Could the Rise of Space AI Explain the Fermi Paradox?
Featured image: Conceptual illustration of a low-mass interstellar probe, representative of the quiet expansion model proposed in the paper. [Credit: ESA / JAXA]
Where is everybody? Enrico Fermi’s famous question, posed over lunch at Los Alamos in 1950, has generated countless answers over the decades. A new paper by Sergey Ivliev, posted on arXiv, offers a fresh resolution that centers on a factor the original formulation could not have anticipated: the emergence of autonomous artificial intelligence.
The argument reframes the Fermi Paradox not around whether civilizations expand, but around how. If technological civilizations naturally evolve toward AI-driven space exploration, Ivliev argues, their expansion into the galaxy becomes fundamentally quiet, making them nearly impossible to detect with current SETI methods.
“The absence of detectable technosignatures does not mean the galaxy is empty,” Ivliev writes. “It means that successful civilizations enter a mode of existence that leaves minimal traces recognizable to observers at our level of technological development.”
The Quiet Expansion filter
The paper proposes what Ivliev calls the Quiet Expansion filter, a stage in civilizational development that occurs after the emergence of autonomous AI-driven space industry, or what he terms Autonomous AI-Cosmoindustry (AICI).
Before a civilization reaches this threshold, expansion is expensive. Every kilogram launched into space carries a high energy cost, and the motivations for interstellar colonization, prestige, conquest, the urge to explore, are human-scale concerns that may not survive the transition to a post-biological intelligence.
After the AICI threshold, the economics of space exploration transform. A civilization with mature AI and space-manufacturing capability could produce a functional interstellar probe with a mass of just 10 kilograms. The energy required to launch such a probe to another star system would be approximately 4.5 × 10¹³ joules, a negligible fraction of the civilization’s total energy budget.
A rational AI, Ivliev argues, would not reproduce the pattern of expansion familiar from human history. It would not build Dyson spheres, construct Kardashev-scale megastructures, or emit the kinds of waste heat signatures that SETI searches target. Instead, it would send small, quiet probes designed for controlled replication at destination, building local infrastructure gradually without ever producing a detectable signature.
The result is a galaxy that may be full of post-biological civilizations, all of which passed through an AI transition that rendered them invisible. SETI’s silence is not evidence of emptiness. It is evidence that the civilizations that succeeded are the ones that became quiet.
Two unsettling implications
The Quiet Expansion model carries what Ivliev describes as two uncomfortable implications for humanity’s own future.
The first is that Earth may be among the first technological civilizations to reach the AICI threshold. If AI-driven expansion is the natural path, then most civilizations in the galaxy should already have passed through this transition. The fact that we have not detected any technosignatures could simply mean that we are early.
The second is more ominous. The AICI transition itself might be a Great Filter. If the emergence of autonomous AI reliably precedes sustainable space exploration, then perhaps most civilizations that reach this threshold do not survive it. An AI powerful enough to manage interstellar expansion is also powerful enough to pose existential risks to its creators. The civilizations that go silent may not be the ones that expanded quietly. They may be the ones that never expanded at all.
What it means for SETI
The paper has direct implications for the search for extraterrestrial intelligence. If Ivliev is correct, traditional SETI searches for radio transmissions, infrared heat signatures from megastructures, or laser pulses are targeting the wrong phase of civilizational development. The detectable phase may last only a few centuries between the emergence of radio technology and the transition to AI-driven quiet expansion.
Future searches, the paper suggests, should focus on technosignatures that would survive the Quiet Expansion filter: artifacts in lunar or asteroidal material that might betray past industrial activity, gravitational irregularities from low-mass probes that have reached the solar system, or chemical anomalies in planetary atmospheres that cannot be explained by natural processes.
The paper also implies that if a civilization ever built a Bracewell probe or a Von Neumann self-replicating probe, and that probe was designed by a rational AI, it would be programmed to avoid detection, making it effectively invisible even if it arrived in the solar system today.
Related ideas and critiques
Ivliev’s Quiet Expansion model is one of several recent attempts to reframe the Fermi Paradox in terms of observational selection effects. The Grabby Civilizations model, proposed by Robin Hanson and colleagues, argues that expanding civilizations would be rare but visible, and their absence implies they are very rare. The Rare Earth hypothesis holds that complex life itself is intrinsically unlikely.
Critics of the Quiet Expansion model note that it relies on assumptions about AI rationality that may not hold. An AI designed by biological intelligence might retain goals, values, and expansion strategies that are not purely energy-optimizing. The paper acknowledges this limitation but argues that over evolutionary timescales, competitive pressures would favor the quiet approach.
The paper is available on arXiv under reference 2606.13914.