I’m Conflicted About Artifact SETI

I’m going to frame this blog post into “pros” and “cons” of artifact SETI, with a lot of references to three specific papers we read about searching for extremely advanced ETIs(Annis (1999)Carrigan 2009, and Wright et al. 2014b).


  • No dealing with Schelling points!
  • No making assumptions about xenopsychology and alien motivations!
  • No need to even consider a METI approach!
  • This is actual, performable, observational astronomy.
  • Huge budgets aren’t needed. Carrigan (2009) used IRAS sources which had already been catalogued.
  • Actual constraints can be placed. Annis (1999) made a (perhaps arbitrary) statement that no galaxy out of his ~130 galaxy survey had a Type III civilization harvesting 75% or more of the available stellar energy.
  • Interesting objects can (and will) be found during the search for outliers, which is far less likely in communication SETI. Carrigan (2009) found 16 objects which were (weak) Dyson sphere “candidates”, which mostly indicated an unusual IR signature.

The Con:

  • We can only look for extremely advanced civilizations, because we need the artifacts themselves to be detectable by our technology at interstellar and even intergalactic distances. The three papers we read for this week were all focused on Kardashev Type II or Type III civilizations. The sheer enormity of lengths of time involved versus (assumed) civilizational lifetimes means that we’re way more likely to come into contact with a civilization more advanced than we are than one at about the same age. But the idea of searching for them makes me a little squeamish.

Any astronomer can tell you that the distance scales of space are not really built for/intuitive to humans. I personally find that there’s a sort of intellectual harmony in the thought that interstellar travel is just hard and not worth it (not quite in a Hart sense of “physical explanations”). Though the swan songs of “free” energy, FTL travel, and the use of dark energy are tempting, and the excuse of poorly understood current physics convenient (as discussed in Wright et al. 2014b), that doesn’t mean they’re necessarily rational fallbacks. Much of our ideas about how galactic settlement would work are based more on the ideas of settlement from Earth’s history, and though it feels tempting to just zoom out and apply the same metaphors, there are fundamental differences that to me* feel insurmountable.

All of that to say: I don’t find it very intellectually fulfilling to do artifact SETI for Type II and Type III civilizations, Type III civilizations especially. It’s too much speculation. So while Annis makes conclusions about the frequency of Type III civilizations, I don’t find his results particularly compelling or surprising.

*at this point in my education, perhaps I’ll look back on this and laugh