Three possible answers
As far as radio goes, the Milky Way could be in one of three states.
- Alone. No other radio civilization has ever arisen in our Galaxy.
- Not now. Others arose, but none is radio-loud now.
- Unheard. Someone is radio-loud now, and we have not heard them.
A radio civilization is one that gives off radio waves which astronomers at another star could in principle detect, as Earth does today. It is radio-loud for as long as it keeps doing so.
A starting belief is one range for each unknown, such as how often life starts on a suitable planet. The model calls it a worldview. Choose one, and the page shows the chance of each answer before and after the evidence.
Calculating…
The evidence
The model uses observations. are searches that found nothing. is about Earth. A universe counts for less the more it predicted that we would have seen something by now.
Each observation below has two measures. The first is the share of the starting belief’s universes that fit this observation on its own. The second is how much of the change in “Alone” it is responsible for, in percentage points. Each has a switch: turn an observation off and every number on the page is worked out again without it.
Calculating…
It depends on what you believed before
Here is the same evidence applied to starting beliefs. Only of them uses numbers that its authors published. The other are this project’s work: are its own simulation settings, and are its attempt to turn a published position into ranges. The authors of those positions did not write these numbers down, and might not agree with them.
Calculating…
The share of universes that fit is not a score. It depends on ranges for settlement that all these beliefs share, and it includes the early start of life on Earth, which costs every one of them about the same except the Sandberg worldview, whose range for the origin of life is far wider. It cannot say which belief is right.
What decides the answer, and what this cannot tell you
The after-evidence numbers on this page are decided mainly by one observation and two judgments. The clearest case is the worldview of Sandberg, Drexler and Ord, the only one here with published numbers. The chart shows its chance of “Alone” under different settings, worked out now with all observations on and our existence taken as given.
Calculating…
What this cannot tell you
- Which starting belief is right. Every number here is conditional on one. The evidence narrows each belief; it does not choose between them.
- Whether other galaxies are like ours. The observation that does most of the work treats every surveyed galaxy as a copy of the Milky Way with the same history.
- Anything about crowded starting beliefs. A belief that expects many civilizations fits what we see only in a thin corner where almost nobody settles or harvests starlight. Where fewer than the equivalent of universes carry the weight, the page shows no after-evidence numbers.
- How biology works. Everything a belief holds about biology is folded into three probabilities: that life starts, that it reaches intelligence, and that intelligence builds radio. Only the first changes the timeline: where life is unlikely, it also starts late. Every later step keeps the simulation’s timing in every universe, so a belief in which success depends steeply on the time available is not captured.
- What settlement would look like. A settled galaxy is assumed to stay settled and visible for good once the first crossing is over. There is no model of settlers hiding, stalling, or stopping others from arising.
- Much from the early start of life. In this model that observation is worth a factor between and one, and it is nearly the same for every belief except the one with the widest range for the origin of life.
- Fine detail of timing. “Now” is an average over the recent past, and the birth rate of civilizations is falling, so the number radio-loud now comes out slightly high. It also assumes that today’s birth rate held through each civilization’s whole lifetime, which undercounts civilizations that stay loud for billions of years.
- Exact values for very small chances. Every number is worked out from a sample of possible universes. Where too few of them carry an answer, the page prints a limit (“below”) or “not estimable” instead of a number.
- That the inputs are right. Several limits are quoted from this project’s literature review and were not checked again against the papers. The table of constants in the method marks which numbers are published and which are judgments.
- That it has been reviewed. The model is new (November 2026). It was built and checked with AI assistance, and no outside expert has reviewed it.
One more choice: what our own existence tells us
Calculating…
Try your own assumptions
Each unknown has a center and a spread. The spread is counted in factors of ten: a spread of two runs from ten times below the center to ten times above it, with every factor of ten in between equally likely. A spread of zero fixes the value. Probabilities stop at 100%.
The first unknowns matter most. The last two of them are the judgments from the section above: how likely a civilization is to settle its galaxy, and how likely a settled galaxy is to have its starlight harvested. The observations you switched on or off above apply here too.
Calculating…
How many, and how far
The three answers are summaries. Behind them, each universe has an expected number of other civilizations. These charts show how the universes of the chosen starting belief are spread over that number, before and after the evidence. The horizontal scale is in factors of ten, because the unknowns are that uncertain.
What this site’s simulation settings imply for the Milky Way
Calculating…
What would change the odds
Here are findings that have not been made. For each, the model works out the answers again as if the finding were in hand, for the starting belief chosen above.
Calculating…
Method
How the calculation works, every constant and its source, and a check against the published result
Calculating…