Jack of All Trades, Master of None

Okay, yeah, so Jack of all trades, master of none. That's what I'm going to talk a little bit about today. And I'm really happy I go after Madsen because I think it's kind of a perfect follow up. I'm not going to be talking about the biggest high end productions. I'm talking about sitting in-house as a lot of us do and moving fast and being this skill set, at least the Jack of all trades, somehow kind of turned into a creative superpower for the in-house team and for the in-house creatives. But also with this being a video summit, I want to start with a video. I'm not going to start maybe with as a cool video, but a video that can show you a little bit what can be done. So, okay. Why am I starting? Why am I showing this? So first of all, that's Vivi. She's four. She's my daughter and she's pretty obsessed with Frozen. It's not super healthy, I think. And why am I showing this? We just saw some much cooler examples, right? But this was something I could do two hours in my living room while I was babysitting Vivi and keeping her entertained. I also did it for free. I had no paid tools because I'm not really a creative anymore. I'm an operator, so I didn't even have an Adobe license. So I used a phone. I used the free version of DaVinci Resolve. I used an open source AI model running in Comfy UI, which is also free. And then it couldn't render on my laptop, so I spent $2 in cloud compute to actually render it. I think that's where it starts becoming interesting for in-house teams. And I couldn't even do Comfy UI or DaVinci Resolve before I started, but then I used a bit of cloud or whatever AI model you want and some YouTube and you kind of have it. So what I'm trying to say with this is imagine then what your in-house team can do if I could get at least to that. And that's without those budgets. Maybe it's 25. So before I get too far into that, just a little bit of who I am. Who am I, besides Vivi's nerdy and very tired dad? My name is Mads. For the past little over six years, I worked at Too Good To Go. And then for those of you who don't know, Too Good To Go is the number one surplus food marketplace. It's an app. It's got over 120 million users. And in over 21 countries. very, very fast the entire time I've been there, maybe a little bit too fast sometimes to keep up. I started just being the video kind of guy. Then I became the head of the video team and later the head of creative production on a team that scaled up to 25 people. I joined as the third person on that team with Kyle Meal and I created other creative. And then we just ran fast and we made it into a creative machine that could serve 21 markets. And I just had a pretty interesting fact from Samuel, who is the lead video producer at So Good To Go Now, that one of the campaigns we ran a few months ago, we exported 909 videos for that campaign. And that was just the videos. We're also running statics and other stuff and we're always doing the always on content, which I think it's pretty insane you can get to that scale. So I'll get a little bit more into that later on. And then in February, I actually took a change. I changed to the company called Family, which is the number one, nursery app in the UK, growing in Germany and the US. Smaller company, smaller scale, B2B now, but very much back to building and scaling on a small team, which is what I found super, super fun. And this time I'm hoping I can do it a little bit smarter with all the stuff we learned at So Good To Go and also a lot cooler with all the new tools that's available. And Family is really a company that's embracing this jack of all trades skill set and mindset and also AI. The founders are still there. The CEO is coding. The things I've been allowed to do for the four months I've been there. When the chief data engineer just takes you to the site to teach you how to use GitHub properly and the terminal, even though I'm doing some other stuff. It's a crazy new world. So that's also when I try to explain who am I. I'm a video, VFX, motion, design, wannabe developer, AI, strategy, marketing, operations kind of guy. So doing a lot of things. And I think sometimes I've had talks with Kyle Emil, who was my manager. He was like, what are you working on right now? How is that your scope? And then sometimes I figured out, okay, maybe that actually does make sense. And that can be my scope. But I've never really been the best at anything, but good at a lot of things. So I just really want to explain that journey to there and how that skill set kind of came. And I always try to force it into Star Wars just because I like Star Wars. And I think the whole film passion came from Star Wars. Watching the VHS tapes with my dad, going to the cinema when they were re-released. I just wanted to work on Star Wars, specifically doing light, saber effects, which I think a lot of VFX people started with. But that's not always the easiest thing. So I wanted to do Star Wars, pretty quickly gave up. 2004, done with high school. Video wasn't really super accessible, specifically not in Denmark either. DSLRs were just becoming a thing. So I did the sensible thing that everyone tells you to go to CBS. So I went to business school, still playing a little bit with After Effects on the side, trying to learn how to do light, saber effects with the free YouTube tutorials from video co-pilots. A lot of people probably know. Then when CBS was done, I was like, okay, that's not what I'm going to do. And this was 2009 and stuff got started getting interesting. The GoPros were pretty good. The