Interview #8: Andrew Roth

Noah: All right, so I'm here with Andrew Roth. Drew, why don't you go ahead and tell me about yourself.

Andrew: Yeah, thanks, Noah. I I'm Andrew. I'm the founder of DCDX. It's a Gen Z insights and strategy firm, helping brands understand next generation of consumers. Running that for about five years. And also recently spun out a company called Offline Helping Brands discover and connect with in-person communities. A couple months ago. So that's getting up and running. Been doing that. I'm outta Chicago recently and 26, almost 27. Getting, getting up there. But yeah. And met you a long time ago. Way back. Long time.

Noah: Very cool. So how'd you get started with DCDX and just give a little background about what you guys do.

Andrew: Yeah, I, I started in 2020, was in school, kinda finishing up COVID COVID education and pretty big disconnect in culture between organizations and young people. And so we started as a way for brands to help give them a better understanding of how to communicate with young people, how to build products and services that really reach and resonate with a young audience. And honestly thought it would be a summer project and then forget about it, but got some good traction. Chipotle became our first client, and after that, things went uphill. And obviously, UPS and down since then. But that's where we've been over the last five years, working with a pretty wide variety of clients across industries from a lot of technology. Spotify, Hinge, a lot of dating apps. To L'Oreal and Beauty Crocs in fashion. So yeah, pretty wide range there. And like I said, recently did some work with Eventbrite. They led us down this path of in-person connection, how people are gathering today and getting away from Al in many ways. And no, I'm just kidding. And led us to spin out and a new company called Offline as well.

Noah: Very cool. And what does offline do specifically?

Andrew: Great question. We are. Building the infrastructure for in-person communities to monetize through brand partnerships and advertising. So run clubs, dinner parties, game nights, all these different communities that have popped up in the last 12, 24 months. And mostly how young people are meeting and gathering in a post-college, post COVID world. They are. Struggling to monetize ticketing is really the only thing. And so we're bringing in brand partnerships into the equation and kind of like same way creators and influencers took off. You're an NIL guy, like their whole world of that marketplace opened up really quickly and seeing the same thing happen in kinda this community side. And so trying to quickly take up kinda build some infrastructure on that. Super cool. Super cool.

Noah: Yeah. Alright, so the big topic today is AI. I've been following and tracking the space for, I don't know, six, 12 months now. And then building it for the past month or so, maybe two months now. And I just wanna ask you your general take on the space, and then we can get specific into how you're using it on a day-to-day basis if you are. And then how you're using your business workflows and things of that nature. So just first broad question, what's your general take on where the Al space is today?

Andrew: My take is ask me tomorrow. 'Cause it's changing so quickly. I think everything, everything is just accelerating at a pace that I've never seen tech, do even, 12 months ago, like the way we integrated it looks fundamentally very different from how we do it now. And I just think that's, again, that's continuing to change very quickly. So we, we use it, we leverage it. I think there's so many benefits to it. We also think about the consequences of it too. A lot of our work with Offline and is about it's kinda almost benefiting from, the, because AI is making it easier to not to just fabricate things, but create and keep people online in, in those circles. There's also a strong bull case for, offline and bringing people together. So I think there's, there are pros and cons in many different ways, but it's obviously not going anywhere, so.

Noah: Definitely. Which AI tools are you using primarily?

Andrew: I would say mostly GPT and Claude honestly just have both of them up almost all the time. Bouncing, sometimes I'll ask them the same questions, you'll give me a better answer and just, rerun and things like that. We're doing a lot of tech on the Offline side, so starting to build with some no code, local low-code, low-code tools, Versa. Figma honestly, is a great new AI tool Figma make, and just a couple other things we're working on. Yeah, pretty widespread in terms of how we use it, but never like Gemini or I don't even know what being, what's the Bing one? I don't know. Yeah. On any of those guys. Cool. Yeah, I think I love Claude. I love Claude Code specifically. Super helpful. Yeah.

Noah: In development, what's your take just in general on how those companies as a whole are doing, or which ones, if you're following any of the macro stuff with those companies.

