← The Journey with Morgan DeBaun
The Journey with Morgan DeBaun
AI in Action: Practical Tools and Strategies For Your Life & Business with Allie K. Miller
00:42:21
Loading summary
Transcript57 lines
- [00:00]
Morgan DeBont
Hi friends, it's Morgan debann and I have some exciting news. My book, Rewrite youe Rules the Journey to Success in Less Time with More Freedom is now available for pre order. Pre orders are so important, not just for me as a first time author, but for the message of this book. Ordering this book helps amplify the message that more people can break free from the grind, rewrite their own rules in life, and live a rich, juicy life full of joy and empowerment. So if you resonate with any of the content that I put out over this last few years, this podcast, any of the stories I share on social, and you've been inspired by my work at Blavity or Afrotech, pre ordering is the best way to show your support and your gratitude for any of the work that I have done. Plus, you'll be the first to get your hands on the book when it launches next spring. And that will mean literally the world to me. So make sure you check out morgandebond.com to secure your copy and join me in this movement to create a life of freedom, abundance and more importantly, to have the tools and the frameworks that you need to take control of your life. Hey everybody. Welcome back to the Journey podcast. I'm excited today because we have the incredible Ali K. Miller here. We are talking all things AI. You all know this is one of my favorite topics and I'm personally on a mission to help all of you all be more advanced with how you are using AI in your everyday life to help you work smarter, not work harder. There's so many more tools today than there were even a year ago and the tools that are existing today have become so much more advanced. I am constantly even challenging myself to say, okay, but is there a better way to do this? Like is there a different tool that's come out? I'm tracking things like looking at venture capital funding to see, okay, who's raising these huge rounds. Let me go and check out. There's certain VC firms that I think even more recently have been doubled down on a lot of the existing technology. And so it's been really fun to see all the different ways that people have been using it in their day to day life. Also recently at Blavity, we just turned on Gemini for our Google kind of workforce. I've just been testing it with maybe like 20 or 30 of our directors and employees to see how they've been using it and it's been fascinating seeing like different people use it for Google Docs versus email. And so I'M really excited for Ali to be joining us today to hear more about how we should be thinking about building AI into our everyday lives. AI is not here to replace you, but it is here to help you do more things. So, Ali, welcome to the show.
- [02:36]
Ali K. Miller
Thank you for having me. I also, I did see that you're creating AI twins of yourself. And so I feel like even though I'm going to be sharing how I use AI, I definitely want to hear how you're using it too.
- [02:47]
Morgan DeBont
I'm testing it, I'm testing it. Everyone thinks I'm a little crazy. They're like, why would you give the Internet, like, your face and stuff? I'm like, first of all, the Internet already has my face. So, like, what are we talking about? Hey, everyone, I'm Morgan DeBont, a passionate entrepreneur and life advisor. With the Journey podcast, you'll discover that success isn't about the destination, it's about the journey. I'm sharing stories of amazing people who've taken control of their lives. Join me on my own journey to discover the secret sauce behind reaching success. With permission from no one else. Ali, tell me from, like a day to day perspective, like, how you incorporate AI into your current workflows.
