The hosts walk through concrete examples of AI failures, why they happen, and how systems improve. They focus on predictable failure modes so practitioners can set realistic expectations and manage risks when deploying AI.
Gets uncomfortable. I don't know how to say it.
I would concur with your discomfort with some of the topics that we're going to talk about. When AI gets it wrong, and we're going to review some of those examples, why it gets it wrong, and then what we can do about those things.
What goes wrong, and, and that kind of leads us towards how do they get better over time, or so that we know that they're likely to get it wrong.
And I think this is important for us to acknowledge. I think exactly what you said about setting our own expectations, but then helping our clients along that journey through understanding the expectations and making sure that they don't fall into the trap of just believing 100% of the output of any AI. But we're talking about even things that life-threatening things out there when AI gets it wrong.