Andrew: Yeah, I there's a great podcast I listen to from on Lenny's newsletter from the CEO of Philanthropic. I think the way they, are approaching AI from a pretty human first lens is impor is very important. And I prefer using Claude where possible. Not just because of that. Not just because I think a great tool, but also because of their kind of moral structure of how they want the future to look. Yeah, honestly I think. Sam Altman's in a money race to get just, not even about money, but just about being the first to do something cool. And I think AGI is going without limits on that end is what it seems like people, people are, have left there openly to go other places that are more moral. So yeah, that obviously is concerning. But it also is so integrated into everything now that it's hard to ignore. And yeah. I think generally like, pro companies that think about the consequences of what they're doing.

Noah: Yeah. That's not a bad stance to have. In terms of DCDX and business specifically, maybe a bit less with Offline, but where are you using AI? What workflows are you building with AI? I know you mentioned a little bit of build no code stuff. Specifically, what type of workflows are you using for AI?

Andrew: Yeah, I think the most common one is honestly, on calls or interviews, things where I'm in conversations like just transcribing note taking, and then syncing into, automated workflows after that. I actually, Notion has been doing, stepping up their AI game too, and so Notion is our like home for everything because we keep, all our notes in there. But now all the conversations I have go into Notion and all the calls are logged and recorded in there, and so we can easily plug those into the next steps. I also typically, like I, it allows us, allows me to feel freer too. Engaged, like more genuinely in a conversation because otherwise I've just been type, typing out notes and making my computer shake left and right. I think that's been really helpful. And I'll build like knowledge bases in Claude and GPT just uploading like different. Decks or project scopes around it. And then like when I have a conversation with someone, I'll throw that in the knowledge base. So it's just like a living and breathing brain around the project, which is cool. Yeah, we're trying a lot of different things. I think it's all still all experimentation, but at least there's some things that feel pretty routine. And now if I go into a call without my handy dandy AI assistant, I don't really know my memory is is tested.

Noah: Yeah. That's valid. Same.

Andrew: Same. Yeah, it's still, we're still super early in the space.

Noah: Obviously everything is somewhat experimental, different people are trying different things. Some things work, some things don't. Have you had experiences or what are those experiences where you've tried something with AI, it hasn't worked out, and then how have you tried to overcome that?

Andrew: Yeah, I think, I just think there's certain things that we've learned to use it for and not to use it for. I think writing is just like a really, writing and create. Not just creativity in general, but like the, it is so clear when people use AI written like copy and even our, people who are working on projects. Like I know when there's something that is like a part of a deliverable that's been written by AI, it has a pretty clear syntax and it all just feels very similar and familiar which is not how it should feel, especially in the work we're doing and very like sharp insights. So I think we, we. Not only how we communicate it to the team, but I think we tend, and I tend to use it for more operational or like organizational things and recall and just I think things where it's less about net new and more about either recovering or revising or synthesizing something that is from the past. Or, getting, even getting up to speed on now. But I think the danger we found is like using it to make progress forward versus catching up on to on where things are. And we'll throw a, we'll throw a, we'll get a, you have to do the work to get like a template. You have to, you have to put your own energy and time into getting a template, right? But once you get that then you can throw a conversation transcript against the template and just say to re, recreate that and it does it well, but you have to set that foundation first. So I think some of those things are things we've learned through just trying to, you can spend an hour trying to tweak something, but if you just take 20 minutes to do it yourself and then have that template going forwards, that'll save you a lot more time going forwards too.

Noah: Yeah, that makes sense. I call, I call those templates, workflows primarily. I think when we were pre GPT five, it seemed like everything was going so fast that AGI was knocking on the doorstep. ASI was maybe coming soon. But there was no, no holds barred. Everything was going just incredibly fast. And once GPT five dropped, we saw a bit of lack of excitement in the general AI space in terms of, okay, maybe AGI isn't coming tomorrow. Maybe things are slowing down just a bit. But what that also means is we're also starting to see a lot of focus towards those workflows, those templates being used in a business setting where people are not necessarily replacing employees with AI, but updating their efficiency with AI. And with that in mind, how would you say your efficiency has changed with the use of AI? Whether it's what you mentioned with not having to type notes in meetings and actually being more attentive. Or just kinda any other workflows you're running, where would you say just 10%. A hundred percent. 500%. How would you say your efficiency is has changed with AI?