- [03:29]
Ali K. Miller
Yeah. So just as a bit of background, I started in AI almost 20 years ago and have worked in this space very heavily for the last decade, have worked at IBM, have worked at aws, and now work myself. And so I just seen AI through a lot of different lenses. And so I know a bunch of your audiences, entrepreneurs or corporate professionals. And so literally every single person listening to this or watching this, you're gonna have a different way of using AI. So I'll try and give you a few of the examples that I use. But ultimately it's gonna come down to your awesomeness and creativity and brilliance to figure out how to customize it into your own life and bring it in a way that feels really authentic. So couple, couple ways to maybe kick it off. And Morgan, I know that you said that people call you crazy. I think that that's like, kind of the goal, right? With any single new technology, right? Internet comes out, and I'm sure people are like, I'm going to put my restaurant menu online. And everyone's like, why on earth would you do that? If people could just come to your restaurant and open the menu? So all these things feel crazy for the first, depending on what the tool is, maybe even couple weeks. And then you realize that the whole world has changed and that everyone around you is actually the crazy person for not figuring it out. So one example of that that I do think feels a little crazy when I tell people is that I have a 10 to 20 minute conversation with AI every single morning. And so depending on what the task is, I either use ChatGPT or hotter AI ChatGPT. If I want it to be a back and forth conversation, there's a new feature called Advanced Voice Mode. You'll see it abbreviated as abm. And that is the really natural sounding voice that you can interrupt, that you can ask it to speak faster, you can ask it to speak in a British accent, you can ask it to only speak, you know, without verbs like whatever makes you happy. But that's going to be the more natural back and forth conversation. If it is instead that I want to dictate to AI for 20 minutes straight, then I'll use Otter AI, which is a transcription tool you can try for free. Both these you can try for free. And so for that I'll have that conversation. Because when I start my mornings, I have a to do list, right? I'm looking at either a post it note with a bunch of different things that I have to get done, or maybe it's a note on my phone or whatever that is. And so I might want to send in that to do list to AI to have an AI system like pre deliver as much as it possibly can, or I might have one to do list item that says, you know, prepare a business proposal for Comcast, whatever. And so I might get on a transcription tool and I might talk through this business proposal. I might be on a walk through New York City. There's dogs, there's honking, and I'm talking through what I want to get done from this business proposal. And I'm giving it every little bit of context. I'm saying, you know, my client thought this way, but actually there was this confusion point over here. And the real thing that I want them to get across is this, but I'm going to say it in this way, but could you make it sound more business formal? And so you can add in all of that insane context that to a human would feel like you are completely overwhelming them, right? If you were just delivering this to an intern and you spoke straight at them for 20 minutes, that would be insane. But to an AI you dictate straight for 20 minutes and then you can send that into either just a prompt or a pre built up or something and you can say, turn this into a business proposal, right? And so that walk, that beautiful morning walk to get your coffee just turned in to productive time and then you still as a human will go through and review it and edit it. But oftentimes my morning is kicking off with some sort of conversation or to do list automation.
- [07:23]
Morgan DeBont
That's awesome. And I'm curious because I actually do something relatively similar. Like I will look at, I keep a weekly list of like all the things that I didn't do the last week and then decide if I need to keep doing them or not doing them. And then I will dump some of those things into ChatGPT or CLAW, depending on what I'm actually asking it to do, and say like, okay, prioritize this. Can you write any of these, like, email, draft these emails for me and like, things like that? One of my challenges has been training ChatGPT on me and who I am. So I actually created my own GPT within ChatGPT where I basically uploaded all my previous newsletters, like my bio transcriptions of some of my podcasts, where I felt like I kind of gave more of my background of who I was. And that has been really helpful at it, coming out with outputs that sound more natural to how I communicate. You know, I'm not a wordy person. I'm very direct, you know, so I had to be like, ma'am, I would never say this much.
- [08:21]
Ali K. Miller
We regretfully inform you. Like, just say the words.
- [08:25]
Morgan DeBont
Yeah, yeah, just say immediately no. Okay, so I'm curious, like, how have you like, customized ChatGPT or Claude to you? And like, how would you recommend somebody approach that?
- [08:38]
Ali K. Miller
Yeah. So one thing that you're clearly ahead on is figuring out that the generic AI that everyone has access to without giving it specific information, specific context, specific examples. Right. All of your amazing newsletters, without that context, it's going to come out incredibly generic. And by the way, for people who read a lot of, you know, job applications, or for people who read a lot of social media comments, it becomes very obvious which ones are AI generated. So even if you don't provide the context, this is an important piece. Even if you don't provide the context and the output looks good, it will look to others as generic and you will have not learned how to use these AI tools. So first step is definitely just figuring out how to talk to, you know, ChatGPT or Claude or Gemini or any of these things. And what you get back is likely to be very generic. A good, good, good prompt or a good system in your case, for that GPT includes tons of context. And for your example, it's context on you and how you write. My big tip is that these systems can be very skewed toward the examples that you give it and they are not yet smart enough in air quotes for those who are listening to really pick up on okay, this is how she does newsletters. And the way that she does newsletters is more witty, but the way that she does social media comments is more, you know, like short and blunt. So it's not going to be able to pick up on that sort of nuance. So what we do and what I certainly suggest, even clients and their marketing team, is that you actually kind of have to have a different GPT or system or prompt or whichever ua you want to go after it. You have to have one for your newsletter, one for social media posts, one for Instagram captions. You might have one for newsletter subject lines. You might have one gossip tweet, CTA's. If you find that no matter how many examples you give this thing of, you know, your entire newsletter, it's just not getting the Morgan voice or the Ali voice or the John voice or whatever at the end, scrap it. Create a new GPT or create a new, you know, SOP standard operating protocol procedure like create a, create a new set of documents and examples just to solve that last thing. With an extra literally 20 minutes of effort, you can immediately solve that problem where it's not quite getting you. And these models are changing all the time, updating all the time. I love that you are encouraging your readers to stay up to date. Certainly reading daily news is probably overwhelming, but just hearing from friends, having this as text conversations, you need to not be updated on every single hour. But when a new model comes out, you should see that as an opportunity to say, man, it really didn't understand what I wanted out of subject lines last time. Let me test it this time and see if the new model cracked it. If the new model gets me better, like that's actually a really interesting time to lean in and experiment.
