man what the f*ck
@manwhatthefukk 
, no, once you plug in really large numbers its wrong, that's what the (x,y) plot shows. There is no end, it goes to infinity, so in essence:
0+1+...+ = (infinity*(infinity+1)/2 = (approximately) infinity*infinity/2 or infinity^2/2. Let's make subsets the of the lower and upper limit of the set to make it easier for you to understand my point.
Let's describe the upper subset (1,000,000,000=T)as {...,T,T+1,T+2}, and the lower subset as {0,1,2,3,4}. For the lower subset there's no problem. 4*(4+1)/2= 10 which is close to 4^2/2= 8. Now for the upper subset (T*(T+1))/2 = 5.000000005e17 (17 0's). T^2/2= 5e17 (17 0's). T^2 = 1e18 (18 0's). In other words, you dividing a T^2/2 = T^2-5e1. IIf it's already that little of a difference at a trillion, at infinity, the difference between n^2/2 and n^2 is truly negligible. So essentially, the bigger the number gets, the smaller the difference between n*(n+1)/2 and n^2. So once you get to really large numbers,
you are saying basically that 0+1+2...+infinity+infinity+1 = infinity^2, which is why the x,y plot never approaches a finite number (e.g. divergent), because that's not true. Think about it man, 10*10 = 10+....+10 ten times. so how many MORE times would you have to add infinity on itself for it to equal infinity squared?

And on top of it, you are decreasing the number you are adding by each time. That would be like you expecting 10+9+8+7+...=10^2, that's crazy talk. As the axiom approaches infinity, the math behind it becomes ludicrous.
You believe it's true at infinity by faith, not by fact. That is Godell's point.