Okay, a quick summary of the argument (without special characters that evidently do not show up for everyone):
The degree to which a hypothesis is
probabilistically supported by evidence is equal to the probability of the hypothesis given the evidence minus the probability of the hypothesis alone. Let "ps" be probabilistic support:
ps(h|e) = p(h|e) - p(h)
The degree to which a hypothesis is
inductively supported by the evidence is equal to the probability of the material conditional, where the evidence is the antecedent and the hypothesis is the consequent, minus the probability of the same material conditional given the evidence. Let "is" be inductive support:
is(h|e) = p(e -> h|e) - p(e -> h)
Here's the deal: for any evidence which probabilistically supports some hypothesis, that same evidence will also inductively counter-support that same hypothesis. In other words, probabilistic and inductive support for the hypothesis move in opposite directions given the same evidence.
Another way to say this is that as the conclusion of an inductive argument becomes more probable given new evidence, the
logical inference from the premises to the conclusion becomes weaker. That is, the evidence actually supports the conclusion less and less even while its probability keeps rising. Whatever is responsible for the increasing
probabilistic support, it is not anything resembling an inductive or partially deductive inference.
This creates a dilemma for atheists (most are empiricists/Bayesianists): every time they observe something that is not God, the (subjective) probabilistic support for God not existing increases, but, logically, the same evidence
inductively supports the conclusion that God does exist.