- [11:53]
Morgan DeBont
Yeah, absolutely. I think that's a really good point. And I think that I have maybe thought that AI was a little more advanced than it was because I definitely had one GPT. And then to your point, I was having it do my newsletter and then I was having it to help me with social captions and it was just messing it up. So I had to create a GPT just for newsletters where it was like start the newsletter with this and then you go into like two or three quick facts and then you do this, you know, and I had to actually Tell it what I the structure of the output and to give you just.
- [12:22]
Ali K. Miller
A sense of skill. Because this is always helpful for me when, you know, talking to someone who's using a tool in a new way. Just for my newsletter, we have five GPTs. Just for the newsletter. Yeah, absolutely. That's awesome. Because each of these sections are treated so differently, right? And maybe these systems can be stacked in a more interesting way or we could set up an AI automation to be able to think through these steps and kind of call an API in different sections. But just for everyone on my team to think through that level of granularity, you know, for your, maybe it's for your pitch decks, right? You have a bunch of entrepreneurs in your audience. Maybe it's your next pitch deck. Okay, should you have one GPT for your entire deck? Absolutely not. You should have one for the headline of your pitch deck, one for the headers of each slide, one for the email that goes with the pitch deck that you send, that's catered to that VC that is based on their website, their most free, frequent, you know, pillars of investment, their last three investments, their whatever. Like that is the type of granularity that you don't necessarily need from the get go, but that is what's going to make these systems skyrocket on performance for you. I, like, a lot of people freak out and I'm sure you've heard the same. A lot of people freak out that they're not using these tools to the full extent, right? And like, oftentimes one of the solutions is to think smaller and to think in some tasks and to not assume that whatever this model or platform is, is the silver bullet for all your problems and that you can throw the kitchen sink at it and it will do it all. Just break it up, ask it one thing at a time. That's how engineering works too, you know, start with one thing, make sure it doesn't break. Keep going, keep going, keep going.
- [14:11]
Morgan DeBont
Now, I think what you're saying is actually really important. It's even important for me to hear because, you know, sometimes I think it can be, it can feel like I'm doing too much. If I had 50 GPTs, what I'm saying, like totally. No, like that feels like, start there.
- [14:28]
Ali K. Miller
You'Re not going to start, right? But now that you're at the level you are at, maybe 50 is actually the right number. I mean, you have so many things going on that literally maybe 50 is the right number. My team, I think we, we definitely have more than 50 and we probably use 10 very regularly, 20 decently regularly, and then, you know, there's a long tail where it's just solving this one freaking problem that annoys us. You know, once every two weeks, once every three weeks. But you'll scale up. You don't start at 50, but super users who are non engineers might absolutely have 50.
- [15:04]
Morgan DeBont
So I haven't gotten to this level of advanced yet, but I'm curious what the entry advanced level would be for someone who wants to. You can actually set it, connect, have your AI run through different steps. So like you have four or five GPTs to write the one newsletter, but is there a way that you go to another tool that says run through GPT1, then go to GPT2 and then input all these outputs into this little place?
- [15:33]
Ali K. Miller
First of all, if no one has told you that you have like a product manager mind of thinking, that was my first job. Okay, there we go.
- [15:41]
Morgan DeBont
I was.
