What’s the most L’esprit de l’escalier[1] advice you’ve ever gotten?
I want to hear advice that at the time you either misunderstood or left you utterly baffled but some time later, you had a “Eureka” moment. A rushing in of insight long afterwards where you finally were able to apply the advice and understand it.
My own examples aren’t that good but hopefully are illustrative:
In GLSL it is good practice not to use conditional statements, or to use them for very small branches.
I am not a coder by any stretch of the imagination so at first this seemed like an arbitrary rule that made no sense. It was only when I saw the output of a GLSL to ARB compiler that I understood why this is good practice. The Eureka Moment was seeing two virtually identical slabs of code, of course they aren’t identical in that the values they are processing are very different depending on the conditional.
From then on I began to take the order of operations and conditionals much more seriously and understood why conditionals should be avoided.
Of course, for the tiny Shadertoy fragment shaders I am writing with very few loop iterations, if any, the overhead is negligible.
Another example that comes to mind is the heuristic from the famously contrarian Nassim Nicholas Taleb:
Go for the surgeon who looks like a butcher over the surgeon who looks like surgeon
The insight was completely lost on me when I first read it. Now I understand that it is the notion of “in spite of appearances” or costly signalling. If a surgeon doesn’t look like a clean cut surgeon and is still in the business, he must be damn good at the actual business of a surgeon, and probably better than someone who has all the right appearances.
An analogy I might make is to dating, never take dating advice from someone conventionally attractive. There is a line from a Jim Belushi sitcom lampshading the typical—slobby husband, photogenic wife dyad which I’ll paraphrase:
“Do you want to take advice from someone who married down (i.e. his TV wife), or someone who married up (Jim Belushi’s character)?”
Another analogy—Don’t take career advice from someone born into money. Take it from someone who had none of network or advantages.
So, two examples of “what the heck could that mean” that only later made sense and appeared to be good advice.
This might seem to invite broadbrush “deep” sentiments like “oh they told me to cherish these days, but it wasn’t until years afterwards I realized what I was missing” but I’d be much more interested in technical or nitty gritty Eureka moments like not using conditionals in GPU code.
That moment well after the party when you think of the perfect retort or reaction. Like the “Jerkstore” episode of Seinfeld. There is also a Dot Hacker song called “wit of the staircase”
If a surgeon doesn’t look like a clean cut surgeon and is still in the business, he must be damn good at the actual business of a surgeon, and probably better than someone who has all the right appearances.
This assumes that both surgeons are evaluated by the same set of criteria (so if one scores lower on something, he probably scores higher on something else). That is not necessarily true.
As a counter-example, imagine hospital A that only hires professional surgeons and pays them high salaries, and hospital B that hires anyone and pays them low salaries. The “surgeon who looks like a butcher” probably works in the hospital B.
What’s the most L’esprit de l’escalier[1] advice you’ve ever gotten?
I want to hear advice that at the time you either misunderstood or left you utterly baffled but some time later, you had a “Eureka” moment. A rushing in of insight long afterwards where you finally were able to apply the advice and understand it.
My own examples aren’t that good but hopefully are illustrative:
In GLSL it is good practice not to use conditional statements, or to use them for very small branches.
I am not a coder by any stretch of the imagination so at first this seemed like an arbitrary rule that made no sense. It was only when I saw the output of a GLSL to ARB compiler that I understood why this is good practice. The Eureka Moment was seeing two virtually identical slabs of code, of course they aren’t identical in that the values they are processing are very different depending on the conditional.
From then on I began to take the order of operations and conditionals much more seriously and understood why conditionals should be avoided.
Of course, for the tiny Shadertoy fragment shaders I am writing with very few loop iterations, if any, the overhead is negligible.
Another example that comes to mind is the heuristic from the famously contrarian Nassim Nicholas Taleb:
The insight was completely lost on me when I first read it. Now I understand that it is the notion of “in spite of appearances” or costly signalling. If a surgeon doesn’t look like a clean cut surgeon and is still in the business, he must be damn good at the actual business of a surgeon, and probably better than someone who has all the right appearances.
An analogy I might make is to dating, never take dating advice from someone conventionally attractive. There is a line from a Jim Belushi sitcom lampshading the typical—slobby husband, photogenic wife dyad which I’ll paraphrase:
Another analogy—Don’t take career advice from someone born into money. Take it from someone who had none of network or advantages.
So, two examples of “what the heck could that mean” that only later made sense and appeared to be good advice.
This might seem to invite broadbrush “deep” sentiments like “oh they told me to cherish these days, but it wasn’t until years afterwards I realized what I was missing” but I’d be much more interested in technical or nitty gritty Eureka moments like not using conditionals in GPU code.
That moment well after the party when you think of the perfect retort or reaction. Like the “Jerkstore” episode of Seinfeld. There is also a Dot Hacker song called “wit of the staircase”
This assumes that both surgeons are evaluated by the same set of criteria (so if one scores lower on something, he probably scores higher on something else). That is not necessarily true.
As a counter-example, imagine hospital A that only hires professional surgeons and pays them high salaries, and hospital B that hires anyone and pays them low salaries. The “surgeon who looks like a butcher” probably works in the hospital B.