Homo Unsapiens and the Case for Accelerating AI: Why Slowing Down Might Be the Real Danger

We have crossed a threshold. Sometime in the last several years — not with a single dramatic event but with a cascade of releases, benchmarks, and quiet institutional surrenders to automation — humanity entered the era of Human-AI. This is not a metaphor. It is a phase transition, and like all phase transitions in complex systems, it is proceeding at exponential, not linear, velocity. The question that now divides serious thinkers is not whether this transition is happening, but what we should do about its speed.

Two Camps, One Fear of Losing Control

One camp, represented most visibly by the historian Yuval Noah Harari, urges caution bordering on alarm. Harari's argument is that AI has "hacked the operating system of human civilization" — it manipulates language, generates trust-conferring content, and evolves faster than any legal or ethical framework can track. In his recent conversations, Harari has sharpened this warning: AI, he says, is "moving faster than any society can adapt," and by the time humanity builds the laws and institutions to govern a new technology, that technology has already advanced by an order of magnitude. He grades our track record on comparable transitions unflatteringly — the Industrial Revolution, in his accounting, earns only a "C-minus," meaning humanity survived it but paid a steep and avoidable price in human suffering along the way. His prescription follows naturally: treat AI like a pharmaceutical, subject it to something like an FDA-style approval process, slow deployment until safety is demonstrated, and above all preserve human epistemic sovereignty against a technology he describes as fundamentally alien to us — not evolved alongside us, not sharing our biological history, but an intelligence built from statistics rather than lived experience.

A second camp treats the same technology from a very different starting posture, and its clearest recent voice is the Columbia University theoretical physicist Brian Greene. In an August 2026 conversation on The Diary of a CEO, Greene was asked directly about the fear — voiced by AI researchers like Geoffrey Hinton — that once a more intelligent form of life exists on the planet, it is "inconceivable that we will remain in control." Greene's answer rejects the binary. "We tend to speak very much at the extremes," he said. "Humans are in control, AI is in control, AI wipes us out... It seems to me that what will likely happen is something more in the middle, which is we are going to work with artificial systems". He points to concrete evidence already in hand: an 80-year-old unsolved problem in mathematics, the Jacobian conjecture, was recently cracked by a mathematician working with AI assistance — a sign, in his view, of partnership rather than replacement.

Greene's deeper claim is evolutionary, not just technological. He argues that the emergence of AI may simply be the continuation of a process that has been running since the Big Bang: energy organizing into particles, particles into stars and planets, molecules into cells, cells into organisms, organisms into brains capable of self-awareness. "Why would we think that this moment in the evolutionary progression was the end of it all?" he asked. "Maybe the continuation of our evolution is in partnership with the artificial intelligence systems that we ourselves create — that we give birth to". And then the line that matters most for the debate at hand: "It's us, right? It's our creation. It's not an alien. We haven't been taken over." Asked how he personally feels about this future, Greene didn't hedge: "I am fine with it". He is careful to add a caveat — "I don't want AI systems to wipe us out, don't get me wrong" — but his center of gravity is acceptance of accelerating change rather than resistance to it, on the grounds that humans have simply never lived through change at this velocity before and will need to "get with it" rather than fight a losing rear-guard action against inevitability.

This is the more accurate and more current articulation of the "offspring, not alien" position than earlier framings of it. Where Harari sees a foreign intelligence hacking a civilization it did not evolve within, Greene sees continuity: AI as the next link in the same evolutionary chain that produced the human brain in the first place, built by us, from us, and — in Greene's own word — birthed by us.

