Episode Transcript
Kyle James (00:00.61)
Hey, welcome to the ad chronicles podcast. I'm your host Kyle James. And today we'll be talking about how a digital design company called MetaJive is using AI inside of their own business. And we'll share the exact steps that you can take in order to implement AI for yourself. Now, before we dive into that, listen closely. Are you looking to implement AI inside of your own company or maybe just struggling to get your AI to stop hallucinating? Speak to GPT trainer. GPT trainer literally builds out and manages your AI for you.
eliminating hallucinations for good. Go to GPT-trainer.com. I promise you, it'll be the biggest time saving decision that you've made all year. Trying to set up AI on your own is like trying to build a house from scratch. Sure, you could do it, but the time of frustration is going to take you to get it finished. It may not be worth it. It's a thousand times faster and safer to hire professionals. Once again, that's GPT-trainer.com.
Today I have with me on the show, Dave Bitten, is the CEO of Metajive and a creative strategist who has led over 350 digital products at the intersection of design, technology and innovation. With deep experience in building user centered experiences, Dave brings a sharp perspective on how AI is reshaping the future of creative work. So excited to have this conversation. Hey Dave, welcome in. How are you doing?
Dave (01:22.785)
I'm doing great. Thanks for having me, Kyle.
Kyle James (01:24.962)
Yeah, man, for sure. So give us some background here. Like how did MetaJive come to be and show us a little bit of paint that picture of the origin story of you found in the company.
Dave (01:32.638)
Yeah, so, manager can be an accident, right? So, I got a project when I was in college to help some buddies, build a website for some snowboards. did that. then graduated, just going to get a job, but instead I got a contract. helped build out Disney world.com as I was looking for a job when that contract was done, I got another contract and then, turned out that was project runway. Then I got to work on.
Xbox 360 launch. Sorry, I'm starting to date myself as an old guy there. And then I just kept intending to get a job, but kept getting another contract. And eventually I needed some help and I had a development partner for like 15 years. And we worked together, we grew our businesses separately. And then, yeah, we just slowly kept leaning into this being a real business.
Kyle James (02:26.732)
Yeah. How did you, I gotta go back to like the contract part. Like that's, it's pretty rare. Most likely must be out of college. Like I got a job. Yeah. They gave me an offer, but in this case, like you went from one project to the other. Like, how did you get, how did you go from like, how'd you get Xbox 360 and Disney? We're like, those are not exactly small projects. so how, what was that process there?
Dave (02:42.396)
Yeah, so while I was looking for a job, I was interviewing all these small studios doing innovative work, like things I really wanted to work on. And the one guy was just going, hey, want to give you, he's like, these guys want me to do this thing, but I have a studio. Like, I can't just leave and do this. But I don't have any more room for more people on my team. So like, I'm just going to introduce you. And then I got that. then
Right, Xbox 360 wasn't like Microsoft hiring me, it was like an agency who I'd helped with a pitch, was like, oh, come help us with this thing. The Project Runway is the funniest story because on Project Runway, the accountant knew my Disney World contract was up before I did. And then she's like, you should come hang out with me and my husband. He's involved in AIGA. I went, bought a round of beers. Her husband's friend.
Right, just super random. happened to be the president of AIGA. But his friend's like, oh, my buddy John, right? So it's like six degrees of separation there. He's like, my buddy John needs some help. And they just want to hire a person, not an agency for this project, because they don't think it's going to be successful. And when I talked to him, goes, oh, Brian says you're good. like, I don't need to see your portfolio. And I was like, snuck my way in there.
Kyle James (04:04.046)
Wow. That's, that's like the power of referrals and like the network in that regard, like that obviously took you, cause that's like rare. It's rare to have like this kind of types of opportunities. Like it doesn't come to on everyone's on everyone's table in this case, like because you had those relationships that just tied in from one to the other.
Dave (04:20.07)
Yeah, they were all really fresh relationships, so I got really lucky.
Kyle James (04:23.126)
Yeah, that's so cool. That's, that's incredible, man. So now you're using AI over at MetaJive. I love what you guys are doing. And like, tell me like, why, why did you start using AI in the first place? And like, what types of challenges were you trying to solve with it?
