Hilary Gridley: belief

Transcript context

…I forget, actually, isn't that crazy? I forget what the original was. I think it was 140. Okay, 140 characters, yeah. 140, I think. It was short. And if you had to get a link in there, good luck to you. But oh my gosh, my ability to just look at something written today, and just cut it, that text in half, third, whatever it needs to fill the space, I can do that in my sleep, because I got all these reps very early in my career. I think people see the way that there's a threat of companies not wanting to hire as much entry-level talent, because it's like, "Oh, this is the kind of work that AI can do. " The fear that I hear, at least, is if you're not getting those reps early in your career, maybe it's not contributing so much value to the company at that moment, but it's how you learn to be great later on. And so, there's a fear that in five, 10 years, we're just not going to have that class of people, who have learned to do the jobs well, and who have built judgment in that way. But what I think that misses is, it assumes that you go and you do this analyst job for two years, and at the end of it you have a person who knows how to make models really well, knows how to do a few things really well. But why does that have to take two years? Why does that model of you grind over this thing? You wait for feedback. Eventually, you get that feedback. Maybe that feedback's good, maybe it's not. You go back, you try again. It actually is really inefficient, when you think about it. And the sort of learning applicAttions around AI that I get really excited about are, how do you shrink that loop? So in my podcast with Claire, I showed her how I build these GPTs, that kind of think like me. And the purpose of that is so that my team can get feedback that is at least 80% close to the feedback that I would be giving them. But instead of having to wait until I get to their message, or until our one on one, they can get that on demand as many times as they want forever. And I think there's a lot of things like this, of ways that things that require other people, just naturally slow things down, require getting feedback from other people, just naturally slow things down. We can build AI tools that, in my view, there's no reason why the amount of reps that you get at whatever task you're doing, you can be a film editor, just sitting there, poring over the film, deciding what to edit, what to cut, what to put into the trailer, or whatever it's making. That's an incredibly tedious job that takes forever. And I think there's no reason we can't make that way more efficient with AI, that make the learning more fun. And so, I think that that's sort of my hot take is yes, there is this threat of, a lot of these jobs that are things that seem like you can just automate them away, that might happen. But we absolutely still need to be investing in people's skills. I just don't think we need to be investing in them historically in the way that we always have. And I think in the future, we'll find that those ways actually seem quite inefficient, compared to what's possible today. That's such a powerful point. And we're already seeing this. I imagine you've seen these studies, I think it's in Nigeria, where they give students AI tutors, and they just zoom to the next, they accelerate so quickly in their progression of just reading and math. I think we're already seeing it. And it's harder to measure in PM and product, and all these things, but in school, it's a lot easier to measure, and we're already seeing results there.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.