Adam shares how Growth Assistant cut social post creation from four hours to one using a fine-tuned AI model, boosting impressions by 20%. He outlines five ways to get teams using AI: an "AI Guild" show-and-tell, focusing on real wins over speculative projects, removing fear through experimentation time, protecting learning hours instead of mandating budgets, and surfacing hidden tool features so teams don't have to hunt.
Chapters
How do you get AI integrated into a company in a way that's not distracting, that actually makes people more productive?
Let's get all of these people that are figuring it out into a room. Let's cross-pollinate ideas. It takes him 5 minutes to train a new model, and they cost $0.75 to train. 20% more organic social impressions because his content was better. There's too many tools. Just click that other button. In the same tool you're already using. If I get you excited about something, I don't ever need to work again 'cause you're gonna run with it.
So I run a company and I've noticed a couple of different things, Adam, when it comes to AI. Some people, if I show them anything AI, they roll their eyes and they don't wanna pay attention or they humor me. Other people are so down some kind of AI rabbit hole that they are these geeky gadget obsessives who are getting distracted by it. What I wanna understand is, How do you get AI integrated into a company in a way that's not distracting, that actually makes people more productive and gets them to enjoy the whole process? I wanna talk to you about how you do it. And you gave me a win. This guy, Achint, who works at one of your companies, Growth Assistant, what did he do and how did it impact his business? And then we'll go into the steps.
Yeah, let's start from the impact. So 20% more organic social impressions because his content was better. Why was his content better? Because we introduced one tool to him. That doesn't cost much money, that didn't take him more than a day to learn. And now instead of him writing his social post and it taking 4 hours, takes 1 hour. He goes and he interviews his team with all of his extra time now, and he gets so much more interesting ideas and concepts out of his team. And that's getting him his 20% higher results.
Okay. And the tool that he's using is what?
Uh, it's actually one that we've done a video on. It's fine tuning. He's taking these like basic chat models. It takes him 5 minutes to train a new model to say, okay, I want to build a new version of ChatGPT, but one that can talk in my voice that's like relevant to our brand. This is one of those things that I think sounds really nerdy, but actually takes a couple minutes to do. Just about anybody could do it. He's not a technical person. He's built a bunch of these and they cost 75 cents to train. So this is not an expensive thing either. It just slotted right into his existing workflow.
So where an ordinary chatbot would help him write, but make it more generic, He now gets to use a chat experience that is trained on the most successful posts that have worked for him. Okay.
That's right.
All right. Here's how we're gonna do it. We're gonna talk about the guild, which is a big way that you've introduced it. We're gonna talk about real wins. We're gonna talk about stimulating curiosity. We're gonna talk about the AI budget. And then finally, let's talk about curating tools. But first, the guild. What is the guild as you've introduced it at your company?
The guild is a, is a bottoms-up one. So we're, we're a venture studio, right? We own, uh, a couple different businesses. And I see a handful of people really getting a lot out of AI. I see the majority of people just not unchanged yet by AI. And then a couple of people that are staunchly against it. So the guild, I said, well, let's get all of these people that are figuring it out for themselves. Let's cross pollinate ideas because they're all finding tools. They have real problems that they're actually solving. And I want them to teach each other and then also me. How to make it work. If I can learn what made these people get over that hump, well, that's going to help get me the other 85% of people in these organizations to also get over the hump.
So it's like a show and tell where everyone is showing and explaining what's working for them. And you're the moderator. How do you moderate these things?
Yeah, I want to let people talk. I want to let people show what they liked. I want it to be a very real space. And so I don't, as the kind of nerd in the room, I don't want to make it a nerdy room. If you're demoing something, you say, hi, I tried this prompt and here's where it worked really well for me. I'm going to ask you questions because I want other people to see, oh yeah, this is a process. This is learning. And it's, oh, I tried too. And then I gave up, but this person just kept trying for 10 more minutes and then it worked for them. Okay. I could do that too. So it's, it's about these like genuine small wins. Again, bottoms up. We're talking little, little rocks that are going to build up.
Why bottoms up?