Dave (04:36.009)
So I think how we use AI at Metajive is pretty differentiated. We're lucky. We work with Google on AI projects. We help them build really interesting things, like the front-end interface for Flow TV, where we built a bunch of prototypes for what became Flow, like the VO thing. Obviously, we didn't train VO. We didn't do that. We have other clients, like the Future Laboratory, which is a trend forecasting service. And we built them a
text-based agent that is trained on every single article, every podcast they're on. These guys are prolific content creators. a lot of standard programming goes into just the plumbing to keep feeding everything into.
the AI infrastructure, whether you're using Vertex or I can't think of the AWS one or Katana or right, like there's all the levels of what you're building there. And then, right, everyone on our dev team uses different CLIs. And then in our design team, we're using it to generate images. We have a whole practice around that. And where we're really excited about it is actually in the
higher end knowledge work, the stuff where you think wouldn't be accessible. So copywriting is something you'd expect, and we use it a lot, but we spend a lot of time and effort tuning. There's a lazy way and an engaging way to search for better with AI. And I think we tend to push, spend the same, if not more time to do the same thing, but get a better output. We also...
have been pushing on it as strategy and continuing to ask deeper and deeper, deeper questions. And just the depth of experience of finding things, you of course have to fact check everything that comes out. But as you keep pushing in, you really get to shape a lot of ideas. So from building AI products for the best and the brightest and getting exposed, that's super rewarding.
Dave (06:53.725)
something like automated onboarding, offboarding, something like that, and then pushing in on strategy, pushing in. Even our designers are putting web design comps in and asking for feedback and training it. And then I'm also recording almost all of my meetings and then referencing, I have automations and batches and things. So at the end of the week, like give me feedback on, like,
Kyle James (07:07.852)
Hmm
Dave (07:22.599)
how I behaved in meetings, what were my strongest points, what were my weakest moments, what were, like, what's all of my takeaways?
Kyle James (07:29.644)
Yeah. Yeah, for sure. So you mentioned something earlier about like the lazy versus like engaging. Like, can you like, can I break that down a little bit? Like, what do mean? Like, cause I think there's a lot of people who, like, I think I understand like the lazy part of it, like kind of just like, what is that, difference between the lazy and engaging? Like what results are typically being pulled from like each of those kinds of like directions when you're pursuing the AI.
Dave (07:49.949)
So perfect example. There's a million of these. But I think in general, this is a really big topic in AI right now. So.
Dave (08:04.305)
Someone on my team, I had asked a question to someone else. Someone essentially put that question into chat GPT and posted like the answer. That was extremely obvious, extremely low grade, didn't leverage their skillset, my skillset, the other person on the team's skillset. And I was just like, never do that again. Like, we're fully in on AI, but it's like that answer.
which comes in a very specific format, right? In a very specific way. Like, you know it when you just see it. And it's like, you spent one minute copying and pasting a question. And like, if I wanted to ask Chatjapt that question, I would have asked Chatjapt that question. So it was like, don't ever do that again. But the other example is, we have an internal newsletter in the holding company we're part of. We had sent out everything for approval and I saw it and the headline wasn't quite right.
Kyle James (08:38.615)
Yeah.
Dave (09:03.13)
And I said...
Hey, I need to get your replacement. Give me a little bit. And I put that headline in. I put the body copy in. I explain the context of what we're doing. And like, you know, the prompt is like this long at the end of it. And I and I don't say rewrite it. I say, give me 10 options. And then I go through each one and go one is OK. Two is good. Three is this. I like this part of four right on every single one. And I go give me five more options.
Kyle James (09:18.541)
Mm-hmm.
Dave (09:35.844)
And then it's that way where I say like, give me five different strategic options and just continuing to push. So if I was just going to write that headline, I could have written that headline and it probably would have been 80 % as good and it probably would have even been faster. But by doing the reps and understanding what's going in and being strategic about what it was, I actually think the copy that came out of it was better. And then because I did that, I said, the body copy could be better.
Kyle James (09:53.42)
Hmm.
Dave (10:04.7)
So because I had already done that work, I said, can you also update this to match what we just did in the headline? Again, we're talking about a total of four sentences of copy. This isn't a book. And then it gave me that, and I said, can I get two more options of this? And then I copied and pasted those and put that in. So I think, again, when you're pushing for better, you're going to burn a lot of tokens. You're going to go through the prompts, and it's not always faster.
