1. Define an infohazard as an argument that is true, but shouldn't be said because the effect of circulating that argument on a hypothetical future world (or present discursive world) is dangerous.

2. Assume, as I do, that to argue something is an infohazard is damaging to our ability to seek truth because the word "infohazard" introduces a new and tedious way to shut down an argument agnostically of its truth value.

(For example: "I agree that the idea that queer people are Born This Way is damaging, but you shouldn't argue against it because our opponents might capitalize on the fact that we were wrong / a queer person down the line might need to hear it to transition / we're in too much danger right now to support infighting over immaterial concerns / &cet.")

3. Therefore, if the "Roko's Basilisk" argument that infohazards exist is sound, the effect of circulating it in our world is a net negative.

4. Therefore, the argument I am making now holds that the argument for infohazards is an infohazard.

5. Therefore the argument I am making now is "a tedious way to shut down an argument agnostically of its truth value" and should not be trusted! ⇒⇐

I was reminded that John Williams composed two pieces for the Olympic Games. The first, completed for the 1984 games in L.A., uses Leo Arnaud’s Bugler’s Dream fanfare as an introductory section. Arnaud’s piece was not composed for the Olympics but had been associated with it since 1964, so it was familiar to fans of the games.
https://youtu.be/QoodIsTZyNg

The second, which I believe was composed for NBC when it picked up coverage rights starting with the 1988 games in Seoul, jettisons Bugler’s Dream in favor of an original fanfare. (I’m guessing this was purely for rights reasons — when Williams records his ’84 fanfare he frequently drops the Bugler’s Dream for a perfunctory introduction.) It also reworks the austere middle section to be much more melodic and hummable. This is the one that’s the real infohazard. It’s also the one that I have hugely nostalgic memories of hearing on the live broadcasts in ’88.
https://youtu.be/i4QQ2SvQcE8

Musically, it’s fascinating to see what Williams does with his second bite at the apple. I imagine that the spec for both pieces was basically the same: about 4 minutes long, opening with a fanfare, somewhat marchable so it can be played at the opening and closing ceremonies, and nothing off-putting. Of course this is the kind of thing Williams can deliver in his sleep.

But both of them have a moment in the middle which is something of a musical breakdown. In the 1984 piece, the music loses its way a bit at 2:15, wanders around, rallies, starts to fragment again at 2:40, and is rescued by the fanfare at 2:55. In the 1988 piece the passage is shorter, rejecting the triumphal march beat and falling apart at 1:55 before being put back on track by a fanfare leading into the march at 2:07.

Who asked for these passages? They’re the kind of thing I think you couldn’t get away with in a piece that was meant as a strict processional march (c.f., e.g., Walton’s Crown Imperial, which has tempo variation for drama but never just descends into milling about, or any of Elgar’s Pomp and Circumstance marches, which have moments of chaos and even some rallantando but never a passage where the orchestra seems to lose the plot). So my headcanon is that they’re Williams’s personal statement about how it goes when you’re on your own in the middle of a grand project. You don’t always get to march straight through.