Luke Arrigoni’s Groundbreaking Vision for AI’s Future

Luke Arrigoni’s Groundbreaking Vision for AI’s Future

Sean: We talked about how to use AI when you’re a small to medium business owner. I’m wondering now, is there a way to use AI if you’re a mid-level manager or maybe just a business owner wanting to improve your life, your personal growth? Is there a way to use AI to be able to make that easier or augment that?

Luke: I think so. If we could separate those two out. I think for small business owners, the vendor approach is probably still the best way is just making sure you get your vendors correctly because most data science programs are like hundreds of thousands of dollars. Like that would be a pretty substantial investment. If I were smaller, I would just pay someone else to do that, right? Yeah. But if you’re a mid-level manager somewhere or maybe a senior level manager, you’ll probably have your manager breathing down your neck saying, What are you doing? What data like, what are you doing with your group’s data? Like, when are we going to do machine learning in your team? Right? And so that’s probably a little bit more stressful, right? Because it’s like this big thing and everything I just mentioned, it’s expensive too, right? So you have a PNL and you’re like, What am I supposed to do with my PNL? And the most expensive service on earth right now, which is A.I.? It’s like rocket science from the 1960s.

Sean: Yeah.

Luke: It’s a fortune, right? It is. And so obviously vendors approach, right? You can, you can go buy it, but what I would do if I were you, if you especially if you’re running engineering teams or you have a small even two or three people on your engineering team, I would get them to start upskilling. It’s not an impossible skill. If you already know software development like your team may, then I would push them to do tutorials and learn online and start to kind of dip their toes in it. What I found is that most engineering orgs, most software developers, they want to learn, they want to be given an opportunity and they don’t really know even how to ask their managers for it. They’re just like, I don’t know how to go to my manager and be like, Can I do data science? Whenever, whenever managers say, Hey, do you want to learn a little data science? I’ve never seen a software developer say, No, I don’t. They all want to jump on it. So I would encourage upskilling your teams. If you’re kind of resource constrained, that’s probably the best possible path.

Sean: It is a little daunting when you look at it right. Learning AI. I don’t think it should be. I mean, with what we discussed earlier, there are so many companies saying that they have AI in their software. It’s just been generalized. Right. But I don’t think it’s still that easy to make a real one.

Luke: You know, this is a good point. So it’s not easy to be great at AI, but for most people, there are some simple problems like that vendors can’t solve because they don’t have your data that a simple machine learning model might. There is a cross section which I trust. Your listeners will know when they fall in that cross section where it does make sense. But you’re right. If someone wants to say, Hey, I’m going to start a state of the art data science program at my company. Yeah, you’re probably in for a world of hurt, right? But if there’s a specific problem when you’re like, I can’t get a vendor to do it, and I have a couple of hungry techs that really want to figure this out, I would trust them. I would. I would give them a little bit of rope to make that happen. See? See what they do with it.

Sean: Right? Right. So we have this. The huge spectrum of what I am from snake oil, which is not real. And they’re just saying AI because it’s hot to companies selling software that has some form of machine learning in there, branding it as AI to companies who are really doing AI like Google and Microsoft and all these companies doing it in the back-end to ChatGPT, which is. We It’s so obvious that it is an eye. What do you think would be the next maybe 2 to 3 years of I as we graduate, hopefully from this pandemic and we face the US recession that’s looming 2 to 3 years. Like for me the most advanced version of AI is ChatGPT right now. What do you think can happen?

Luke: So I think there is there’s two different fields that I would want to touch on. The first is technically, I think everything’s going to become multi-modal, something we actually talked about a little bit where images and text are taken together and these AI’s learn from both. So you could say, Here’s a picture of my washing machine, what’s wrong with it? Right. And it would know like the brand and the model and I tell you like the basic stuff like that’s very near future. Yeah. I think near to mid future, like maybe five years we will have AI that analyzes our social life on a deep level, not social media, something that like lives in our phones that says, Hey, your mom is stressed about something, She just sent you an email and I can tell by what she wrote that you should probably call her and ask her what’s going on. It’ll be this thing that is almost like a personal assistant that can read different emotions through all of our communications that come in and out. It can listen to a phone call and say, you need to call Client X because they really are. They sounded like they were really trying to get you to pitch them on a new idea. Like you’ll have this thing that kind of lives in your head. Yeah, not literally, but you’ll have this thing that lives on your phone that listens to all your communications, that reads all your communications and helps you navigate complicated social systems. And this is why humans were meant for very small social networks. 20 people, 50 people, tops.

Sean: Yeah.

Luke: Right now most people, we’re navigating in the hundreds or even the 1000 some people, and we’re really bad at it. Even the best people are like they’re operating at like 10%, right? Like, yeah, they’re so below failure on it. Right? So tech like this, just imagine ChatGPT. The combination is something that will make everyone really good at being social again, which I think is something that everyone is really craving right now because social media is kind of delivered the opposite. It’s like anti social media these days. I love it. Like I’m on Tik tok, I’m guilty. I’m not trying to be religious about things like turn off, or disconnect Instagram. Like I have an Instagram, right?

