This post mentions an equation from Ender's shadow

Which kind of annoys me because either:

  1. It's trivial to solve using algebra
  2. The joke is that the answer is irrational so it takes "forever" to solve

So the question I have is:

What is the best problem you know of having the following properties:

  1. Can be expressed as a short formula using "high school" math or less
  2. Has an integer (or rational I guess) answer
  3. ..which is less than 10**100
  4. And cannot be solved using contemporary mathematics

The first "almost" example I came up with was something along the lines of

Where  are all prime.  This should have a couple of large prime factors and hence be hard to factor, but notice it doesn't satisfy criterion 3 (since we can easily factor numbers ~10**100**2)

Does anyone have a better example?

My other instinct was Ramsey Numbers but these seem to fail criterion 1 since I didn't learn the definition in any high school math class.

Also thought of Collatz Conjecture which I guess would count if it had a counter-example < 10**100.

New Answer
New Comment

8 Answers sorted by

Measure

4910

Find integers a, b, c, such that a³ + b³ + c³ = 42

The solution for this problem was found recently (2019):

42 = (-80,538,738,812,075,974)³ + 80,435,758,145,817,515³ + 12,602,123,297,335,631³

Can you please add a citation?

EDIT: The original paper (linked from what localdeity found)

6localdeity
I googled one of the numbers and found e.g. this and Wiki.

tailcalled

81

Not sure if that is counter to the spirit of the post, because while the answer is either -1 or 0, it involves an intermediate calculation which vastly exceeds .

Feynman once challenged people to come up with a problem that could be stated quickly but he couldn't solve to within 10% in a minute, and a colleague stumped him with finding .

If you mean something like cos(3^^^3 * pi), then I think that one should be solveable. Maybe try induction.

EDIT: Whoops, I didn't see the floor symbols. Nevermind me. With not-quite-so-large numbers it's an interesting problem that might be solveable by finding digits of pi in linear time, but linear time is of limited help here.

-2Beckeck
↑ is not ^  https://en.wikipedia.org/wiki/Knuth%27s_up-arrow_notation 
2Charlie Steiner
I was too lazy to find the right character. Whoops, looks like I misread the comment though.

andrew sauer

46

Your rules need refining about how large "intermediate values" are:

⌊π^π^π^π^π⌋ mod 10

-High school formula

-Integer result < 10 < 10^100

-Can't be solved w/ contemporary maths

-Misses the spirit of the challenge but obeys the rules

Viliam

40

What is the smallest positive integer n such that both p and p+n are prime for infinitely many values of p?

(The answer is probably 2, but so far we can only prove that it is not greater than 246.)

Source.

Charlie Steiner

40

Maybe anything about Hofstadter's Q-sequence.

In the Q-sequence,  is defined as  (with terms n=1 and n=2 set to 1). So , and so on. This sequence "dies" if  for any , because then the next term would ask for some term like  which is undefined. We don't know whether the Q-sequence dies or not.

Caspar Oesterheld

30

There's a Math Stack Exchange question: "Conjectures that have been disproved with extremely large counterexamples?" Maybe some of the examples in the answers over there would count? For example, there's Euler's sum of powers conjecture, which only has large counterexamples (for high k), found via ~brute force search.

queelius

10

I'm kind of annoyed by the challenge, also.

However, this does provide the opportunity to think a bit more deeply about identity and being able to refer to a thing, like π.

To expand on this, the number n, like the number π, has no decimal representation, but unlike most real numbers, it is well-defined since it can be pointed to and thus named: the equation as given in the post is one such name (it is an implicit representation of n; simple algebra may be used to rewrite it into a more explicit form).

If n became useful enough, we could even give it a simpler name (like we did with π and 2), instead of the compound symbolic expression.

There are a lot of unspecified assumptions in Ender's challenge, such as reducing the expression to some canonical form, e.g., 2 instead of 16/8, sqrt(4), or the only even prime number. It takes a bit of sophistication to appreciate the depth of questions of equality, identity, and representation.

The fact that Ender had not considered any of this while confidently presenting the challenge leaves me feeling a bit annoyed.

1 comment, sorted by Click to highlight new comments since:

Ugh, that equation annoyed me because it looked like GPT-1 trying to write math. I wonder if Ramanujan would have dreamt his conjectures if regularly exposed to this.