Andrew: Yeah, that's a great question. I would've loved to if I could go back and see how fast I was working like before. This happened and compare it to now. So until the time machine's invented we'll wait on that one, but it's probably like 50%. Like I think I have one and a half of me now instead of just one. I think there's there's just so much more I can get done in a day and at a cheaper way of doing it too. Even even legal and financial stuff, right? They can go in. Talk to our lawyer for an hour and you get a bad bill afterwards or just, again, not that we're basing our big decisions off of that, but at least getting 90, 80, 90% there. So you can ask a more intelligent and sharp question that can only take 10 minutes of their time. I think those are things that are just are helping a lot. Yeah, so I generally like probably, yeah, probably 0.5 of me on top, which is great. Which I love. We'll see if we can get a duplicate going soon, but, but yeah.

Noah: Yeah, that's super helpful. What are you most excited about just coming up in the AI space and we'll talk, don't answer the full question yet, 'cause I also wanna ask you what you're gonna see, what you think you might see in five or 10 years. But what are you most excited about maybe in one year or two years of what AI can do for you?

Andrew: I think the really exciting thing is just like the barrier to entry from. For starting things is so low now. Even it has been decreasing, but now just I really believe anyone can start something now. You don't need, I'm not technical, don't know how to code, but we can get, get an MVP up and running and I think there's like a, or there's certain for so long, like the. The mantra has been higher. Be an engineer because those are the most valuable people. And not that's, there's certainly still a value to that, but I think that idea has the kind of the crystal ball ground that shattered a little bit. And I think there's just like entrepreneurship is now, there's a more exciting category to be in because there's the, any idea is a good, is an operable one now. Which is interesting. And so I'm excited about that. I think that means just more. It means it has good and bad consequences. It means there's more good ideas that will come. There's, it means there's also a shit ton more of bad ones too. And so I think there will be some really exciting companies born that are being born now, that are, will continue to be born in the very near future. Which is awesome. I'm excited for that. I dunno if that's the direction you wanted me going with that, but that's what excites me about the space.

Noah: Yeah, that's perfect. Okay, so that now leads me to my next question. Five years down the line, 10 years down the line, where do you see AI going? And if you don't have an answer, that's fine, because it's very hard to predict. Obviously, with it changing every day it's hard to predict a week, a month, a year out. But if you had to just think about five, 10 years down the line, what might you think about in that direction?

Andrew: Yeah, we'll have AGI by then, that is for sure. And I don't think anyone knows what that will be like to be in a world where that is around us all the time. I don't know. I honestly I think that's why a lot of our work and thesis has been around the. Like the swing back or swing against that, which is the offline world. I think there will be, because it'll be so abundant in so many different places. Like the, there will be a much higher value on spaces away from that. That feel more real, I would say, not that things are fake, but I think that feel like an escape from that world. So I think that's where like our, that's where my thesis on the future lies is like, there will. This, there's so many questions on what that will look like. We know it will look like something but what we haven't focused on as much is like the value that will come from just in-person connection alongside of that too.

Noah: Yeah. That's really smart. That's a great build. Cool. So going back to your previous answer too. It's interesting to note just because so many people are building things right now, there's almost like a signature similar to what you mentioned with writing, with building with a quote unquote vibe coding where you can, or I started to see, oh, this website was built with Claude or Claude Code, which isn't a bad thing. I use Claude Code to build some websites and some different things, but it's just funny how we can start to sense. The difference in building between AI and human.

Andrew: Yeah. And it's it's funny 'cause it has a, it has like a personality almost, right? Like I feel like we'll know. And when AGI is there, like what it, we'll be like, oh that's, that's AGI talking, right? I feel like it has this identity to it already, which is scary. But I also think that's what makes. Novel, like thinking even more valuable, right? Like when you are different from that, it'll be very clear too.

Noah: Yeah, absolutely. Yeah. Okay. I think I got about two more questions for you. Yeah. First, personal projects, are you using, or not even just personal projects, just in personal life, are you using chat, GBT or AI? Are you asking GBT what you should be wearing on a date? Are you, how are you talking to AI if you are in your, in personal life, outside of business?