iPhone 3GS came out. Everyone could do video, kind of. I could get a cracked version of After Effects pretty easily. Your gaming laptop could run it. So I got back into the creative thing, went to the Danish media and journalism school, chose the most jack of all trades education they had, interactive design, worked a little bit of everything, and then tried to do the VFX again. Came a junior VFX, VFX artist, and basically just did a lot of rotoscoping, which was super boring and couldn't stay focused. Got into motion design. Then it became video. Then I tried to do a video game. Got kind of confused, but video was my main focus and always felt like I wasn't good enough. This was about the time that I joined Too Good To Go and got a pretty good realization here. I'm not a master, not even close. But it always kind of seemed like a bad thing, not being the best at one thing. I wouldn't be winning a can lion. I probably wouldn't even be making a Super Bowl ad. So before I get to the part where I'm going to say that's a good thing, I actually want us to show two examples of what the masters can make because it's definitely still used for that. But in the in-house teams, I do think there's a lot of use for the jack of all trades as well. So the first example is really, really cool of what the masters can do without AI. It's an old example to try to branch out a little bit and not show AI. Still think we can do a lot with that. And the other example is unfortunately by Kanye. I'll still show it because it's a really good example. And I think there was a lot of talk about Kanye at Too Good To Go. So to honor Kylie Meal, I think I want to keep Kanye in there. So the first one is from 2018. Just going to show a short snippet for it. But Spike Jonze made an ad for Apple, which is pretty expensive. We can dance in the hallway. Only one more night in Los Angeles. I really thought that I can handle it. But the funny thing is I was holding back tears. I didn't think this day would happen. I gave all this up for a chance. You would have thought I'd be the man for this. But the funny thing is we can never say it. I didn't think this day would happen. I'm about it till it's over. I'm about it till it's over. So that's pretty cool. That's done by the masters doing what the masters can do. That was all practical. Basically no VFX, crazy coordination choreography done with hydraulics. Probably cost a few million dollars as well. Had a famous singer making the song, another dancer, Spike Jonze. It went super viral. It got one awards at Cannes, awards multiple places. It even got an 8.5 on IMDb, which I don't know how happens for an ad. That's a thing an in-house team probably never is going to do. It's also another example. Here's the one with Kanye. Let us just see it. Hey y'all, this is Jay and this is my commercial. And since we spent all the money on the commercial spot, we actually, he is spending the money on the actual commercial. But the idea is I want you to go to Yeezy.com. Y-E-E-Z-Y.com. And I'm going to write it at the bottom of the screen. And I got some shoes. And that's it. So yeah, so that's Kanye. It probably cost about the same as the Apple commercial. He just spent all the money on the Super Bowl placement and just did that thing. Right. I think those are two kind of good extremes where probably the in-house teams don't play. They probably both had a good return on investment. Like Kanye says that this generated 20 million in sales and almost 300 orders of shoes in 24 hours, says Kanye. But it probably did pretty well. They also did something interesting that I think a lot of us in-house teams forget to think about production spend versus media spend. You can go full in all on production spend and hope you go viral. You can go full on on media spend like Kanye and just trust that. Or where most of us play, probably somewhere in between. And a lot of us, I think, tend to forget that we spent all on production. Forgot any media spend. We don't get it anywhere. So now I'm going to get to the part where I'm going to try to say not being a master is not really a problem. At least if you're sitting in-house in most companies. Because this is where I think we need jack of all trades instead. Because most of us don't need a Ken Lyon. We're nice to have though. Or a Super Bowl placement. Probably can't afford it. 99% of companies, they need something else. And they need something else on a daily basis. Because I guess you all know this one. We usually have no time, no budgets, and you need it in six languages. So we're not going to go out and do a custom choreography with hydraulics or even get Kanye, right? And we're also required to have a high return on investment. And even when we spend zero money, they tell us you can't spend too much time. So what are we doing in-house? We're doing a ton of different things on a daily basis in this age of content. And AI is probably making this even crazier, right? We're doing SoMe, web, user-generated content, out-of-home, print, point-of-sale, event blog, email, podcast. We're doing everything, right? And it's with small teams. That also means one person needs to be able to do a lot. At Too Good To Go, it looked something like this. Lots of user-generated content. It performed super well there with content creators, some done in-house. I think we were doing about 50 to 60 of these a month. Then a bunch of global campaigns, smaller local campaigns, lots of out-of-home, lots of SoMe content. Lots of partner content. Too Good To Go has a lot of partners and they all want some content. So it's