User blog comment:LittlePeng9/Random Turing machines/@comment-5150073-20130406103030/@comment-5529393-20130406111458

I imagine that BB(55,6) > BB(100,2). (perhaps the product is a good rule of thumb, and 55*6 > 100*2). I'm guessing Ikosarakt just meant proving BB(100, 2) > f_e_0 (4), i.e. programming the function in under 100 states.