- [15:44]
Ali K. Miller
Slam dunk. So, I mean, it's very clear that you've, you've maintained that way of me. That is one of my favorite ways of using GPTs. So maybe it's helpful to hear like the graduation path that I would suggest for people and you're kind of like stage three, stage four, and for folks listening, maybe you're on stage one. Okay, so stage of AI usage, right? Let's say you're brand new ish to ChatGPT or even you've been using it for a year, but you still feel like you're using it in a very basic way. Step one is literally opening up a tool. And let's just say for all intents and purposes, we're just going to talk about ChatGPT as well. You open up the tool, you see a blank page, and all you're trying to figure out is the ui, right? You're trying to figure out the interface. Where do I type in my stuff? Where do I add an attachment? Okay, that's the conversational thread. And so you're just going to ask you questions, you know, write me a poem about dinosaurs, as if you're afraid of dinosaurs and you really want a smoothie, but actually your favorite drink is a boba. Okay, you're just going to try something just to see how it reacts and how it responds. Okay, the next thing that you're going to do is you're going to get out of the generic space and move into the hey, now it really gets me space. What does that mean? That means you're going to elevate your prompt. You're going to give that system your goals. The main task, you're going to break it down into subtasks, you're going to give it examples and then you're going to revisit and revise. So I call it the wiser framework. Establish who it is and identity, give it instructions, subtasks, examples and revise. You're going to go nuts at it. You're going to give it five examples of your newsletter, 50 examples of emails you sent, right? And then it's going to quote, unquote, get you a little bit more. Stage 3 of AI adoption of ChatGPT usage. You are going to move into an efficiency stage. You are going to realize that, ugh, you know, I write newsletters every day, therefore I should build a GPT for that, right? You're going to get into your head about which tasks you're doing multiple times. When my team does something three times, we usually build a GPT for it. And so that's when you're gonna get that efficiency. So you don't do re prompt every single time or go into a thing and find the old conversation. We're, we're past that. So stage three, you're already passed that. Stage four is really turning it into a workflow. Okay. It is stacking GPTs. When you are in a chat GPT conversation, you can literally act a GPT. I don't know. Yeah, no, I didn't know that. Writing a newsletter. Let's say you're writing a newsletter, you can add subject line generator and it comes up with one and then you add the newsletter writer and then it'll edit and then you can add grammar checker and it does that. We even have a GPT that we made public called no More Delve, that removes all of the most frequently used AI words from anything that it outputs. So we're not using delve, we're not using realm, we're not using landscape, nothing. So it'll do that swap out. And so we'll call literally like one after the other after the other so that they're building on the past one's actions and output. The last one is an end to end workflow. That's when maybe you have a Google Drive trigger that every time you upload a photo to a specific folder in Google Drive, it automatically triggers an Instagram Caption writer, a TikTok Caption writer, and it'll also, you know, come up with five songs that you can put as the background based on what the photo is. And it'll tell you the caption that you should put in your broadcast channels. Right. So that's, that's automations. There are no code platforms for that. You certainly do not have to be an engineer to get to that stage five. But that is the sort of flow of advancement. Whatever stage you're at, look ahead to the next one and just start experimenting. It sounds like in the next five minutes you're going to be past stage four and onto stage five. Right. But whatever people are doing these days, just like, no one is judging you for how you're using this stuff. You know what I mean?
- [20:09]
Morgan DeBont
Yeah.
- [20:10]
Ali K. Miller
Like, no one.
- [20:11]
Morgan DeBont
I think that's really important for people to hear.
- [20:13]
Ali K. Miller
Yeah. No one is looking over your shoulder, like, that's a stupid prompt. Why would you use it that way? Try the craziest, wackiest, zaniest thing, the thing that your high school friends made fun of you for. Like, do that inside of these systems, assuming that you're not sharing like PI, Social Security numbers or whatever, and, and just see what it outputs and like, challenge yourself to go to that next degree with these things.
- [20:39]
Morgan DeBont
Absolutely. I love asking ChatGPT First Opinion. Like, there was a trend on TikTok where people were screenshotting their Instagram feed and being like, okay, chatgpt, like, what do you think of me based off of this picture? And I at first thought that it was fake. Like, I thought people were just making it up. And then I was like, let me just try. And I was like, oh my God, chatgpt.
