The Exchanges

Q How much, how much noise should people pay attention to? Because you don't want to be, you know, in a silo, you know, in a mountain, coming up with ideas and a strategy, but you also don't want to be, like, glued to, like, You know, product hunt, 24 seven or whatever it is. So how do you, how do you think about that?

A I think you gotta ask like, where's the noise coming from? Um, the first year of Jasper, there was a lot of noise, but it was all inside of our Facebook group in our community. And it was so noisy and it was crazy and it was fun, but these are all of our customers. And so like, I'm just talking to them every single day, like all day, every day, riffing, laughing, making memes, building cool stuff. Like, and it was, it was wild, but it's like, those are our customers. And so like that noise matters A great amount. And I would say it's not noise. That's, that's signal. And, you know, you should be paying deeply attention to the things that your customers are saying, engaging with them. And I really didn't care at all about the outside world. I wasn't, I probably didn't tweet it all that year. I didn't like do anything, but just like, like the customers, like was my like social, uh, social network. Um, and that was really good. And yet if I hadn't been on Twitter six months before we launched Jasper, I wouldn't have seen GPT three come out. I wouldn't have seen some of the cool demos and I wouldn't have thought, well, I think we could do that for, for marketing content. Like that sounds pretty cool. And so, uh, I think it's a balance. I think, yeah, you've got to like, you know, be in the know, but I think it's very rare. Like on a day to day basis, it's rare that any, I would mis…

AI assessment note: “I think you gotta ask like, where's the noise coming from?”

Q consumer experiences, but first I want to talk about like the, you know, business context, the AI. You know, today a very small percentage of my workflow is AI enhanced. Um, that's changing, but it's small relative. Um, so maybe it's like five percent, three percent, you know, over the next five years, what percentage AI enhanced workflows do you see, you know, the modern day white collar worker having?

A I'd say over 90%. Maybe 95%. And in many ways, it'll just be like using tools. It won't be this Oh my gosh, I used AI in my workflow here. It'll just be, you know, the new generation will be confused at what it means to like not use AI in the workflow. And it doesn't mean that you'll only have like five percent of your job to do. It just means that you'll be able to sit on top of a mountain of AI generated, uh, or AI assisted, uh, work and context and summarization and idea generation and filtering and, you know, all of those things. It'll be like having Just a really incredible assistant or team of assistants working for you that just helps you be more productive and work on the stuff that you want to work on more, but I think it'll be a very large percentage of our time or work will be affected by AI.

AI assessment note: “I'd say over 90%. Maybe 95%.”

Q Love that. Uh, next question from Merkle. How do you think about Moat and Jasper.ai? Could you give us a little more insight into that?

A Yeah, I mean, I think there's, there's defensibility everywhere that businesses can pick up. You know, I don't know, both in B to B and in, in AI, in particular, like, if there's anything that's incredibly durable. Um, you know, it's not like the models themselves are these super moat, you know, intensive, durable things. They'll probably be more commoditized over time. Uh, the app layer, you know, same thing. I don't think we could just stop building on Jasper and it would be this great company, you know, over time, just cause we have this like durable moat. I think there's obviously a lot of product that you can build that just building great product that solves user needs, you know, is defensible. We've got a lot of great data that we collect and that we can use to go fine tune and train and perhaps even build like our own large language models someday. Um, that can like make our output different and better for our users than, uh, any other product would be. Um, and so we get value out of that, like distribution partnerships, brand community, uh, just talent, uh, internal organizational rigor. Like all of these things matter and are hard to do. And I think all of them, you know, add up to, you know, something that's really hard for someone to compete with. Um, but I don't think we've got some silver bullet, even the data stuff, which is probably the most like compelling, or …

AI assessment note: “I don't think we've got some silver bullet, even the data stuff”

Q understand like the zeitgeist a little bit more. I do feel that You, you know, what's, what's been helpful for me is just carving out, you know, once, twice a week where you can just go through all the content and write some notes on it and be like, okay, this is what happened this week. And this is what it means for me. What do you think about that?

A Totally. Yeah. I, I, I don't, you know, I'm not kind of methodical enough to, to structure quite like that, you know, but for me, yeah, it's like, you know, I don't have like Twitter on my phone anymore because like, It's, it'd be hard for me to actually, like, react well, like, on my phone, you know, so it's like, well, if I'm on my computer, I can, like, take, you know, take some quick notes, or, like, ping the right person on Slack, or, like, whatever. If I'm on my phone, I'm just gonna be, like, distracted with, like, my family, uh, distracted, you know, there, probably, like, overreact to something, not get it to the right person, like, a thoughtful way, and it's just gonna, like, go poorly. So I think, yeah, like, I still, like, synthesize a lot of it, but I really try not to react on anything, you know, too quickly. And I just tell our team like, Hey, listen, like our strategy, we've got a great strategy. We've got the things that we know we need to do in the next three months. If we do those things, I think we can all agree that that's going to be really great for us. And if we don't do those things, it'd be really bad. And right now the biggest reason we wouldn't do those things is because we're all caught up and knee jerky and going crazy about the stuff that we're seeing. And so like, unless anything you're seeing Like, like pokes a hole in some part of our strategy.…

AI assessment note: “I'm not kind of methodical enough to, to structure quite like that”

Q Um, what's, uh, what's something that people, the vast majority of people misunderstand about the world of AI? Like what's something that you wish that people really understood that they don't currently?

A Um, I mean, I've said this before twice now, so maybe I'll think it's something else, but yeah, I think I get a lot of it is it's, I see people looking for like, like they've got a hammer and they're just looking to like, we're looking for nails and they're trying to fit this onto everything else and they're not doing the hard work. Like, like you're still building a business. Businesses have always been when you find a group of customers that have a need and you can solve that need. And ideally solve it for, you know, a fair exchange of value that everyone's happy with. And you can do that repeatably. And there's more of those people than like, you can grow a business. Like that is still true today. And I think people think that for some reason that has changed and that AI is just something you can just jam anywhere in the world and kind of create some business. And there's just a lot of kind of copycat, just, you know, it's like on Twitter, just copies of copies of copies. When like, again, I could probably go walk down You know, downtown Austin and just start talking to people and come up with like five ideas that nobody on Twitter has ever thought of. There's no businesses there. And like some of those could be really great businesses. And so, um, I just think it's really cool technology, but it's not a business in and of itself. It's just enhances and like, I think acceler…

AI assessment note: “it's really cool technology, but it's not a business in and of itself”

Q Is too much magic, AI magic, a bad thing?

A I know I do wonder how much we care about like knowing that something is from a real person versus like, let's say, uh, like I don't find my, someone's like, Hey, this Twitter thread was AI generated and they kind of call it out upfront. And then like, I just copy and pasted it. Like, I don't really like want to read it as a person. Cause I'm like, you know, whatever, maybe there's some golden there, maybe there's not. Now, if they're kind of curating it and thinking it through, and I know that, Hey, they kind of used AI, but also like We're very thoughtful about it. Then like, it kind of goes back to, I really want to know this person's thoughts, but

AI assessment note: “I do wonder how much we care about like knowing that something is from a real person”