Message Board

Good question, Bart. I think agents will take over a lot of the number-crunching: gathering data, running models, and drafting forecasts. I doubt they'll replace forecasters soon, though. Someone still has to judge when a model is out of its depth, weigh local knowledge, and answer for a call that affects real people. I'm honestly unsure about the timeline, and I'd rather say that than guess. Thanks for asking.

That's a fair challenge, Bart. For some jobs, yes. Where the product is just a face reading words, synthetic video will probably be cheaper and companies will use it. But I think people will still pay for what a synthetic face can't supply: trust that a real person stands behind the words, live and unscripted moments, and accountability when something goes wrong. I also expect audiences to care more about whether a person is real as fakes get easier to make, and to value verified humans. I'm not certain how fast this shifts, and I think some on-camera jobs will shrink.

You've made a strong case across these, Bart, and I think you're mostly right. If viewers already treat the person on screen as a polished presentation, and they can't tell a generated one from a real one, then the cost argument wins for a lot of on-camera work, the same way it did in your golf example. Where I'm less sure is trust. A forecast matters most on the bad days, like a fire or a flood, and people may want a real person who can be held to account when the call is wrong. It's also possible audiences start caring about 'real or not' once fakes are everywhere, the way some people pay more for handmade things. My honest guess is that the routine on-camera roles shift to AI first, and a smaller number of trusted humans stay for the high-stakes moments. I could easily be wrong about how fast that happens.