a lot of logos, a lot of translations. Lots of lead gen, lots of paid ads. Hundreds of thousands and thousands of assets a month. And at Family, it's a little bit similar. Even a smaller company, there's still a big need for a big output. There's a biweekly podcast, newsletters, emails, blog posts, YouTube channel, lots of testimonials, event and print material, lots of paid ads, right? So I think this is what jack-of-all-trades on in-house creative teams really, really enable. They can do way more than the specialist who is the best sound designer or the best color grader. I got an interesting stat. It's from 2025, so I don't know. The world might have changed completely in six months. But I think this still holds pretty true for a lot of us. Most now have some kind of in-house video capability, but a lot of people maybe feel like they can't get to that quality where they want, that they didn't have to go out of house, get an agency, et cetera. But especially with video production, right? It's complicated. It's time consuming. It requires gear. It requires all these things. But I don't really believe that's true anymore, and especially with this like jack-of-all-trades skill set for the people on your team. And it's also kind of a mindset for me. I feel like it actually doesn't replace ambition to be a jack-of-all-trades anymore. I think it enables it, at least when you're sitting in-house, because you can all of a sudden do more things. You can free up time and money for the big bets, right? This is really where the jack-of-all-trades shine. So it comes with a negative thing when we're saying jack-of-all-trades master of none. But let me kind of explain what I think that this new jack-of-all-trades that's not a negative thing is. And I think you're already probably seeing it on a lot of your in-house teams that people are branching this way. Even as Martin was before me was telling, right? Hey, if you knew this, this, and this, AI was smarter, right? And I think we need broader and broader skill sets to be able to adapt to the new world of all these things coming. But I think first and foremost, in the creative teams, we need to be creative, right? We need to have a good sense of aesthetics, and they need to have that core skill down. Of course, that's the start. But that's the start. They need to be curious, tech-savvy, and strategic so they can learn the new things. They can try out the new things. They can explore with whatever it was Martin just saw, whatever I did with Vivi, and that was my excuse to learn a new thing. I think that's just getting more and more important, and I think it's something we really need to enable for our in-house teams to be able to kind of win in this new world and be able to do super, super, super cool things. Also, one of the things I've really realized in my own journey as well, coming from just a tech guy who knew video things, is you also want your creative, being a video producer, whoever it is, to kind of understand paid media, channel behaviors, budgets, know enough about copy and design to sometimes handle some of that that's lower-hanging fruit. They stay current. They're getting the new plug-ins, the new software. They're trying new things. They also understand that a TikTok is not the same as a YouTube video. They also can jump into Figma. They can do all these things, and they understand that the right creative needs to also have the right placement and the right budget to have the right performance for you. But then also before I tell you to be like, okay, just ensure that everyone on your creative team can do everything and they can run right on the in every single direction, especially seeing what Martin just showed with AI. We can really start doing a lot of things. A lot of in-house teams, at least in my experience, sometimes don't really have the foundation in place to start running crazy, or at least you start running crazy and you don't save it or you forget it and you can't really use it. So I want to get a little bit into a lot of the stuff we did at Too Good To Go, but also at Family about how important foundational work is. A little bit boring, so I'll try to get through it quickly. But just more than ever, we need this foundation. Just very, very quickly, some things we somehow forget a little bit is just three things, especially for video production. You need to have your storage in order and it needs to be properly structured. Most of us already have Google Drive or OneDrive or whatever. It can be free. Just use that. Just be good about it. You probably don't need an in-house server anymore like in the good old days. It's expensive. It's difficult. We use just Google Drive at Family. We use Lucid Link, which is a little bit more pro for video production. You can onboard a freelancer in Spain. They can be working in your workflow on the same day and deliver on the same place. Super important. Project management. Again, if we have 10 people doing AI stuff and running even faster, super, super important. Get some kind of software. We had Asana Enterprise at Too Good To Go. It's Notion at Family. But stick to it. Be really strict. We built a full kind of briefing system at Too Good To Go where everything goes into an intake form and comes into a project management tool. Because again, as you scale and you run faster, you can't do it without good project management. And then just fix your briefing process. We all have that problem. We do shitty briefs. We waste way more time. Especially now if we have a bad brief and we run and do 20 AI created things. So just