NASA’s Galileo mission was primarily designed to explore Jupiter and its moons. In 1989, the Galileo probe started out traveling away from Jupiter in order to do a gravity assist swing around Venus. About a year later it also did a gravity assist maneuver around Earth. Carl Sagan suggested that when passing Earth, the Galileo probe should turn its sensors on Earth to look for signs of life. [1]
Now obviously we know there’s life on Earth. But if we’re going look for life on other planets, it’s reasonable to ask that our methods return positive results when examining the one planet we know for sure does host life. So scientists looked at the data from Galileo as if it were coming from another planet to see what patterns in the data might indicate life.
I’ve started using looking for life on Earth as a metaphor. I’m working on a project right now where I’m looking for a needle in a haystack, or rather another needle in a haystack: I knew that one needle existed before I got started. So I want to make sure that my search procedure at least finds the one positive result I already know exists. I explained to a colleague that we need to make sure we can at least find life on Earth.
This reminds me of simulation work. You make up a scenario and treat it as the state of nature. Then pretend you don’t know that state, and see how successful your method is at discovering that state. It’s sort of a schizophrenic way of thinking, pretending that half of your brain doesn’t know what the other half is doing.
It also reminds me of software testing. The most trivial tests can be surprisingly effective at finding bugs. So you write a few tests to confirm that there’s life on Earth.
Related posts
[1] I found out about Galileo’s Earth reconnaissance listening to the latest episode of the .NET Rocks! podcast.
So did they find life on Earth?
They came up with a list of markers that they thought would be useful to search for. For example, free oxygen. Since oxygen readily combines with other elements, free oxygen suggests something is continually producing it.
“It also reminds me of software testing. The most trivial tests can be surprisingly effective at finding bugs.”
Mutation testing is a neat tool to quantify the goodness of “trivial” tests. I did some work (see links below) where I found that for numeric software trivial tests can be very effective provided that you know the “right” answer to a with sufficient accuracy.
https://qspace.library.queensu.ca/handle/1974/1765
https://doi.org/10.1109/MCSE.2009.157
I don’t remember if it was related but I recall that someone was, for a long time, advocating merely looking for things not in equilibrium (such as your example of oxygen production).
Yes, but proving INTELLIGENT life exists on Earth is a much more difficult problem.
Re: software testing: And make sure the test alive by making sure it fails when the code is broken. It’s very easy to write a test that passes, even when the code is broken. That’s a ‘dead’ test.