Condoms reduce the risk of getting infected from _any one performance_. If a girl does, for example, a dozen scenes (I'm not sure of the number as I haven't calculated it) then the odds of infection are about the same with or without a condom...
I barely passed grade 11 math but can tell this is nonsense.
Think of it as the difference between a coin flip and a roll of the dice.
What are the odds of flipping a coin and having it come up heads? 50%, 75% it happens if you flip twice, etc. It *will* happen if you trying long enough, and the odds of it eventually happening go up each time, so we say it converges on 100%.
What are the odds of rolling a pair of dice and getting snake eyes? 8%, 16% if you try twice. Poor odds, but it *will* happen if you keep trying, in other words, the odds converge on 100%.
With HIV it's more like a coin toss without condoms and rolling 12-sided dice (dodecahedron) with a condom. The odds are lower each time, but the cumulative odds over a long series are the same, i.e. 100%.
I did overlook one factor in my earlier post: not every HIV exposure results in infection. So a guy in gay porn probably needs 25 scenes to reach 50/50 odds of getting HIV even if condoms are used.
The odds in straight porn are clearly far lower: the people are tested so we know they're HIV-, and someone who's HIV- cannot infect anyone else even bareback.
It may well be that there is so little work these days that the girls just don't do enough shoots for the odds to converge and that condoms would make a difference. But historically when the girls worked every day for months, depending on condoms to prevent HIV would be useless: you have to do something better than that.