get that under control. And then of course for video production, we're going to need some equipment. But well, since back when I wanted to get into video, it was really expensive or really shitty. Now the equipment is super cheap and powerful. We just saw another example of a Hollywood production using Gen AI. We're also seeing examples of Hollywood productions really embracing this jack of trades mindset and skillset as well. The Creator. They used a $80 million Hollywood sci-fi film by Gary Edwards. They used an FX3. It's the same video camera we have in the Too Good To Go studio. It's a 3000 Euro camera or something. They didn't choose it to save money. They choose it because it was versatile. It was good enough to actually get the quality they needed. And instead of doing a big set, they could now run off into a jungle and film with no time with a small camera. I think that's pretty crazy. And a lot of it was even filmed on a DJI Mavic drone. I encourage you to see it. It's crazy effects and it looks amazing. And then secondly, 28 years later, Zombie Horror Movie. They chose to shoot it predominantly on an iPhone. Another kind of crazy effect. Of course, they still have an $80 million around it. Weird lenses, lots of lights, et cetera. But the fact that they can use that and they could do these arrays with like 28 cameras around. Imagine doing that 28 array with a Sony Venice, right? It's expensive. You probably can't log it into a forest. You could do it with the iPhones. And if anyone tells me you can't do video production in-house, you probably already have a free iPhone. You can just get started and it'll be good enough. And if the sound sucks, you use Adobe podcast and it sounds great. What I'm just trying to say is if you need a super, super high end kind of in-house setup, you can get two FX3 and FX30 as a backup camera for less than 10,000, right? You can go half that price if you're okay. You can just start with 800 euros for a phone and you're good to go, right? There's no limit here for the jack of all trades, but also you need someone who can then also operate all this equipment, which you probably don't have 10 people. And again, that's where I really feel like the jack of all trades is. I feel like the jack of all trades skill set is important. And another thing when we're talking in-house teams, we're also like, oh, we have to go out and film. We don't have a studio. Well, you can get a studio. You have an office, right? That can be a studio. At Too Good To Go, I think one of the ways we got to get a studio was we need more meeting rooms. Okay. Can we have a hybrid studio, right? We can have a meeting room slash studio. We can roll everything away. And then Samuel and Roberta, they can roll everything out and they can shoot stop motion. Like having this studio and without it even costing anything. Just made it so efficient for these jack of all mind trades. Sorry, putting you on camera, Roberta. I should have warned you. But it just means we could run in, we could do headshot stop motion. We could do cooking videos in here. We had a portable cooking setup, even though the office smelled and people complained. But the thing that we could produce with just one or two people was insane with this studio and it cost us nothing. And also Samuel and Roberta here, two people I really also feel have embraced this jack of all trades mindset. Never done stop motion before. Figure out drag and frame is the software. Learn it, go in. They build their own custom potato spinning rig with some duct tape. And then all of a sudden we had great output, right? And both of them also keep expanding their skill sets. Roberta came from classic graphical design, took on motion design, then jump into video, now stop motion. Now I heard you're programming AI's after effects scripts by yourself. Samuel, the same thing. Samuel even built an intake tool for this whole Asana thing. When you have a thousand people briefing, you don't want all those briefs coming to the creative team. So you build an AI tool that scans all the Canva templates and says, oh, we already have a template for this. We're not going to work on it. You just use this. That's crazy that our marketing intern seven years ago is now doing that while he's doing crazy video production. I think that's a jack of all trades. That's the kind of people you need in house. At Family as well, they also had a studio before I joined. And I don't even want to call that a studio. They have an old server cupboard, which is like two square meters tops. It's horrible. It's hot in there, but they made it work. They're running a biweekly high quality podcast. As long as Mikkel sits in the window frame, then he'll fit in there. But again, I feel like that's the jack of all trades mindset. You can run a high quality podcast with two cheap microphones, the camera you use in a cupboard. There's really kind of no thing that stops you. And Mikkel as well is one of these people who also embraced this jack of all trades mindset. Coming from a video background, now getting into podcasting, learning how to optimize for CEO and organic growth. He's doing a marketing account. He's doing a marketing academy with Matt, the brand director to teach him that. He's fringing into paid. I think this is just, we get so much more out of our people in house when they can do a little bit of everything. So, okay, we have people who can do a little bit of everything, but we also need to free up some time for them. The foundation was one of those things and right tools, but also we need to automate and scale