Yeah, I think bottoms up and top down can, can both work. If I can win a couple champions in each of these businesses, it's going to make any top down effort be so much more effective versus like Adam came in and said, dude, here's some new big initiative. Then he left and everyone's like, yeah, like we'll try the thing, but like, yeah, it's like a side project. I'd rather someone in the organization be won over like in their heart and say, no, This really can help you, but that's only gonna come if it's genuine. That's why I wanna learn and like really connect with this group of people and grow them so that they become the evangelists. All right.
So the Guild is a Zoom room essentially where everyone is doing show and tell. You're moderating, making sure that you give people room to talk, to explore, and you encourage them to share with each other. The next one is Real Wins. And this came up when you told me about someone who had a $60K proposal. What did, what was the proposal and what were they gonna do?
This one proposal though kind of missed the mark for me. It was $60,000 and it was a smattering of, of cost cutting. We hate cost cutting because the most you can accomplish with cost cutting is just removing our baseline of cost. And like, I want to dream multiple times bigger than that, right? A lot of what I see in AI proposals these days is new cool things that are net new to the business. We talk about AI enhanced versus AI first. AI enhanced, Sandra, is like, you're doing a thing, let me inject a little AI and it'll make that thing better for you. AI first is you could have never done this before GenAI was a thing. Problem is if the proposal is all dreamland, you're saying pay me $60,000 to get this, but no one knows if we could do it and we don't really know what it would cost thing. Like, I'm not going to recommend that any of our businesses sign that contract. Dreaming, like we don't have money to dream. We want to have money that makes money. That piece in the middle, which is AI enhanced, like we could make some steps, we can make some improvement, that is going to get the boulder moving in these organizations.
Okay. And so stay grounded in reality, but not so reality that you're petty and coming up with the cheap wins, the real wins for the business. All right. Next one is stimulating curiosity. What does stimulating curiosity mean for you when you're working with people on AI and integrating it?
In my experience, when people have the most fun and they get the most curious and they know everything about the business, suddenly their, their energy starts to flow into the business. If the business doesn't stimulate me intellectually, if I'm not curious, if I'm not excited, my energy is going to go into something, right? And everyone is this way. If you're curious about AI a little bit, and I could stimulate that and show you, oh, you thought it was so cool. But actually, you've only seen like the first half of a percent of what you could do even today. Take Achint again for an example. I taught him how to fine-tune models. And by the end of the week, he had made, I think, 6 or 7. And I had still only made 3 or 4. This curiosity thing, this scales. If I get you excited about something, I don't ever need to work again because you're going to run with it.
I see. And what you said before was also there are a lot of people who are thinking of AI being the killer of their jobs. Right. And what I need to do is get them to understand this is not a replacement, it's an enhancement. Get curious about it. And if you are curious about it and play with it, then you'll take it to a level that I'm not gonna be able to take you to.
This is part of why bottoms up to me makes sense. It's one-on-one, right? Or one-on-6, but I then meet with the people individually too. And I wanna find like what is stimulating for this one person and What's stopping someone from being curious? It's emotion, it's fear. So it's actually most of the job is peeling away some of those layers of just what's stopping you. And in some cases, maybe now is not the right time, right? If you're behind and I don't think AI can help you, well, the best thing I could do is actually step away. It's not say you need to try 2 AI tools this week.
Right.
Well, because that's not going to help them. That's not going to make them curious. Like, wait for the right time to show up. It's Let me help. Let me, let me do that. Oh, like now I could peel one more layer away and they can get a little more curious and excited all on their own.
Okay. Next, AI budget. This is Lauren, our chief of staff, had suggested that we give an AI budget to everyone and maybe we even encourage them to use the whole budget every month on something new. You didn't seem as enthused when I told you that idea. Why?
You know, I don't want to be labeled like the anti-top-down guy. But a budget to me is, is a mandate, right? If I say everyone can spend $60 a month on AI tools, everyone's going to think, well, I'm supposed to spend this, like I'm supposed to do this. And what I would say is like, okay, well, so you're adding to that person's job, right? Which is go and learn these tools and find whatever the heck tools. If it's just money, you're like also telling them to identify the right tools that could be useful for them and play with them and learn them.
