and not convincing recombinations of existing knowledge, delivered with oracular authority? The Metamate case at Meta already showed systems failing to distinguish basic contexts like professional licenses. If they can't manage that, it seems premature to assume they'll generate scientific breakthroughs without rigorous oversight.
Some aspects remain unclear to me. The sheer speed of the process might make real democratic governance structurally impossible. And the promised gentleness: is it a design goal, or a strategy to reduce resistance while consolidating dominance?
What does seem evident is that when someone with such concentrated power claims a historical event of this magnitude is already underway and will be gentle, it's worth asking gentle for whom, exactly.
Gentle for whom?
Sources
1. "Relentless" podcast — statements by Sam Altman on the technological singularity
2. Sam Altman, "The Gentle Singularity" (personal blog, June 2025)
3. Coverage of Prometheus, Jeff Bezos's automation project
4. Reports on the strike at Hyundai's Ulsan plant against the Atlas robot
5. Class-action lawsuit against Meta over the Metamate system
