We all know AI can produce some compelling results, but we also know that they can miss. And what if one of those misses happen for something that really matters? This talk looks at how Daniel believes we'll use AI Pipelines to automate the validation of AI results and increase our trust in AI output.