10

Probably a proof (if any exist) that calls upon Knuth's up-arrow notation or Busy Beaver.

Ami
  • 938
  • 7
    Once upon a time, this was Graham's number: http://en.wikipedia.org/wiki/Graham's_number. I have no idea what the answer is now. – Qiaochu Yuan Aug 15 '10 at 04:44
  • 2
    This should be community wiki. Perhaps closed – Casebash Aug 15 '10 at 06:34
  • 2
    Infinity $\infty$ is used in lots of proofs :) Anything bigger? – Pratik Deoghare Aug 16 '10 at 13:05
  • 1
    I once saw a programming contest along the following lines: Write a C program, under 5K bytes, that outputs the biggest number possible. Assume (contrary to fact) that C can handle arbitrarily large integers and that your program has unlimited computational resources (i.e. memory). The winning entries were amazing. – academic Aug 10 '12 at 18:51
  • 1
    @Casebash: There is a unique answer to the question, so I do not see why it should be a community wiki. – user1729 Sep 19 '12 at 11:06
  • @FrankThorne: Could I trouble you for a source on that? It sounds very interesting. – SiliconCelery Sep 26 '12 at 19:22
  • See http://math.stackexchange.com/questions/443519/is-there-a-way-of-intuitively-grasping-the-magnitude-of-grahams-number – KeithSmith Jul 21 '13 at 04:16
  • @academic I believe this was Moews's Bignum Bakeoff, even more impressive is that the upper limit was 512 bytes! – C7X Feb 17 '23 at 10:13

3 Answers3

16

The mathematician Harvey Friedman observed a special finite form of Kurskal's Tree Theorem. Regarding this form, Friedman discusses the existence of a rapidly growing function he calls $TREE(n)$.

The $TREE$ sequence begins $TREE(1)=1$ and $TREE(2)=3$, but $TREE(3)$ is a number so extremely large that its weak lower bound is $(A(...A(1)...))$, where the number of A's is $A(187196)$, and $A()$ is a version of Ackermann's function: $A(x) = 2↑↑...↑x$ with $x-1 ↑s$ (Knuth up-arrows).

Whereas Graham's Number is $A^{64}(4)$, the above mentioned lower bound is $A^{A(187196)}$. As you can imagine, the $TREE$ function keeps on growing rather quickly. For a discussion on the hierarchy of fast growing functions see here: http://en.wikipedia.org/wiki/Fast-growing_hierarchy

There are other examples of numbers greater than Graham's Number, as can be seen here: http://en.wikipedia.org/wiki/Goodstein_function#Sequence_length_as_a_function_of_the_starting_value, although I'm not sure if this number is larger than Friedman's $TREE(3)$

Elchanan Solomon
  • 30,807
  • 1
    In terms of fast growing hierarchy:$$\operatorname{TREE}(n)\gg f_{\theta(\Omega^\omega \omega,0)}(n)$$ http://math.stackexchange.com/a/1979838/272831 – Simply Beautiful Art Mar 07 '17 at 14:41
8

In one of Friedman's posts on the FOM mailing list, he mentions a number called SCG(13) that is far larger than TREE(3): http://www.cs.nyu.edu/pipermail/fom/2006-April/010362.html

I couldn't find a lot of other information about it, though.

Danny
  • 101
  • 1
  • 4
4

TREE[3] is much larger than the numbers derived from Goodstein sequences for any reasonable input. See: http://www.cs.nyu.edu/pipermail/fom/2006-March/010279.html

The Goodstein function is upper bounded by ε₀, whereas the TREE function is lower bounded by the small Veblen ordinal.

Danny
  • 101
  • 1
  • 4