Kyle James (10:22.551)
Hmm.
Kyle James (10:29.26)
Yeah. Yeah.
Dave (10:34.139)
And a lot of things are not faster with AI. So you better get to better.
Kyle James (10:38.84)
Yeah. I love that. Yeah. I think that's, that's a big deal. Cause I think it's almost saying this point, kind of insulting when someone just gives you like a one shot prompt and then copy paste. Like that's like, didn't do any homework. didn't do no, no sweat was poured into this. And in this case, it's like, Hey, it's gotta be more of that, that chain of thought, like deeper engaging, or you're, digging deeper into the AI to get more like better answers and better content in this case. And like kind of transitioning over.
on this next question is just like for you, mean, there's a lot of different things you're doing with AI, both like internally and externally, but like, are some of those like maybe key results that are worth sharing and maybe some things you've been seeing since you've been implementing a lot of the AI, both internally and externally within your business.
Dave (11:24.091)
Yeah, so I think internally, what I'm hoping for is the like, I was listening to a podcast this week and they were going, yeah, like dev should be getting five to 10 % better per month because the tools are better. And I'm like, there is no way, or form that I'm seeing a 10%. I may be seeing a 10 % annual increase in output, not a 10 % per month.
like increase in productivity with devs. And every time I hear that I'm like, do we need to get a different set of tools? we need to lean in deeper? Like how could we leverage it? But I think because we hold really high standards, again, it doesn't come out in that. That's a KPI. That's one of those low hanging fruit effective KPIs that everyone wants to measure. And like my team is better. I think the team being able to do things that they couldn't do.
which might not be an effective measurement tool, but it's something I really look at. So for instance, our executive creative director, amazing designer, amazing thinker, thought-provoking, not a writer. I have a 30-page document to review of all of the copy for our new website, and he's gone in, like, done that work on 25 headlines on all of the things and written everything.
because he knows the brand, knows who we are, he knows how to articulate, but he's not a writer and he's able to do it with a special writing API or writing LLM. So I think that's special. think we are seeing revenue in creating assets and generating them and really selling that as a service. like we typically don't have a lot of photo and video shoots and like we're not that type of.
Kyle James (13:03.384)
Hmm.
Dave (13:21.347)
design agency, but when we know what we want and clients are comfortable going, okay, I can't own the copyright on this, which is always a fun conversation with the lawyers.
That's one theoretical revenue we wouldn't have had before. And two, like we're really in doing it and really exciting it. And it's making our projects better again. So those are the things I think about. Now on the future laboratories project, the AI chatbot that's trained on all their things, we're noticing, and I don't have all the stats.
right now because, we're, we're working on the case study for it and we're working on it together is what we're really focused on is, okay, we spent money to build this tool. We're going to build this. Let's focus on like, are we lowering churn? Are we increasing the number of times per month someone's coming on? Like what are all their success metrics for like how much value they're providing kind of the same way Netflix does. Like Netflix wants you to watch 14 hours of
TV a day because that means you're getting a lot of value from Netflix. We're looking at all the same metrics Netflix uses for how strategists are engaging with their product.
Kyle James (14:33.902)
Mmm.
Kyle James (14:41.932)
Hmm. Yeah, that's, that's so interesting, man. And I appreciate you sharing the, like the perspective, like even like talking about like the, when you said like the, with your, not your copywriter, but like your, your debt, your, your director, your creative director is like his expertise isn't on the copywriting or like understanding that part of it, but like, because he's got AI, it's almost like that weak part that he did have. It's almost like that's his, I don't say call it a crutch, but like, it's almost like that support beam, like because it's weakness now it's more of a.
Dave (14:54.52)
Grange record.
Kyle James (15:10.912)
Not necessarily a high strength, but a better like strength than it was before previously. think like, think it's.
Dave (15:17.433)
Well, sorry, one of things to clarify there is like, it is a strength. Like if he wasn't doing it, we would be one spending money hiring copywriter and like, I love our copywriters, but I actually think a lot of the copy is better because he has taste and he has comprehension and he knows the brand as well. So it's not, we don't have to explain to someone and onboard them to the project because he's doing it, which again, I think the copy is better than when we had a copywriter.
