OK, where to start? Well, we are arguing about pure logical outcomes. Here are our wannabe professor's options (and for ease of reference, I shall refer to the four grid boxes as A:False, A:True, B:False and B:True. And I shall refer to him as Ron):
Now, there is an immediate problem with his logic that should be obvious. Spotted it? Because apparently none of Ron's friends can.
Yes, it is in A:True. Why? It is simple.
Whilst Ron assumes that A:False will lead to "global depression", he assumes that A:True does not; he has inserted a smiley face instead.
However, by Ron's logic, if A:False leads to global depression, A:True must also do so. By Ron's definition, it may be true that global depression is better than the alternative, as represented in B:True, but it does not mean that it doesn't happen.
This rather changes the grid and, simultaneously, utterly undermines Ron's logical argument; what this means is that Column A is now a lose:lose situation, whether GCC is happening or not.
Thus, by Ron's logical extrapolation, the best thing to do is to choose Column B and do nothing because, in that way, there is a 50:50 chance that we will be alright. If, however, we choose Column A, there is a 100% chance that we will definitely lose out.
I could, of course, also argue about the extremity of his positions; the likelihood of a 20 foot water rise—yes, he quotes Gore and, yes, the likelihood of that happening in the next 100 years (let alone the next few decades) is 0—or increased Katrina-style hurricanes, etc.
However, all that Ron asked is that we debunk the logical argument which I have done: quod erat demonstrandum.
As Ron says, spread the word.