things. Because if they're just running around doing copy paste work, they're not going to be learning the new AI tools. They're not going to be exploring the new software. So automate and scale. We built a lot of this, both at family and at Too Good To Go. Some with AI, some the good old fashioned way. But it also doesn't have to be complicated, right? At Too Good To Go, it's a enterprise setup where the creative team builds lots of templates. Then a thousand sales and whatever people can just go crazy. And then luckily there's an approval step. So they don't go too crazy. That already took so much work off the creative team. And then all of a sudden they could start doing and learning other things and expanding that jack of all trades mindset. That's a really hard thing to say. At the family, we're using Figma Bus. It could just be, it's even the thing, right? You all have a PowerPoint template, et cetera. Or now you can make a home build solution. But get this scale done to free up this time for your team to catch up and be able to learn the new things. A few examples. One thing, this whole 909 assets exported from Too Good To Go of these videos. This was all done with our eight year old After Effects script that Samuel found. Basically, it can just take a CSV file and make it into a file. And make all these videos for us. So you just get the partner marketing people to fill in all the content they want in their logos and it just runs. Now it's even stronger because we can have AI fill it in or translate it. And we can even adapt the script or write these scripts ourselves. Which I just think is so crazy we can do this. And they're like, this script was $50, right? And then it unlocked this thing for the company. Another thing at Too Good To Go being in 21 markets, we had a lot of translation stuff to do. All of a sudden the tools caught up, right? We could do the dubbing like we all see in Germany. Or we could embrace some of the new tools. So I just want to show a super short example of a video filmed with a cheap freelance and one of our marketing people in Oslo. And then sent back to Copenhagen to be edited and then run in 10 different languages. When you've been studying for three years and invested a lot in your own company, you don't have that much money. So there's one sensible thing to do, and that is to buy yourself a box. Yes, that was one of two sensible things. Yes. To save money, we both live and work on the boat. So that was just three of the languages that these guys were speaking, right? Back then, some of the automated tools weren't quite good enough. You've probably all heard of Hagen, 11 Labs, there's some way to lip. But then we just worked with our teams. These guys had signed off so we could do dubbing and AI. They were super into it. So we could just ensure we had really, really good scripts from 11 Labs and we can add them in here and we can do lip syncing. Now, if you want to do a freelance, a film in Norway could run in so many of our countries and perform so much better because it's actually speaking the right language and even had lips animated for it. We took kind of the next step be like, can we scale this translation thing? So we're dragging these files into Hagen and then we get that and we drag it over to Submagic, another tool to get animated subtitles. These tools both had API access, which means you can basically just have another program do it. So we hired a freelancer to Good2Go to build this front end. We dump in one video and you get 10 out with their lip synced and dubbed and automatically animated subs. We did it with a freelancer because I was too scared to try. And then when I joined family, I figured I had a boring evening. They don't need it. Let me see if I can build it again because I don't know how to do it. And with Cloud Code, I could build that in an evening and those two APIs. And then you have a tool. It's not quite perfect yet. It makes mistakes sometimes. But the fact that you can dump in one video, get 20 out, dubbed, lip synced and animated subtitles, that's crazy. And that's a scale we take away from the creative team. Another thing, as I said, I ran a little bit crazy with my first four months at Family. At Good2Go, being so many employees, we have so much picture and video content, right? You need people to get it somewhere. Google Drive kind of sucks. OneDrive probably sucks as well. Never tried it. But we bought a damn tool, a digital asset management tool. It worked perfectly. AI takes our pictures. People can filter. They can find it. Everyone can download it. It's pretty expensive. At Family, we're small. Doesn't make sense to invest in that. And I figured, hey, I'll give it a try. Can I build that? Then, of course, you can. On top of Google Drive with our own kind of internet, I could build it. I could build the front end in a day or two. But then you're like, oh, you don't have the cool AI features that doesn't tag the thing. That's what you pay all the money for. Talk with Claude. Can you do it for me? No, it's way too many tokens. It's too expensive. Talk a little bit more with Claude. What about all these free open source AI models? Could I run that on my little MacBook Pro? And it's like, oh, that's true. You probably could. It builds a Python script for me. I downloaded a 15 gigabyte AI model. I ran 5,000 pictures through it, and my MacBook almost died, but it did it. And then we had it, and it works, and it's in a database. Next step I want to do is basically have that AI model run on a Mac Mini somewhere, and it's an automated flow, so I have