Kyle James (15:36.974)
Mmm.
Kyle James (15:46.978)
Yeah. Cause like almost like you had to like, if there was a copywriter, now you have to like almost hand it off to them and they may not have that same perspective as your creative director has. And therefore it's not like cohesively working together, which can be a challenge. And it's a very interesting perspective. Never thought of it like that way, but
Dave (15:55.99)
Yeah.
Dave (16:03.533)
Yeah, also personally, I use it a lot for...
getting feedback before the real feedback. So again, just like the designers putting in and saying, hey, give me feedback. Tell me you're a designer, you're an editor from this publication, you're a client of CMO of Fortune 500 company, whatever it is. Give me feedback on how you're interpreting this. Because again, I think assigning a role to your GPT is the best tip I've ever gotten. Just saying you're a Michelin star chef, like help make my dinner better.
Those sorts of things are amazing. And so we have a board meeting, we're owned by a holding company now. I put the entire presentation in and said, give me all of the questions I need to be ready to answer. And then evaluate, give me feedback on those answers. And so I had like, and I did that one through voice and I had like a probably 25, 30 minute conversation with the
Kyle James (17:00.814)
Mm.
Dave (17:10.38)
GPT and it was it was really helpful and I'd say 80 % of the questions that it grilled me on
Got asked. So it had pretty good foresight. I almost wish I could have changed the tone to be a little more board-like in how they asked as opposed to write like, I find especially Chetchi Bt gets very agreeable very fast. I don't need, like the number of times it'll say to me, like, I know you don't have time for imperfections. You want things done the right way and you want straight talk. Like I'll give you straight talk. And it's like, yeah.
Kyle James (17:21.228)
Mmm, wow.
Kyle James (17:29.389)
Yeah, right.
Dave (17:49.867)
but with a chipper attitude, right? you're just like, don't, like it doesn't have to be, I don't need my agent being super, super happy.
Kyle James (17:51.438)
Yeah, yeah.
Kyle James (18:00.462)
Yeah. Yeah. Kind of like a little bit of like people pleasing sometimes. Oh, I agree with you a hundred percent. I understand where you're coming from. Like, do you really machine? I don't know. But like, yeah, having a little bit of pushback. Cause like you need that challenging perspective. Like that's typically why like you have different teammates is because you're like, Oh, I need a different perspective. That's going to challenge on like our current, you know, vision that we have. And so, and so now tell me this, Dave, like you, you guys have obviously you're using AI personally inside the, inside the team and obviously externally as well, but like.
Dave (18:07.64)
Yeah.
Kyle James (18:30.082)
What are some of those maybe upcoming AI initiatives that Metajive has planned and like, where do you see it playing some of like the biggest role in your operations next?
Dave (18:39.905)
So we have a whole roadmap of places we want to implement AI within MediJive. And some, and the ones we're gonna address first are the lowest hanging fruit. The ones that are like, help accounting do accounts payable, right? An invoice comes in an email box, that invoice needs to get approved by someone, it needs to get added to a spreadsheet, a deal needs to go and it's just like, here's a bunch of flows, right? But no, you can't.
really automate it without AI because you need.
actually interpret what the value is in that. Does the invoice have a PO number, et cetera, et cetera, and fill all that in, dissect that information. And so there's a bunch of just internal operations, because everything I want to do is take the team off the busy work and let AI just run with those. We're also just really excited to build these things and talk
about them and talk about what we learned and share them with other people because the more we put out, the more we get back. So when I talk about like automated accounts payable or how we're going to keep human in the loop, because that's obviously a big thing. You can't mess up. can't not pay people. You can't pay them too much.
Kyle James (19:53.582)
Mm.
Dave (20:05.973)
the more we get back, tons of feedback. We did this and we ran into that, or I'm interested in doing that, we're working on this project. And I think that's the most exciting thing about this community is that like a year ago, I would have felt really confident actually that like we're gonna have, we're gonna pivot our business in a lot of ways and build AI tools for people, build in automations, build in.
these things. And I'm starting to go.
