Graham's number is VASTLY less than TREE(3), so you lose, actually. Graham's number is around A(64)(4), i.e., the result of iterating the number of Knuth arrows 64 times starting with 4 Knuth arrows, while A(A(187196))(1) is an extremely weak lower bound on...

The maximum length of a sequence of labeled trees, with 3 labels, such that the Nth tree has at most N nodes and no tree can be embedded in any later tree... this being the definition of TREE(3).

See http://en.wikipedia.org/wiki/Kruskals_tree_theorem

Meh.

Take however many symbols it takes you to describe your large number. Call the symbol count N. Then my counterpoint is the Nth busy beaver number. That's a big number.

You might as well say 'One plus the biggest whole number nameable with 1,000 characters of English text', if that sort of thing is allowed.

Are you alluding to Scott Aaronson's essay on Who Can Name the Bigger Number? [1]

[1] http://www.scottaaronson.com/writings/bignumbers.html

Yep!! The point is, if you say 'Take however many symbols it takes you to describe your large number', that's not well-defined and is fundamentally ambiguous due to the nature of language.

Of course the whole large number pissing contest is extremely ageist.

This all just means that countable instances are bad ways to talk about the uni(lol)verse.