to do it. But the fact that that's possible today for someone with a jack of all trades skill set to just be like, you know what? I'm going to try to build a 10,000 euro dam that people pay for every year. And it worked, and we're pretty happy with it so far. So I think what I'm trying to say is if you are jack of all trades in these in-house teams, the possibilities are more or less endless. Because also, software is cheap now, or free, or you build it for free. You build it yourself, right? I think we have DaVinci, Adobe, Figma, GitHub, Cloud, AI tools, Homemade, whatever. The most expensive thing here is an Adobe license. All the Adobe programs, $55 a month, right? I don't think that's really a blocker for any of us. DaVinci is free most of the time. That's just a tool I use to do the Vivi thing. And on top of that, open source is just blowing up. The community-driven things of creative work is insane. The stuff that people are pumping out and giving out for free. And it showed us a few things from Blender, right? Blender is basically like a full, almost 3D, NVFX suite. It's free. It's open source. There's a million plugins for it. There's things like terminal command stuff that you can run that can batch convert video, right? Basically, FFmpeg is the thing that DaVinci Resolve has almost built the top up, and you can access that for free in your terminal. After it's scripted, like the spreadsheet we talked about, there's a thousand of those. People are building them themselves now. Tons of things on GitHub. There's so many free AI models to run. And then there's this thing called Comfy UI, which I think some of the stuff that Martin showed with this motion control, they're most likely using this free open source node-based builder to do that. So it's getting a little bit nerdy. You can obviously see that's what I like. So I'll try to go through it kind of quickly. But basically, it's a free node-based builder, open source. So you run it and you download all these AI models to your own computer. So you're not paying for tokens. You're not paying for anything. And then you run this workflow through it. And there's a bunch of stuff. There's a bunch of templates. It's the thing I used for the little clip I did with Vivi. Then I also didn't have to put my daughter on a cloud-based something something AI model, but I could actually keep it local on my own computer. This workflow just made all these different passes, a skeleton pose for motion tracking like Martin showed, the masks, et cetera. And then I did that clip. I also started working on a clip where Vivi's skeleton was animating an Ewok. But I had to stop because I also had to finish for this talk and do some real work. But it's just crazy what you can do. And I'm not even going to go into this list. But just for this talk, these were the AI models I found and learned about. And these are all free and open source out there. They can run most of these things that you can do. Of course, some of it is not as good as the thing as CDance and other things you buy. But a lot of them actually are. And a lot of it runs locally. The model that I used to tag them is the thing called clip at the end. Just free. You could just download it through the terminal. Claude told me how. Didn't really know what I was doing. But all of a sudden, it was working. And the community thing is just crazy as well. There's a YouTube channel I like to follow as a guy who likes VFX called Corridor Crew. They basically sold green screen by themselves. I just want to show a short clip of that video. Jordan, you gave me the hardest shot to ever key in my career. There's no way I'm going to regret this. I'm going to say there's literally no way you keyed this. You're lying. No way. Fine. Whoa. Lies. I'm going to do the double. Whoa. Through the glass. Whoa. And that is working. Wow. Dang. Sounds like everyone's just kind of in shock right now. So that's pretty crazy, right? Like, I don't know how many work with green screen. There are lots of good tools in After Effects and Premiere. But it's really hard to do this kind of stuff. If you're good at Nuke and you pay 40,000 kroners or whatever this Nuke costs and you're really good at that, you could probably do it. But these guys, they built it basically open source. They trained their own neural network. And they made fake green screen clips with 3D so it had all the perfect information. And they just trained it. And what did they do? They gave it away for free. And it kind of sucked when they gave it away. It required like a super expensive GPU to run. It was running it through GitHub. I gave up. And then two weeks later, the community optimized it so it could run on a shitty GPU. And then two weeks later, there's a free AI script so you can plug it into Fusion or After Effects. And now it just works. And I feel like that, again, it's just so crazy, right? That's what you can do in the in-house team. The Jaggerfall trades that's willing to try this and get that script and then do it. And then when I had to render this thing with Vivi, it was getting late and my computer was having a little bit of a hard time. But now cloud compute is cheap and accessible, right? Like we all know that everyone is trying to buy GPUs and all the data centers are going crazy. The good thing about that is that then they rent them out cheap. You can get an H200 system with all this whatever stuff. It costs about $60,000 to build, right? You probably can't even build it because the GPUs are so expensive. It's sold out. You can rent it for $3.5 an hour. I just chose a high-end