- [20:59]
Ali K. Miller
But you give me the greatest part of what you just said is I saw it and I tried it. That is. That is actually the minority of what people do. Most people see it and they go, I bet that doesn't work. Or they'll see it and go, that's cool. But so few people would scroll, scroll, scroll and go, whoa, whoa, whoa, I gotta check this out. Screenshot test and get it back and see how it works. I have to tell you, one of the craziest moments I had was with Gemini, with Google Gemini. It under it, it allows for video inputs, right? Not just photo, not just documents, not just links, whatever. You can input a video and ask for input. We could even put in a video of this podcast. And that's what I thought. So I scrolled through my TikTok feed very fast, right? Like, I wasn't watching every video for 20 seconds. Scroll, scroll, scroll, scroll, scroll. I took that video, I put it into Google Gemini 1.5 Pro. And I said, based on this video and this video alone, give me, 20 gifts that I can buy for this person that cost under $100, that can be shipped to the United States in under two weeks and that are personalized with a monogram. Do you know what I mean? Like, that's the sort of thing where you're just like, I do that and you throw it out. And I gave. It has no information on me. It was just that video input. I could have used your TikTok.
- [22:27]
Morgan DeBont
Totally.
- [22:28]
Ali K. Miller
It was unbelievable. I'm, I'm someone who loves to give gifts and I'm bad at picking gifts. And so, like, I love using ChatGPT for gift advice. I love using it for second opinions. It is a free or cheap second opinion. I mean, it's, it's, it's fun.
- [22:47]
Morgan DeBont
It's truly, it's so fun. It's so fun. I'm like, give me a problem to solve. I'll.
- [22:52]
Ali K. Miller
I mean, okay, wait, how did you switch your mindset from thinking about it as, you know, this tech. I know your whole background kind of sets you up a lot better for this, but like, what, what was that mindset shift for you?
- [23:05]
Morgan DeBont
I think for me, I really, I don't like to do the same thing over and over and over again. And there's part of my life and part of my roles and responsibilities that I do have to do over and over again. Like when I decided to be a public entrepreneur or started to decided to do this podcast, you know, then it comes with certain other responsibilities. Like you should be saying the weekly, the email, you should be posting your content on social. You should be doing these things because, you know, why do the podcast if you're not going to tell anybody about it? You know, so when I start to do something, it's not typically because I want to be doing the administrative work of it. And I was also getting to the point where I had a lot of virtual assistants. You know, I had a lot of people working in like the Philippines and I was constantly on upwork trying to find somebody to do this one specific thing that they would be good at. Because I've learned over time it's very hard to find, you know, one person who could do 10 things. It's much easier to find 10 people that do that one thing every for you every week.
- [23:56]
Ali K. Miller
Week.
- [23:56]
Morgan DeBont
But they're 99% accurate at that one thing. But at some point, I don't want to be managing all these people. Like, I really don't want to be a large people manager because at Blavity, I already have 100 plus employees so in my personal life, I'm like, I don't want to be dealing with people management. So I think AI has been a perfect solution for me because it fits my personality and how I like to be spending my time. And I can now say, you know what? We used to have a copywriter that wrote our weekly email every week, and she wrote the associated captions and all the things. She was great. I don't need her anymore, you know, and that's helpful for me.
- [24:35]
Ali K. Miller
You know, there was one thing that you said in there that just stuck out to me, which is that finding someone at, you know, who's good at 10 things is very hard, but finding someone who's good at one thing is easy. It's exactly like AI. It's exactly what we were just talking about with having AI just do that one thing really well. And then the second that stood out to me is you almost have to have like some of the highest self awareness of who you are and what you want to be doing in order to figure out how to use AI best for you. So for you, one of the insights of I don't want to be a large people manager outside of baby, because you're already doing that. That's an insight that you'd kind of only have if you really sat with yourself or having the insight of, I don't want to manage and find all these people on upwork. You had to kind of go through that experience, like going through your daily calendar, revisiting at the end of the day what you just spent all your hours on. Like every single day. We have 24 hours. Are you really spending an hour coming up with the grocery list? You know what I mean? Or we only have 24 hours. Are you really spending an hour editing your emails for grammar? Like, is that.
- [25:41]
Morgan DeBont
Yeah. And the answer was what you were.