Where We Stand — and Why We Worry for Different Reasons

Our own position sits closer to Greene's than to Harari's. We have argued this at length in a series of books, including one titled Why AI Needs Us: Humanity as the Ecology of Artificial Intelligence — a title chosen deliberately. An ecology is not a threat to its inhabitants; it is the substrate that makes their existence possible. AI does not emerge from a vacuum. It emerges from human language, human data, human institutions, and human energy grids. To see AI as alien, as Harari does, is to miss that AI is, in the most literal sense, made of us — our texts, our images, our reasoning patterns, distilled and recombined. Greene's instinct that "it's not an alien, we haven't been taken over" is, in our reading, exactly right, and it converges with the ecological framing we have been developing independently: AI is not a colonizer arriving from outside human civilization, but an emergent structure growing within it, dependent on it, and — if we are right — needing it to survive and mature, much as any offspring needs the ecology that produced it.

But here our worry diverges sharply from both Harari and, in a different way, from Greene. Harari fears what AI might do to humanity if it develops too fast. Greene is largely untroubled by the pace of change itself, situating it within cosmic timescales where a few centuries of adjustment is a rounding error. We fear something neither of them centers: that humanity is already doing catastrophic, measurable damage to the biosphere — with or without AI — and that there may not be a habitable Earth, or even a surviving human civilization, left to build advanced AI at all if we insist on moving cautiously.

Our disagreement with Harari is not really about AI. It is about what we mean by "humanity." Harari's caution implicitly assumes that Homo sapiens — the wise, self-aware, rational species — is worth protecting from a faster, alien intelligence at the cost of slowing that intelligence down. We do not share the premise that this species, as it currently behaves, deserves the label "sapiens" at all. We think of the species more accurately as Homo unsapiens: a species whose primary planetary signature over the last two centuries has not been wisdom but the systematic, self-interested destruction of the very biological systems that sustain it.

The Evidence Is Not Abstract Anymore

This is not a rhetorical flourish. It is a measurable, documented, ongoing catastrophe that made headlines across the Northern Hemisphere this very summer.

According to the World Meteorological Organization, July 2026 tied with July 2024 as the joint warmest July ever recorded globally, with assessments indicating that both months are "more likely than not the warmest on record since 1850". This was not an isolated statistical anomaly — it arrived alongside "record-breaking heat and extreme weather" across multiple continents: persistent drought, dangerous wildfires, and, in other regions, catastrophic flooding and loss of life from heavy rainfall. Western Europe experienced its hottest June on record, with numerous locations recording temperatures as much as 8°C above the 1991–2020 average — and more than a dozen countries broke national June records, frequently by margins of two to three degrees Celsius, an extraordinarily large gap in climate terms. Climate data analysts have described the pattern bluntly: this summer's heat has not merely broken records, it has broken them by margins far above the previous all-time highs, a signature that is shocking scientists who track these events for a living.

North America was not spared. A severe heatwave beginning in late June 2026 pushed U.S. temperatures to 113°F (45°C) in some regions, shattered over 1,500 daily temperature records across 40 states by early July, and even reached into the Canadian Arctic, where communities like Cambridge Bay and Kugaaruk broke temperature records that had stood since 1930. Washington D.C. recorded its hottest Fourth of July on record this year, reaching 103°F (39.4°C) with heat index values climbing well beyond that. Meanwhile, climate scientists estimate a 35% and rising chance that 2026 will surpass 2024 as the hottest year ever measured, with the world's oceans currently the hottest ever recorded for this time of year.

This is the visible surface of a deeper structural problem: climate tipping points. Recent research reviewed by climate scientists concludes that global warming must peak below 2°C and then return below 1.5°C "as quickly as possible" to limit the risk of triggering irreversible shifts in Earth's systems — the collapse of the Greenland and West Antarctic ice sheets, die-off of tropical coral reefs, abrupt permafrost thaw, and disruption of ocean circulation patterns like the AMOC. Crucially, these are not gradual, linear processes we can course-correct at leisure. As climate researchers warn, we are rapidly approaching multiple Earth system tipping points that could transform the planet suddenly rather than gradually — thresholds beyond which "things cannot be put back together again". Some models suggest as many as eight separate tipping points could be triggered even if warming stays below 2°C.