Sean: No one’s judging you.

Luke: Right? But like, it’s true that it’s anti-social. Yeah. And like, the definition of anti-social these days is you meet a group of people and everyone sits on their phone on Instagram. Yeah, it’s like it’s become an anti-social tool and I think that this kind of AI stuff that will come up will actually help us be more social again. So I’m, I’m excited. I think that’s what the future holds.

Sean: That is what will be very useful. I do agree 100%. I’d love to see that. It’s like from how you were describing it, it’s like, it’s like Jarvis or Friday, right? Iron Man’s.

Luke: Not far from that.

Sean: Do you think, though, do you think that with so many concerns about mental health right now in suicides, that it could it could it could go to that level where you could talk to it and could fix some of your worries or anxieties in your head?

Luke: Possibly. I think there is , I know right now there’s a lot of effort around how we can use language models. And that’s what ChatGPT is. We call it a language model, a large language model. These large language models, how they can help with mental health. But I think the unfortunate part is, especially in America, there’s a huge mental health epidemic and no one really knows, even as humans, how to help other people. So we go on that whole, how do you train an AI to do something that even that’s nothing we’ve never been done before. You can’t. So if humans were really good at helping other humans with mental health, I would say, yeah, I would be the perfect way to step in. But unfortunately humans are really bad at helping other humans with their mental health. So we don’t really have a data repository to teach A.I. how to do that yet.

Sean: Yeah, yeah, you’re right. 100%. That makes sense. So again, yeah, we can’t program the we’ve got no data working data model that is like 100% successful. Right? And it’s a case to case. Everyone’s had a different history of upbringing in the past. It’s complicated. But it’s very interesting that it’s very mind blowing as well when you say it’s going to help us. With how we manage our social life. So is that something that you think is currently being worked on right now to be able to like who in his who’s going to allow access to that kind of AI. Right. Who will know? Every phone call, every email, every message in every social media platform or communication platform.

Luke: I think to start it’ll be a corporate tool, a sales tool, right? Like you want to make more sales. Well, you put this tool on all your sales calls, put this tool in your email like it’s going to be about money, right? And then from there, people will say, well, you know, it helps me navigate 150 clients, right? Like, that’s incredible. I’m always knowing what the client needs, what, where, when and what they’re thinking about. Like, I bet you I could put this thing for my social life, right? And that that jump isn’t going to be far. Jump.

Sean: Wow. That’s going to be an amazing tool. That is to be an amazing tool. Okay, so I think that we got a lot of insight about what could happen, what’s happening now. What would be your advice to people listening in? I mean, they’re scratching their heads right now and saying, Yeah, I’d love to use AI for my business, but I don’t know where to start. Like, what would be your advice for them?

Luke: So there’ll be different tiers, so I’ll go down the tiers real quick. If you’re really, really small, like it’s just you, you’re running an e-commerce shop that your vendors will wonder what to plug in. You buy all those plug-ins, each your margins, right? If they’re promising AI, make them explain what it is. If you’re a little larger and you have your own company, you have maybe 20, 30, 40 employees, like you’re making it now. At that stage, you might want to consider asking one of your developers to learn it or something along those lines. Otherwise, still vet vendors. And then if you’re a mid-level manager, you definitely want to start encouraging your technical staff to learn at least basic data science and everything above that. You should hire basically.

Sean: Yeah.

Luke: Yeah. Like that’s you’re talking like if you have a, you know, a $10 Million budget. Luke Who do you call? It’s like, I think those people give me a call. You know, our team will help you out, but like, that’s not really most people, right? Most people were not making those kinds of decisions. We have real practical decisions. We have real constraints to our business. It’s what makes us really good at business. It’s like anyone can be good at an ecommerce shop if you don’t have to care about your expenses. Right. I’ll just spend $20 million a month on marketing and make $5,000 and call it a success. It’s, you know, like a bullshit approach, right? But like, what makes everyone good at business is how you can be successful with your constraints. So with that being said, everyone with constraints, you probably do need to do stuff with machine learning. Try and get as many other people to do your machine learning for you. Get your vendors to do machine learning. Get other tech staff to learn a little bit. Don’t dive head first. If you have constraints with resources on building your own machine learning program, it’s tilting at windmills.

Sean: Right? Cool. Hey, look, if people want to get in touch with you, they want to get in touch. They want to get to know more about AI machine learning. They want to pick on your brain. Where is the best place to do that?

Luke: Connect with me on LinkedIn. I’m very accessible. Shoot me a message there. We’ll set up time. I’ll send you a calendar link. I don’t really mind. You can also reach out through my site and someone on my staff can talk to you as well. That’s africa.com ARRICOR.COM. But if you just type Luke Aragon into Google or LinkedIn, I’m the first hit everywhere. So just go ahead and reach out and say hi.

Sean: Awesome. Hey, Luke, thanks so much for being on the show. Really appreciate you. We are better for it. I love being here.

Luke: Thanks for having me.