Click here. Scroll up for the conversation, and the photographic evidence.
The short version: Brad Anderson is the Greatest Living American Hero. I posted a nonsense cocktail recipe as a joke on Twitter, and he actually followed the instructions to the letter. He drank the whole thing and livetweeted the photo evidence.
The recipe (and why it's relevant for this blog) for the "Instant Regret":
In a collins glass, mix equal parts:
heavy whipping cream
garnish with hash browns
The 140 character limit cut off the remainder of the instructions, which were to serve on a bed of chilled mayonnaise and chase with olive-soaked havarti.
And here's why Mr. Anderson is the Greatest Living American Hero: the "beverage" is called the Instant Regret, and he quaffed it regardless. Right there on the label, it's clearly not euvoluntary. Did he care? No. He knew the hazards and he defied them anyway, thumbing his nose at an overcautious nanny culture. His suffering became for one bright, brief, shining moment the immense pleasure of anyone there to witness his heroics.
The lessons are twofold. First, it is better to regret something you have done than to regret something you haven't done. And second, pernicious regret arises under conditions of information asymmetry, not from information incompleteness. Brad knew what he was doing. And so I doff my hat. Well done, sir. Well done.
Also, I invite you all to think of the general equilibrium of ubiquitous warning labels. There's really no good way using text alone to signal the severity of risk or the nature of the conditional probability involved (not in a way that makes much sense to a semi-numerate public anyway).
Warning labels suffer a pooling equilibrium problem: t/f? Explain.