No, the Copernican Principle does not apply to time

A common error in the service of a worse argument.

Hate to pick a fight over this kind of take, but Freddie de Boer’s latest is pretty aggravating:

I would call for a Temporal Copernican Principle, an admonition that commentators on modern issues, especially AI, operate from the general assumption that we are not occupying a particularly important time period within human history. (To be clear, the ordinary Copernican Principle also has a temporal element, but that’s invoked in cosmological terms while I’m interested in human terms and a human timeline.) We should always operate from a stance of extreme skepticism that we live in a particularly important human moment…

First off, the Copernican Principle does not apply to time. You can go to any place in the universe and it will look roughly¹ the same, but if you move forward or backward in time things look radically different. People used to think that the Copernican Principle might apply to time back in the days of steady state theory, but now we know that if you are looking for a privileged vantage point in the universe all you need to do is travel back to the Big Bang. Measuring time elapsed since the Big Bang has an absolute meaning in a way that (say) measuring space based on distance from the earth does not.

I clarify this from the outset because Freddie is not the first person to try to apply the Copernican Principle to time. Astrophysicist James Gott still peddles his clever “Copernicus method” for predicting the future using a lot of the same assumptions as Freddie — insisting, for example, that our temporal perspective is always probabilistically random. I am not going to wade into all of the (often pretty technical) mathematical and philosophical debates about this, but what I do want to flag is that whatever this is, it has nothing to do with Copernicus. Call it the Temporal Mediocrity Principle if you like, but calling it the Copernican principle just confuses people into thinking, as Freddie puts it, that “the ordinary Copernican Principle also has a temporal element”. No! That hasn’t been plausible since the early twentieth century.

Another reason I think this is a bad argument is that it invites us to treat very different claims about the importance of this era as equally implausible. That’s the whole point: it’s an intellectual shortcut, a way of skipping past the specific merits of arguments.

Do we live in a special age where artificial intelligence poses a historically unprecedented threat to the human race? No, that is stupid.

Do we live in a special age where climate change poses a historically unprecedented threat to the human race? No — the threat is entirely precedented. When you look closely at those 250k years of human history that Freddie breezes past, what you find is that entire civilizations have been wiped out because of climate change, and that even the human race writ large has been in serious jeopardy once or twice. There would be nothing “special” about that happening once again, nor would it be particularly unusual in the grand scheme of things for humans to be the latest species to go extinct because of climate.

There are good reasons to conclude that one of these arguments is much stronger than the other, and that instead of throwing money into artificial intelligence R&D we should probably be investing in things like carbon capture and solar energy. But if we take Freddie’s argument seriously there’s no reason not to just eyeroll at both of these possibilities while pretending that your blind cynicism has some basis in logic or Copernicus or whatever.

1. Though the Copernican principle has served scientists well, it’s become increasingly clear that it is not strictly correct even concerning space. The universe, it turns out, is not homogenous — at the largest scales it seems to have a distinct structure that essentially corresponds to where galaxies appear, and that structure is thought to have emerged from quantum fluctuations during the Big Bang. One can also, from quantum physics, extrapolate a whole philosophical framework which supposes that the perspective of every observer is privileged in a sense, but those ideas are a bit too whimsical and speculative for my taste.