Right.
So you basically added, uh, what, how many hours a week to, to their plate? A couple?
That's fair. There's a big—
Depending on how curious they are, it might even be a lot. It might be, if they get real distracted by it, it might be 10 hours a week. Are you gonna also relax your, your targets for them?
Right.
Say, well, hey, I wanted you to find 10 outleads this week, but because of this new AI initiative, you only need to find 8 because I want you to spend 20% of your time on this other thing. No. And it can put people in a hard spot, which is The most curious people that are the most excited are going to get the most behind and then the most stressed out and you can kind of kill their spirit in that way.
So what you do instead—
Thoughtfully.
Yeah.
Done thoughtfully, it might work. You create space for it. The AI Guild is a great example of this. It's literally an hour on the calendar where we're just going to talk and get curious together and you're paid to do it. Like, this is part of your job right now. Like, we're going to do this thing and it's a little bit more like, what do you want? What do you need? I'm, I'm not giving them anything they're not asking for. I'm not asking them to do anything that they're not asking me to keep them accountable to. So it's more about discovering like, what are the very unique things that people are interested and excited about and then enabling that. And that might mean like carving out a part of their job for it. That's, oh, like you actually shouldn't have this responsibility right now. We're so excited about AI as a business. I want you to spend 20% of your time. On like learning how to apply AI. So we're going to take this one big task and we're going to reassign it to someone else. Like that, man, I'm, I'm all in on that.
All right. I see that. Let's go to the last item, which is curating tools for them. So instead of saying, go look online on Product Hunt, on the App Store, go watch YouTube videos, which to me is exciting, you're saying curate.
Yeah, there's, there's too many tools and there's a lot of overlap between the tools. Right now, I, I'm one of the people that signs up for trial after trial after trial. The problem is like, there's too many tools back to this, like resourcing. If you put that on your team, well, you actually have to like allocate time and energy for them to do that. What I think is more useful is, look, I'm already doing this. That's why we've created this AI Guild space where people can come to me. They can show me the tool they're using because I've tried 60 different things. Like, oh. Well, actually, I think this one, all the, all the stuff you like about this and the stuff you're frustrated with, that's actually just one other feature. Can I give you a really quick example of that? Yeah. So everyone uses ChatGPT, right? The guild meeting last week, one of the people was presenting, hey, here's how I use, uh, projects in ChatGPT. It's easy, easy to use, upload documents and then you can chat with it. But it like loses the why of the documents and it feels like I'm just talking to regular ChatGPT. And another person said, well, Have you ever tried the custom GPTs? She said, no, I've never clicked that button before. Like, what does that do? Well, it turns out custom GPTs, it's a lot like projects, but you can also tune the model. You actually train it a little bit. There's no training in projects. And so we walked away from that conversation being like, don't buy a new tool. Don't spend any more money. Just click that other button in the same tool you're already using and all the frustrations you have about not being able to train the projects folder, how to look at your documents. Oh, it's already right there. And you know what? The person who has done the custom GPT, she had never used a project before. People are so close. They just need someone that's kind of done it before. I, there might even be a world here, Andrew, where like AI guilds become a thing. I just want to like join this thing and be guided gently on my own journey. I think almost any business could do this, create space for a guild. Where you just find the person that's the most advanced, that's the most excited. You cut out part of their job for it. You give them a little space and say, I want you to run a guild every Friday from 1 to 2 PM, get everybody in the room, moderate a conversation. Any business could do this this week. There's nothing stopping you. And I can almost guarantee that it will level up everyone in your org. At the very least, they'll share tools among each other.
Yeah.
And at the most, they'll all get really, really creative and really excited and they'll start to find new new, more exciting ways to plug AI into the business and, you know, grow the top line, not just reduce the cost.
So interesting that you use projects because I am someone who does have custom bots on ChatGPT, but until like Wednesday of this week, I did not touch the projects feature and I didn't realize that it solved the problem that I had before. I think it was just like, eh, it's for other people. All right, so this has been incredibly helpful. We've got 5 different ways for people to use and teach their people about AI and get them incorporated into this new world and to teach each other. Thanks.
Yeah, thank you.