- [25:43]
Ali K. Miller
Yeah. Yes. Which is, which is. Which is part of that evolution. Like, I don't, I don't knock anyone for, for being in that space. What I knock people for is hearing the podcast conversation that we're having right now and still doing the admin tasks that they don't want to be doing in two weeks.
- [26:03]
Morgan DeBont
I agree.
- [26:04]
Ali K. Miller
That's what I knocked people for.
- [26:05]
Morgan DeBont
That drives me nuts.
- [26:06]
Ali K. Miller
Yeah.
- [26:07]
Morgan DeBont
And I even do that with my own employees. I'm like, what do you mean you? I mean, there was a time when I used to write an email when we were fundraising, you know, venture fund. I would draft the email and then I would send it to my co founder, Aaron and say, okay, check this Email, make sure it sounds good, you know, check grammar, Just give me another pair of eyes. You know, we didn't have Grammarly seven years ago, right. Like there's all these manual things we had to do to just function. I'm like, if I meet founders now and they're telling me like they don't use AI to help them customize their pitch decks, to customize the email outros of the intros that they're asking, you know, the formidable emails that they're writing. I'm like, you're writing this from scratch?
- [26:49]
Ali K. Miller
Yeah. What we talk about, I tell people like I'm allergic to a blank page, you know, like, like AI should be getting you part of the way there. And then you'll still edit the heck out of it, make it awesome, build that relationship with that person. But it is, I mean just to give you a sense of like the productivity equation change that is happening in startups today. So your entrepreneurs that are listening, maybe they're building a tech based startup, right? Maybe they're building a, an Etsy shop, whatever. It doesn't really matter what they're building. The productivity gains that people that are using AI are getting. One example, I have a startup, they have five people at the company, the company has millions of users and they spend 15 minutes a day on customer support. 15. And the only reason that they're able to do that is because of AI. Right? They don't have a customer success manager or anything like that. They are able to scale and to use their venture funding, to use it for compute, to use it for advertising, to use it for attending conferences, to have those in person meetings. Right. That you still need to be able to build a business. Like that's where they're dedicating those hours and those dollars not toward something that they don't want to be an expert in. The prediction among most AI experts is that we're going to have unicorns that are a couple people and eventually one person. Right. AI is a massive scaling technique of what your awesomeness is, but it is also a helper for the things you don't want to be an expert at.
- [28:42]
Morgan DeBont
So yeah, I completely agree.
- [28:43]
Ali K. Miller
If you are an unbelievable entrepreneur, it'll help you on the literal tech or the Etsy shop you're building, but it's also going to get rid of some of the tasks that you don't want to be doing.
- [28:53]
Morgan DeBont
Right. So let's talk a little bit about people. I think it'd be ten times easier to start a business today that it was Five or six years ago because of AI. But I'm in a situation now where we have to do. We are literally doing an AI transformation. Like we are doing a change management, very peaceful approach, but a change management within the company to transition our process tools, Systems to be A.I.
- [29:20]
Ali K. Miller
Yeah.
- [29:22]
Morgan DeBont
Based. And I'm having a challenge, I think, not getting everybody on board. Everybody's excited because everybody knows that this is going to give them their own competitive advantage and like, help make their lives easier. But I'm having a challenge getting all the tools to integrate with one another. So like Salesforce with Slack, even though Slack is owned by Salesforce or Asana with Slack. And Right. Like we have all these DocuSign, we do all, we have all the tools. I'm curious, what is your advice when you're talking to companies and organizations, even like a nonprofit, that knows that they could be using AI but they're not exactly sure where to start? Like, how would you approach that?
- [30:03]
Ali K. Miller
So lots, lots of, lots of things in one. The crux of what I think you're asking is just like, where to start, right. Assuming that you are trying to transition to what I call like an AI first culture. And so as background, I started the AI for Startups and Venture Capital Group at AWS and then ran it and grew it. And what I could tell you is that a lot of the companies that are able to build AI first more quickly, and so the people that are going to be able to adapt really fast right now are just more likely going to be smaller teams and smaller companies because they can pivot and shift a lot faster. The really, really, really big enterprises that also have very deep pockets can also pivot very quickly for certain departments or certain business lines for everyone else. Here are a couple, you know, pieces of advice of what I would think through. The first is that as much as we want to think that this is a tech problem, and as much as we want to think that it's about what model to use or which tool to pick, it is first and foremost a people problem. And so your team might be fully excited and fully bought in, but the majority of companies are not in that position. I suspect that a lot of people listening or watching this are at companies where they are the person that sees the future and they know that AI is really big and going to be bigger and that they're surrounded by people who love writing emails from scratch.
