Most Companies Suck At AI. Here's The 3-Step Fix

**Eric Siu** (0:00)
All right, so let's face it, most people suck at AI, which means most companies suck at AI. So if you're watching this right now, the question is, how do you make your team better at AI? I'm going to show you exactly how you can make everyone good at AI. Let's get into it. The first step is you want to make sure that you have some type of repository where your team can access your skills. Ramp has a, they famously have like a dojo, their product is called Glass, and you can basically go there and you can access certain skills. And so the way we think about this is that you can see, like, you have people that are maybe on the sales side of things, and they can click on this and see, oh, here's a lead dossier pipeline, here's a lead dossier builder, here's a cold outbound optimizer. And then you can see that there's upvotes, things like that. And so, you know, Ramp came up with that, and then my friend Sayada, he showed me that his skills has like upvotes and things like that. So I thought that content is really important because we have people at my team right now, you think of skills, like a skills.md file, that is the ability for you to take a lot of the processes that you have and then hand them off to someone else, and somebody else just injects it. They get infused with the skills that you have. Also, they can fork these skills into a version that they like to edit it to their liking. And then maybe that version becomes really interesting for other people to use, and that can spread throughout the organization, right? And also you want to think of this having like a skills repository, almost like an app store, like an Apple app store. They don't necessarily delete any apps. It's more so the best apps will rise in top. And so the more votes, the more usage a skill gets, the more it's going to rise in the top. And the important thing here to consider is that if you have a skills repo, you want to make sure that in your dojo, for example, when people are onboarding to a company, let's say they're onboarding for sales. Okay, let's give them these skills over here. If they're onboarding to account management, okay, let's give them the skills. If they're onboarding for paid media, let's give them the paid media landing page turned on the advance and build an AI conversion ops, right? So there's a lot of things that you can pull in here, and then people can add in whatever skills that they want to add over here. And then there's metrics that you can see as well. So the whole idea here is that you are trying to get everyone to align in the same direction. You want all the vectors aligning in the same direction. That way your company is going to be able to move a lot faster. Because the problem with a lot of companies is that you start to add more and more people. What ends up happening is that I started going this way, you start going this way. We're fundamentally misaligned and we end up going nowhere. If we're moving in the right direction, we are going to be making progress. And here's the other thing. You want to see participation by team. You want to see individual leaderboards. You want to see number of completions. You want to see skilled contributors. You want to see the funnel as well. You want to see teams by runs. So these leaderboards are cool and you can decide how you want to gamify this within your organization. Number two is you want to think about this concept of having a pod of one or more so an army of one. Think of it this way. A pod used to be maybe one developer, one designer, and one product manager. Well, now a pod can actually be one. So if you're a really good engineer and you happen to have some skills in design, some skills in product because you have the power of these AI agents now, you can actually do a lot of the stuff. That's no different than maybe at my company, Single Grain, maybe someone doing AEO can do a lot of the work condensed into one now. So we are now thinking of this as a pod of one apprenticeship. By the way, you can see me making a lot of these artifacts where it's simple for me to use a design.md to make whatever artifacts I want to explain things easier. Like how I'm doing to you right now. So basically we're running a six week selective apprenticeship, and then you can see we have different cohorts, there's different content and things like that. Maybe the way to think about this is this, right? So we're basically piloting a six week pod of one apprenticeship where we want everyone, we want these people that are selected to be the first that have test batch to see if we can actually make them a pod of one, because we've seen that happen to my company. The way this is structured for us is for me, I'm going to spend time with individuals each week, but I'm also going to do a 60 minute cohort where I'm just showing a bunch of things. I actually tried to sell one individual on my team, and I basically showed him the ropes. I probably met with him 13 times or so, and he's just flying now because he understands the AI piece and he understands what needs to happen. But he said more important than anything is I helped set the tone in terms of what my expectations are, what the speed is, and that helped him a lot. He kind of just took it in a roll with it. I'm not even going to give myself a ton of credit. It's more so I was able to help guide him a little bit. He was able to take it and fly with it because he was talented, and he still is talented. Real quick, if you're enjoying this video, you got to check out Single Brain. So Single Brain, these are the managed revenue agents that are unified by a single brain that lives inside of Slack or Microsoft Teams that your team can chat with. You can do data polls. You can strategize with these agents, and these agents can do, we're talking scaling ad creatives up to a lot. We're talking about creating email infrastructure for Outbound. We're talking about AEO campaigns. There's just so much to be done with revenue agents today and just revenue opportunities. Go ahead and check it out. Just go to Single Brain with a B, singlebrain.com, and we'll see you on the other side. And so basically, the way I see it now is I can rotate people around. The whole idea here is that if you're thinking about your team now in terms of pods of one or armies of one, then you are adding a lot of leverage. And I think a lot of people are probably going to be a little confused in the beginning. Maybe there's some anxiety. That's why you want to do it in a test batch initially. But if you want to get everyone up to speed when it comes to AI, I think testing this on a small batch first is going to be very helpful. And then you combine it with the skills piece. We start to teach people how to do that. And then what happens is after you run this initial back to the works well, the other people that you tested in this first cohort, they can coach other people. And now you have a whole pod of one apprenticeship that keeps on going. And that is super high leverage. Now, the third piece, when it comes to AI transformation, is you have to set certain standards in your company. You have to create these artifacts, drop all these artifacts. And these artifacts, by the way, can also become skills. But I actually learned about this from my friend Eric. And he talks about this concept known as command by negation. I will do X related to Y by Z, unless otherwise instructed. So these three words are very important because naturally, let's say the larger organization is, there's going to be resistance to AI. And they're going to say, I'm waiting on the training. Oh, I'm waiting on this. Oh, I just didn't have the time. So here's the thing. Whenever you get a blocker, you have a default rule now. I will do X related to Y by Z. As an example here, I will move the approved homepage sections into design related to the Q3 launch by Friday, noon, unless otherwise instructed. I will hold the disputed headline and mark it as swathable. Okay. So this is its client feedback is late. So design assets aren't ready. I will use the existing approved template related to the webinar promo by 4pm, unless otherwise instructed. Final creative can replace the placeholder without blocking QA. So now you have a team that they have these algorithms, or they have these certain rules where it's just like, hey, we're not going to wait for this. We're just going to keep moving. We're all going to march in the same direction with these skills in one way or another. Maybe we can fork some of them, right? And then we can arm people with kind of these suits. So let's say like an Iron Man suit. Everyone gets their unique Iron Man suit, right? That's what the skills do. But then when you add in these rules where it's just like, hey, our default path is just forward. And that is a rule that's going to help your team move faster as well. And then within the world of AI, you have to move very fast and people have naturally unblock themselves as well, because these are changing every single day. And then when we talk about the pod of one, the army of one, if people are able to think like that, that and they're able to move very quickly and they're able to be high agency, unblock themselves, and they're able to operate themselves with skills. The way this all turns out is that this becomes a video game and it becomes fun for everyone. And some people aren't going to want to come along. That's okay. At least you set the standard. Because you set that standard, you can decide how you want to go about moving forward with these people and it's a conversation you have with them. And then for the rest of the team that wants to move forward, great, everyone's having fun now. Everyone feels a lot better when it comes to AI. They feel like it's here to help them with the work. It's not there to necessarily eliminate anything in front of them. It's out there to take their job away. And that's how all this works. So literally, you rewatch this video, share this with a lot of your friends. This is how you move things forward. This is how you take your team from going to not so good with AI to pretty damn good when it comes to AI.