David Kipping's Cosmological Hart Tipler Conjecture turns Fermi's question into a parameterized equation, and the numbers, on the paper's own assumptions, tilt toward loneliness.
David Kipping, the Columbia astronomer who runs the Cool Worlds Lab, has spent his career on stubborn, large questions. His latest paper, posted to arXiv on June 2 and submitted to Astrobiology, takes a swing at one of the oldest: is anyone else out there? The result, in the author's own framing, is "not encouraging." But the more important thing the paper does is give the field a sharper instrument for asking the question in the first place.
The Cosmological Hart-Tipler Conjecture keeps the spine of an argument first made by Michael Hart and Frank Tipler in the 1970s and 1980s, and swaps the specificity for a more general term: an "artificial infection." A colonization wave, a biological pathogen, a self-replicating AI probe. Anything that emerges in one place and propagates through space is fair game. Kipping then writes a three-parameter equation for the process: a spontaneous spawn rate (lambda), a propagation speed (u), and a start time. Critically, he folds in the Hubble-Lemaitre constant (cosmic expansion) and treats it as friction on the wave. That last move is the substantive update. Earlier formulations treated the galaxy, or even the local group, as a closed box. Kipping's box is cosmological.
The headline numbers come out of that equation. For a wave moving at a tenth of light speed, starting 4.5 billion years after the Big Bang, half the observable universe is "infected" by today if the spawn rate exceeds roughly one per million galaxies. At near-light speed, the threshold drops to one per billion. The stronger claim, which Kipping makes explicit: if even one in a hundred thousand galaxies has ever spawned one of these infections, 99.9 percent of cosmological volume is filled at a tenth of light speed. The paper does not measure any of these things. It asks what they would have to look like, given assumptions, for the silence to be doing the work the Hart-Tipler argument needed it to do.
Universe Today's coverage of the paper opens with the famous Fermi lunch: "Where is everybody?" That anecdote is a fair way in, because the whole point of the framework is that it lets the old question be parameterized. The knobs are visible now. A reader can ask: is the propagation rate the bottleneck, the spawn rate, or the assumption that an infection, once started, runs forever? Each is a different place to look, and each has a different answer.
The biggest objection to the Hart-Tipler argument has been on the table for forty years. Carl Sagan and William Newman, in a 1983 paper called "The Solipsist Approach to Extraterrestrial Intelligence," pointed out that the math assumed a uniformity of motivation across billions of years and trillions of colony-years. Real civilizations, they argued, do not behave that way. A culture that builds one self-replicating probe does not necessarily build a billion, and a culture that builds a billion does not necessarily keep building for a billion years. Kipping is candid about this. The Cosmological Hart-Tipler Conjecture inherits Hart and Tipler's structure, which means it inherits their vulnerabilities. The contribution is not that the new model escapes the Sagan-Newman line. The contribution is that the new model exposes the exact parameter on which the Sagan-Newman line bites.
That matters because the alternative explanations for Fermi's silence (the Rare Earth hypothesis, the Great Filter, the Zoo hypothesis, the Dark Forest, percolation theory) share, in Kipping's reading, the same uniformity-of-motivation flaw. They differ in where they put the bottleneck, but each treats something like "civilizations behave predictably at scale" as a load-bearing assumption. A parameterized Hart-Tipler test lets the field ask, for each of these, what the parameter would have to be, and whether there is any independent way to check it.
Kipping is careful about what the model does not do. It is a preprint, not yet peer-reviewed. The 1-per-million and 1-per-billion thresholds are not measured. They are outputs of a model that assumes the infection, once spawned, propagates as written. If real infections fail to propagate, stall, or get outcompeted by background conditions, the thresholds are wrong in a way the math does not capture. The paper's own framing of its result as "not encouraging" is the author's read of his own model, not a measurement on the sky.
The Fermi lunch question is not answered here. It is, however, slightly better asked. A parameterized model with a known set of knobs is something a research program can attack. A philosophical argument about uniformity of motivation across cosmic time is something a research program can mostly just admire. The Cosmological Hart-Tipler Conjecture, even as a preprint, is the former. What to watch next is whether other groups pick up the equation, change the knobs, and see which assumption, if it gives, is the one that breaks the silence into a softer sound.