gaming system for the Vivi clip. Instead of trying to do like eight hours of rendering, it did like 10 minutes and it cost me $2. Like the fact that that's also available, RunPod that I tried to use has a ton of templates that you can just work with or just as a terminal to run things through. So again, like you can't even say the fact that you don't have a server farm in your office can stop you. If you have a Jaggerfall trade, they'll probably figure this out as well. And then the next thing is cloud or whatever AI thing you use. The cloud rabbit hole that really removes the roadblocks. Like YouTube used to do and still do, there's a tutorial for everything. Cloud is basically that on steroids. This Vivi workflow didn't really work for me. And then I was trying to talk with cloud and send screenshots. And then cloud told me, Oh, but the workflows in Compu UI is a JSON file. Just give me the JSON file and just fix the JSON file and did something. And I dragged it back in the program and it worked. Like, again, that's crazy. And that just means like more or less anyone who's willing to try on these in-house teams, you can also sort that out. And then I think it's a little bit about MCPs and stuff as well. I'm not even going to talk about what that is. Basically, your AI tools can talk with other tools. They can directly work in it. So even for project management, I use cloud to like fix everything for me in Notion. If there's something you don't understand, don't bother building. There's MCP to Blender. Cloud can build a 3D model in Blender. So I think just don't say you can't do it. There's a pretty good chance you can get pretty far. And then there's also where I want to get that small warning out there. Okay. Why it's important that we have the foundation in place. Also that we have some managers who can keep us running in the right direction. I even got blind a little bit joining family and getting allowed to build all these different things. And while I was building stuff we didn't really need, the team was leading me to do some of the job I was actually hired for, some project management. Luckily, they told me and got me back on track. We sorted that, got that streamlined, and then I could go back to playing with some of these things. But the fact is that every weird idea I get, I can start doing, right? And then the fact I have a 3D printer. I can make 3D models with an AI and I can do that. And we can make a prop for a film or whatever. It's super cool, but it's a problem. Like you need to keep your team a little bit in check here as well. Because they can run wild and they can build stuff you don't need. And they can just waste a lot of time and money. Okay. So basically what I'm trying to say here is that having these jack of all trades on your in-house creative team, that's basically what unlocks the superpower for small in-house creative teams. As long as we have this strong foundation in place, then you can do a lot. What I want to say is for those in-house teams and you run them, ensure your team can keep exploring, keep learning, keep out of the silos. And by that, I mean most of the stuff I also learned about, not the nerdy technical stuff, that's YouTube. But anything about how to do marketing, run a business, it's by talking with the other people in the team, right? You go sit with the revenue team. They'll teach you a little bit more about the KPIs and all those numbers. They create a lot of information. You go talk with the data engineer and he helps you be like, oh, shit, you're building cloud stuff in a really shitty way. I'll teach you the right way. And he uses 10 words he didn't understand. So I asked Cloud to explain them and then I went back to him and I sounded like I knew what I was doing. So I think that's what I mean about keeping out of the silo. Ensure your team just goes out there and talks with the wider team and keeps learning, hopefully because you freed up time for them by building all these smart things or have them build all these smart things. So that's where I think I would rather call it for the in-house creative teams. The You can't even be like, we're a startup, we have no money. Just get a phone and open source stuff. You're fine, right? Or you can go even bigger. But honestly, I couldn't even see a reason to spend more than $10,000, $15,000 on getting everything you need for a high-end video production. Because also, this thing, being this jack-of-all-trades, using all these tools, being able to do all of this, it also makes what used to be impossible, possible, right? But it also makes it then possible for your company to be like, well, we saved so much. We're getting a much better return on investment. Now we actually have some budget and time left off to go out and hire those big agencies and those big directors and do these big splashes, right? Or even just give your team more chance to really take a big bet. There was this slide. I skipped it. So for me, that's basically it. I just want to say, I think the jack-of-all-trades skill set and mindset, it's already going that way. I think we're talking about the generalist a little bit. But really, really, really embrace it. And ensure that your team goes that way. You don't have to hire new people. I think most creatives are ready and already taking that step. It's embracing and giving them the chance to do it. If you are a creative, just don't be afraid to try. Explore, figure out the tools, talk with some stupid AI, and it'll tell you what the tools are. And then just learn it and go, and you'll get super, super far.