I think that's gonna be part of our business, but I don't think that's the whole future because I think so much is gonna happen in...
want to call it a browser, but it's not in the dialogue box. Clearly that experience needs to be upgraded. I hope that's what Johnny Ive was working on with OpenAI. He's not a software guy, but that's a whole other conversation. how are we getting pertinent information? I don't think the world is going to be happy with black text in a white box for another two years. I also go, we have to get
Kyle James (20:52.376)
Mm-hmm.
Kyle James (21:01.154)
Yeah.
Dave (21:17.628)
MCP and A2A is ready for a lot of our clients. We have to get all of these things working so an agent can order a hotel room, book a car, help with all of these things that our clients do, move, lessons, schedule a trial, things like that, full A2A. Like we're gonna build a book a meeting, A2A.
bot for MetaJive, not that we expect many people to use it, but it's going to be there. It's going to be proof of concept and yeah, right. We'll link it with Calendly and make sure it all happens and just get it so the bots can start talking to each other and just, just build experience. And then I think AEO and how that's structuring, like a webpage experience is a really important piece of what's going on. And then
Kyle James (22:00.012)
Yeah.
Dave (22:16.969)
actually providing AEO for other people is becoming a bigger part of our business. But there's kind of no part of what we already do that's not going to be affected by it.
I say it every year and I say, the puck's speeding up and all these things, but I never quite expect it. Like, I'm still surprised when it keeps gaining momentum. It keeps outperforming expectations really exponentially.
Kyle James (22:48.014)
Yeah. Yeah. It's moving. It's moving at a, it feels like a light speed because like one moment you have something built out and next moment, like, you know, like they always hear a lot of conversations about like prompting as an example, like, Oh, the prompts going to go away. The there is still a prompt for sure. But like you could see, like even with some of the reasoning models now, like where you put in a single prompt and it's very basic to send us along. AI is taking that and going, okay, let me try and interpret what they're actually trying to do and feedback. I like a fine tuned front end prompt.
And then actually go into the actual machine learning on the back end with the AI where it's pulling that data. And so like, I think it's going to get like more user friendly over the next couple of years, and especially like towards a lot of like internal usage, external usage. Um, we'll start seeing a lot of, a lot of probably new business pop up, come up with different solutions. So really eager and really curious to see how it plays out over the next couple of years. And David, as we start transitioning out of the podcast here, man, like where can people learn maybe a little bit more about you.
and little bit more about MediJoy that you'd recommend them check out.
Dave (23:48.233)
Yeah, so I think LinkedIn, find me on LinkedIn. I think it's literally linkedin.com slash Metajive is me. And I'm posting a lot of things, a lot of things about AI there. Metajive.com is the best place to get ahold of us. Use the mail form, but send me a message on LinkedIn. Let's talk. I'm also lifeasdave on basically every social media service. I'm pretty easy to find. So yeah, it's just, it's a really exciting time.
and
Dave (24:21.524)
It's always exciting in technology. So we made CD-ROMs, I went on the bottom of Slurpee's cups. Like that's obviously not gonna happen anymore. We started with Flash. And I remember in 2009, I said, oh, it's too late. We can't transition to build to be like a mobile app company. Like everyone who's done it is already doing that. I think I might've called that one a little early. I think there's plenty of room to continue building mobile apps after 2009. And...
Kyle James (24:46.126)
Yeah. Yeah.
Dave (24:50.9)
AI is just another transition, another way we're going to find a way to create value, help people out with technology, leverage design and tech. it's just another exciting time. It's like, been fortunate enough to work through two or three transitions.
Kyle James (25:08.433)
Yeah, it's going to be, uh, it's going be very interesting and thanks again, Dave. It's great having you on the show today. And thanks again for everybody who's listening in. Remember, please, if you're trying to implement AI into your own business, don't try and do it yourself. The time is just the, could cause it may not be worth it. Schedule a call with GPT trainer and let them build out your AI for you. Once again, that's GPT-trainer.com. Thanks again, Dave. That's great. Man, it's really good having you on the show. Okay. I appreciate you being on.
Dave (25:34.961)
Yeah, thanks so much for having me.
Kyle James (25:36.588)
Yeah, for sure. And sounding off for now, have a great rest of everybody and looking forward to seeing everyone on the next episode of AI Chronicles.