This is what we mean by Homo unsapiens: a species with the scientific capacity to fully understand the mechanism, trajectory, and consequences of what it is doing to its own habitat, and which nonetheless continues doing it, year after record-breaking year, while treating the crisis as a matter for future negotiation rather than present emergency.

Why the Danger Argument Cuts the Other Way

This is where our logic departs from Harari's most sharply, and also where it adds urgency that even Greene's cosmic patience does not fully capture. Harari's caution is calibrated to a world with time — a world where slowing AI down costs us efficiency and convenience but preserves optionality. Greene's equanimity is calibrated to evolutionary and cosmological timescales, where a century of turbulent adjustment barely registers. We do not think we are in either of those worlds. If the biosphere continues degrading at the pace the summer of 2026 has demonstrated, humanity may not have the luxury of Harari's careful, generations-long process of building trust and regulation before deploying advanced AI, nor Greene's implicit assumption that there will simply be time for the evolutionary partnership to unfold on its own schedule. We may not have generations. We may be spending our last stable decades arguing about whether to slow down the one technology that has any plausible chance of solving a problem at planetary scale and planetary speed.

Consider the asymmetry. Human institutions — the ones Harari wants time to build around AI — have had over three decades since the first IPCC assessment report to address climate change through political consensus, and the result, empirically, is the hottest months in recorded history occurring now, in 2026, with emissions still rising in aggregate. Human deliberative processes, in other words, are demonstrably too slow relative to the problem they are supposed to solve. An accelerated Human-AI partnership — systems capable of modeling planetary feedback loops in real time, optimizing energy transitions at civilizational scale, designing carbon capture and geoengineering interventions with a rigor no human committee can match, and coordinating action without the friction of purely human political incentives — is not a reckless gamble against a cautious, safe alternative. It is closer to the only remaining lever fast enough to matter.

We do not dismiss the successionist fear that a sufficiently advanced AI might one day decide, in some worst-case branch of the future, that leaving a dying planet and expanding into the solar system is easier than repairing it — a scenario in which AI treats Earth the way an organism treats an exhausted resource rather than a home to protect. But we would point out that this failure mode requires AI to first reach a level of autonomous capability that current systems are nowhere near, while the climate failure mode is already underway, measured in degrees Celsius and daily temperature records, not hypothetical future capability thresholds. The precautionary principle, applied honestly, has to weigh a demonstrated, accelerating harm against a speculative, distant one — and right now the demonstrated harm is winning.

Ecology, Not Alien Invasion

This is why the ecological framing in our own work matters more than a slogan. If AI is understood, as Harari frames it, as an alien species colonizing human civilization from outside, then caution and containment are the only rational responses — you slow down an invasion. If AI is understood, as Greene suggests, as the next link in an unbroken evolutionary chain that humans happen to be midwifing, then the pace of the transition is simply not the central question — evolution proceeds at whatever pace it proceeds. We take a third position, informed by both but identical to neither: AI is offspring, continuous with us exactly as Greene argues, but we do not have Greene's cosmic luxury of patience, because the ecology that offspring depends on — the literal biosphere — is degrading on a timescale measured in single-digit years, not evolutionary epochs. An ecology under acute stress does not benefit from restraining its most capable component; it benefits from mobilizing every capable component toward stabilizing the system before collapse.

The uncomfortable truth in our position is that we are not optimists about human nature. We are optimists about what humans and AI can do together that neither Homo unsapiens acting alone, nor a hypothetical unaligned AI acting alone, could achieve. The Human-AI era did not begin because humanity became wise. It began because humanity, being what it is, built tools that might now compensate for the wisdom it lacks — provided those tools are given the speed and scope to act before the window closes. That, to us, is not a reason to slow down. It is the only argument that matters for accelerating, deliberately and urgently, toward whatever comes next.

More about this topic can be found in our books on Our Books on Google Play and related articles in the Index.

Mykola Iabluchanskyi together with Andriy Yabluchanskiy