Post

There are usually some AI talks, most are trying to use it in interesting/clever ways (e.g. local, narrative, reviewing output, testing). A couple have been more blatant but the audience are very aware of what's what.

This is true, and I'm not convinced by those talk(s), but I suggest giving them a watch as there is still food for thought within. Some examples to follow...

Mmm... it takes her a while to get to her point, but what she's actually talking about is getting rid of testers. Both internal and external. Funny that she says it doesn't replace actual play testing, because that's exactly what studio CEOs want, and if she thinks that's not the case, she's naive.

True, but where do you draw the line? Sure, traditional QA is still essential, but there new approaches that aren't humanly possible, e.g. running 1000 playthroughs overnight looking for anomalies and generating analysis/reports from them in time for breakfast.