Andrew: Yeah. Still a lot. I if you see me around Chicago talking in my AirPods, it's most likely to catch BT and having a conversation which I guess is less personal, but still work-wise. But I think just like I'd sometimes helpful to talk things out loud and when friends aren't available. Have a conversation. Someone is and they are on my phone. I do that a lot. I do not a lot, not that often. I'm not crazy. I do it a lot. Yeah. It's nice. It's fun, they care for you. They so yeah. Cooking definitely a, if I got like random ingredients left over, I'll be like. What can I make with this? Some trips, planning trips, just figuring out where to go do stuff. Yeah, I think those are probably the most common ones. I'm trying to think if there's yeah, that's probably it.

Noah: Yeah. That's what it's best for right now. Yeah. Super useful. Okay. And then if you were to talk to, say, someone who's just graduating high school and starting in college, or someone who just graduated college and is looking to get a job just in any, not in any specific field, but just in general, what would some advice be that you would give to that person? In either case?

Andrew: Yeah. I think the, I would, it'd be so fascinating to be going through high school or college right now with these tools. I think there's a lot of I just I would honestly, it would be a caution to. It would be a yes. Learn everything you can about them and how to work with them. But also do not be dependent on them. 'Cause I think that's the dependency on them is what, like that when people say AI is gonna take your job. Like that dependency is the things that AI will take over. And it's make sure you maintain your independence in thought. Because if you follow. If your thought follows AI thinking, it'll all blend together and there won't be those new things that come out of it. And so I think. That's the risk that's posed to a generation coming into the workforce to use these things is a sameness that kind of starts to blend together. Where everyone will come in with the same capabilities, with the same skill sets, with the same like ability to execute on things. And some people will be marginally better at using AI. Yes. But the difference won't be in how they use it. It'll be actually how they don't. And I think that is like something that is easy to overlook in my opinion, in like how this will. Get integrated because everyone, you and I are not the only people using this. There are billions of people across the world. And so like how do you stand out from them in a job hunt? You have to right. In your own different ways.

Noah: Yeah. That's a really good answer. Cool. I actually did wanna bring up one more thing. Yeah. Another on the subject too. I've seen a lot of reports recently. I'm actually gonna be talking about this with a former professor I had as well. But I've seen a lot of reports about AI psychosis and things of. Gen Z and younger people talking to AI almost like a companion and things of that nature. And when GPT-4 switched to GT five, we saw a lot of revolt because the personality style was different. And we've seen cases where people are almost losing their minds. Talking to the AI and assuming it's a real person almost, and there's some there, there's a lot to get into there, but just surface level kind of what are your thoughts on AI psychosis in general and how to maybe put in safeguards or avoid dependence on AI or over personal dependence on AI, maybe.

Andrew: Yeah, and I appreciate you bringing it up. It's an important topic. I honestly, I have a friend who is experiencing that. It's scary. Like it's not there aren't just stories. Like it's really, it's true and it's happening. And I think it will only happen more, especially when we're facing a loneliness crisis. And like an easy, convenient solution is solving it with, with t I wish I had better answers for what the safeguards are. I just, I think ultimately it comes down to if we continue to prioritize speed and like firstness in this, then that will happen as a consequence. It's what happened with social media and like how we all became addicted to our phones. Like it just is, it was build and then think, and I think. I think that that just needs to be flipped. And even if that means like not being first to AGI, which like, I, I just think, I mean it's unfortunately that has to be the trade off. That might have to happen. But there is an arms race for these major companies to build AGI first and I hopefully hopefully first in this case, does not mean best. And I think that's like ultimately what has to happen because people like philanthropic are taking this into account for what they want. They're building the, the morals and commandments into into the, into their like thinking process there. So yeah, I don't know. I just I hope we're able to address it because that's, we're building the future and if we can't build that, then we don't deserve to, to be there.

Noah: Yeah. It's a problem. Yeah. Cool. Yeah, it definitely makes sense why you choose anthropic as your go-to in terms of company. 'Cause that's, that is something that they're, they seem to be stronger on in terms of their direction. All so anything, any plugs, anything you wanna address that we didn't get to today?

Andrew: Think plugs, no, after this, go touch some grass, gotta get gotta get outside, talk to people. I'll be typing up on the computer too, but always important to do that. So that's my plug.

Noah: Awesome. All right, Andrew, I appreciate your time today.

Andrew: Thanks, Noah. Appreciate you.