Hallucinations, non-deterministic results, and continents-worth of energy consumption—these problems have no real solution. We’re stuck with them. How can we factor these uncertainties into the productivity gains we get from the tools? What if a 30% productivity gain is as good as we get? I will cover the promises, the realities, and the current/future states of Generative AI as it applies to testing software. We will acknowledge the unmovable realities getting in the way of the unstoppable fantasies and help you separate the benefits from the grift.
