I’ve already done some training for my company where the ‘people’ were AI/LLM created. You figure this is first generation so it’s only going to get harder to distinguish between it and real people.
Holy moly.
It gets irritating how much the ‘presenters’ agree with each other, but I ran this past a couple of people and they were taken in.
I find it hard to tell these aren’t humans so much as overly bouncy presenters of the kind I usually avoid in podcasts.
I’m late to the party but this is the first thing about AI that’s actually impressed me or given me pause. Bob – was this literally just a matter of giving it your post and letting it run with it, or did you do any tuning? The overview link seems to suggest you put in the “source”, press a button and done. If so … this is pretty impressive as a first cut implementation.