- [31:43]
Morgan DeBont
Yes.
- [31:43]
Ali K. Miller
Don't want to use these tools and they're probably losing their minds and pulling out their hair.
- [31:48]
Morgan DeBont
So we're even Getting instructions ali that say, don't use AI. Like I've gotten RFPs from agencies say, do not put this RFP in ChatGPT. I'm like, yeah, yeah.
- [31:59]
Ali K. Miller
I mean, that's the thing. There are. Everyone's going to have their own, you know, risk tolerance or everyone's going to have their own rules. And so if you're at a company that you truly believe is banning AI holding you back, you know, not letting you use these tools and you feel that if you dialed forward and fast forward five years, where is that company and therefore where are you and where is your career? Like, I heavily encourage people if you're at a company that you believe is stomping on the brakes and is not going to be adopting AI in the next say like a year or two, right? Some companies just move slow. Nonprofits move slow. That's okay, you gotta leave, get another job. I know it sounds really easy to say and I mean that kind of in like a joking way, but like that's why I'm talking about this like longer timeline. You need to start looking for a company that lifts you up. You should always be looking for that. But we are in this AI age where actually part of lifting you up means giving you access to tools that are getting stupid things off your clip. It would be like joining a company 10 years ago and they have no, you know, messaging tool or whatever, or they have no spreadsheet tool. It sounds so basic. But in five years we're going to go, wait a second, in 2024, y'all didn't have access to AI. Are you insane? And we're going to see the people who have access to it and we're testing it out now are just going to know how to use these tools a lot faster.
- [33:28]
Morgan DeBont
So yeah, completely agree.
- [33:29]
Ali K. Miller
Being in that culture that is AI forward, that is willing to be AI first is that is the number one thing. And getting everyone on board. Where I would start after that is starting small, starting with either like some sort of horizontal tool like ChatGPT Enterprise or Microsoft Copilot, right? Or any of these. If you have an engineering team, you can also host your own open source model and build your own. But some sort of horizontal where everyone has like safe access to a state of the art AI model, not GPT 3.5. We're not going back in time. Access to a state of the art AI model. Right. So that's spot one of where I would start. And spot two, if you're at a large company would be figuring out More like true AI use cases. And so maybe one of your use cases is for every RFP you receive that you immediately break it out into like a structured CRM and then you auto upload that and then you automatically send a welcome email to that person. Right. That is a discrete set of tasks that that AI can do. It is something that is needed for your business. It is something that is feasible today, but that's where you could start. And so companies are going to start with thousands of different use cases. So I'm not going to share, you know, where people should start, but figuring out what matters for your business and what AI can do today is absolutely where you'd start.
- [34:57]
Morgan DeBont
So Ali, I guess just in closing for those people who are really trying to test different edge cases and like really expand their minds and really try new things, what are some of the like unconventional tools or things that you've been using or playing with recently?
- [35:16]
Ali K. Miller
Okay, first thing that I would recommend is that you kind of have to not do this alone. Like you kind of have to find your squad of AI weirdos. And by that I mean like people who are just so excited to experiment and share. And so maybe you could just have like a group chat. Like six people doesn't need to be at your company, but people who are texting back and forth going, did you try this? Did you try this? Did you try this? If you are going totally solo and you're not following people on social media who are testing and pushing you, you're going to likely stall out and plateau. And you need that, that mechanism that's kind of pushing you to keep trying weird stuff. That weird stuff might be a new tool, might be a new prompt, might be a new interface. Like I think one thing that is probably massively underutilized is multimodal AI. Like I think people are still pretty trapped in a text based chat bot. People are typing back and forth and hoping that that solves all their life problems. And they're not using voice, like dictate have a conversation back and forth. They're not submitting a photo. Like I literally take a photo of my outfit, send it into chat and I get fashion advice. I have an NFC tag on my mirror that I tap get advice back. I'm not a fashionista. Like it helps so much. I will take a video, right? And I'll input it into Descript or something like that to be able to automatically edit that video. I'll have a question on my mind and I'll ask perplexity and then use it to ask the next question. I have a deck, a keynote that I have to give and I'll go into Canva and I'll just see what new features are there. Or there's a new tool that I invested in from the team that like founded Uber Eats and it's called Cove AI and they have like a whole open space. Like, there's no chat bot interface. You ask it a question and it's literally running like four windows at once. Right. And that just launched. Like, there are so many new interfaces that are going to come out, so many new modalities. That's where I would kind of go nuts. You don't actually have to test out 40 different tools. Right. Especially if you're overwhelmed with how many tools are out there. You can find your core 2, 3 tools and go crazy deep on that and still be way ahead of everyone else.
