Post spooks

Post spooks

Other urls found in this thread:

schmid-werren.ch/hanspeter/publications/2014elemath.pdf
en.wikipedia.org/wiki/Borwein_integral
twitter.com/NSFWRedditVideo

...

...

...

That strikes me more as a floating point roundoff error than anything else

well you're wrong

Did you evaluate that nasty integral symbolically to check? Prove it.

schmid-werren.ch/hanspeter/publications/2014elemath.pdf

Everyday I go to sleep I feel as if the world is only getting worse
en.wikipedia.org/wiki/Borwein_integral

nope its correct. just a product of infinity being weird

>what is mathematical induction taught in first semester of every single STEM programme

>saying mathematical induction but not even knowing what it is.
Nice post.

If you trust the "first few terms of a sequence" and extrapolate them towards infinity, you clearly dropped out in your second week you fucking teenager

want to know how i know you don't know shit about math?

>what is mathematical induction
ok, try and apply induction to that problem, ill wait.

it has nothing to do with "infinity being weird"

...

>Mathematician spook
For all consistent formalized system [math]F[/math] which contains Peano arithmetic, [math]F \,\not\vdash\, \mathrm{Cons} \,F[/math].

>Physicist spook
For all isolated system [math]\Sigma[/math], [math]\mathrm dS \,>\, 0\;\mathrm{J \,\cdot\, K^{-1}}[/math].

>Computer scientist spook
[eqn]\vdash\, P \,\vee\, \neg P[/eqn]

>Imageboard spook
[eqn]\rm {\color{red}A^{\displaystyle \color{yellow} u}}_{\displaystyle \color{green} t}{\color{cyan} i}^{\displaystyle \color{blue} s} \color{indigo} m[/eqn]

Oh shit

>still insulting people while not knowing what you're talking about.
dude stop. Just stop. Kys if you cant even be bothered to fucking google what you:re saying.

The integral of sinx from 0 to 2pi is 0.

The derivative of sinx is cosx.

Always found stuff like that too perfect and thus spooky.

>integral of sin x from 0 to pi is 2

that one spooked me out in high school

Finding the slope as delta x approached 0 and summing an infinite series of terms into a finite number spooked the shit out of me when I learned Calculus. I understood how it worked, and I did well and din't have any problems doing or understanding it, but to this day it spooks the shit out of me that its possible.

...

>mfw after taking an """applied""" complex analysis course (what engineers take) this just doesn't feel that special anymore

...

A




u








t







i





s





m

>forgetting the original logician spook
[eqn]\left\{ x \middle| x \notin x \right}[/eqn]

The usual "proof" of the identity is presented as a syntactic trick that involves shuffling around the definitions of i, sin, cos and exp together with the oh-so-rigorous ... notation.
Which is really a shame because it downplays the geometric interpretation of the identity, from which the entire field of complex analysis pretty much follows.

>If you trust the "first few terms of a sequence" and extrapolate them towards infinity,
is this really what you think mathematical induction is?