- [37:43]
Morgan DeBont
Absolutely. I think that's. That's awesome. I was just. Every year at Afro Tech, I do a giant entrance of whatever is the trendy thing. So back in the day when scooters, when all of a sudden all the scooters, you know, lime and, you know, Lyft had their scooters, I like rode in on a scooter, you know, which at the time felt revolutionary because people were very confused about scooters. One year I had a metaverse avatar walking around a fake Afrotech world and, you know, did the welcome through that and I was like, wait, I'm here in person. I thought, we're all here in person now. We're, you know, come out. So this year I was going to have my intro be my avatar version of me giving the intro and welcoming everybody to Afrotech and then being like, okay, so you can get off the stage now. Like, the real Morgan is actually there.
- [38:33]
Ali K. Miller
Yeah. People, you can. I mean, it's not a clone. Right? Again, air quotes. But you can create a synthetic avatar of yourself. There's Synthesia did and hey Gen to do the video. You can do your voice with 11 labs. You can do the speech or script with Claude. Like, you can tie all these things together and you can have an AI ally that is taking over your next whatever. Even though you can do that, and you should absolutely do that for your keynote, there are so many times where people are probably using it in spaces that they shouldn't. And so maybe this is my moment where I just get to say, test all these tools, figure out where they are unbelievably successful and where they fail miserably. And even with that knowledge in mind of, ooh, I can make an entire AI Ally. You then have to make that very human decision of just. Should I? Right. You get to make the decision of whether AI Ally takes all of your C suite meetings for the rest of your life. The answer for me is no. But the answer to, is it going to be the greatest keynote kickoff? Absolutely. So it is always going to be context specific, environment specific, people specific, but just testing out these tools, that is that first, first step.
- [39:56]
Morgan DeBont
Absolutely. Well, you all, this has been such a fun episode. Allie, thank you so much for coming on the show today. I hope for everyone listening in, they've got an inspiration to try something new like that is my only purpose in life. And the only purpose of this podcast is to help people on their own journey, wherever they're starting. So whether you are level one or you're getting to level four, on which you sis wherever you are. But Ali, you have so many resources. You've got your masterclass, you've got all these different things. Where should people connect with you? How can they get more in your world and be in your tribe since you are, you know, at the cutting edge of so many of these new technologies?
- [40:35]
Ali K. Miller
Yeah, I appreciate that. I have every social media channel and it's Allie K. Miller, mostly on LinkedIn, but certainly posting more and more videos on TikTok and Instagram and all that. For folks that have masterclass, I know that Serena Williams has a way cooler masterclass, or Gordon Ramsay is a way cooler masterclass, but if you want to nerd out on AI, I just released my masterclass and honestly, there are so many free resources out there. Some of them are going to come from me. But anything that you find, just grab it and use that as a moment to take advantage of this moment to set yourself apart, to pull other people in with you and lift everyone up. And I hope that you guys have many more AI episodes to come and that this is just one of the few that you've already done, but I can't wait for everyone to keep using this stuff.
- [41:31]
Morgan DeBont
Amazing. Well, Allie, thanks for being on the show today. And you guys leave us a rating. Leave us a question. Let us know what you think and how you're using AI in your life. See you guys later. Thanks for listening to the Journey podcast. If you enjoyed this episode, make sure you leave a review and head to our Instagram and YouTube to leave a comment. I look forward to hearing how this podcast has made an impact on your own journey.