Celebrating Everyday AI's 700th Episode

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life. Most AI tools that analyze sales or customer support calls turn those conversations into a text based transcript. But text only transcripts miss most of the value and Modulate fixes that. Modulate's new Velma Voice Native AI model and their ELM technology actually understand what's happening on those calls. It picks up the valuable tone, timing, emotion and intent that all AI transcription tools can't provide. So whether it's for sales, customer support or voice agents, Modulate's new Velma model helps you capitalize on what text only AI tools miss. Demand more from your AI today with Modulate Modulate at Modulate AI for almost three years, I've spent almost every single weekday talking to the smartest people in AI, reading about the latest AI advancements and putting it all into practice and hopefully helping some of you along the way. And I feel that over the years I picked up a few helpful tips along the way. I mean, that's what this podcast is all about. But I know sometimes those best practices tips and tricks are spread out out all over the place. So once in a while we do these kind of special milestone episodes that I think are a nice huge knowledge dump and maybe even a gem for those of you staring AI implementation in the face with a lot of questions. So today we're doing one of those milestone episodes as we celebrate everyday AIs 700th episode. Don't worry, I'm not going to rattle off 700 facts and stats. Don't worry, ain't nobody got time for that. But we're going to be doing the 7 times 10 times 10 route and going over the 7 ways AI is reshaping how we work, the 10 AI workflows that actually deliver ROI today and 10 AI skills every professional needs to know in 2026. All right, I'm excited for today's episode. I hope you are too. Thank you for joining me. And let's get into it. Welcome to Everyday AI. If you are new here, my name is Jordan Wilson and this thing's for you. It's an unedited, unscripted daily live stream podcast and free daily newsletter helping everyday business leaders like you and me keep up with the non stop AI advancements. How to make sense of it, cut the BS and just grow our companies in our careers. So if that's what you're trying to do, awesome. You're in the Right. Place starts here. But stick to the next level. Make sure to go to our website at your everyday AI dot com. Sign up for the free daily newsletter. We're going to be recapping not just today's show, but also all of the daily AI news that you need to know know. So it's one of those special episodes, but we actually did another pretty good one on our 600th episode. So if you missed that one, hit rewind a little bit and go check out episode 600. We did. The six AI myths you should stop believing, the 10 AI systems you must learn and 10 AI trends you can't afford to ignore. So make sure you go check out episode 600 if you missed that. But let's get over and talk about episode 700. That is today's and live stream audience. I'm curious, out of 700 episodes, how many do you think you've listened to? 5, 10, 700. Let me know. I'm curious. But like I said, we're going to be running down these things into three categories today. Episode 600 I think was kind of the map. And this is the game plan. All right, so maybe if you're listening on the podcast, hit pause, maybe go listen to episode 601st if you missed it, and then come back and hit play on this and let's get into it. So let's first talk about the seven ways that AI is shaping work. So I'm letting you know these aren't predictions. These are already happening. And I think that the traditional corporate hierarchy was designed for manual knowledge work. It was made for smart humans and domain experts to be able to go synthesize and personalize information and to create value for businesses. And I think that that design is becoming a little obsolete. So let's get into these seven ways. So way number one, that AI is reshaping how we work AI first, right? This is an easy one. You've heard me talk about AI operating systems. This is an example of that. But right now you aren't using apps, right? And you're going to be using apps even less. Right? Another good example. This changes all the time, right? So a couple of weeks ago OpenAI kind of got rid of connectors and they moved everything to apps. In the last couple of days, right, we've seen Anthropic's Claude cowork really take off. Less than checking my watch here 24 hours ago, anthropic also released their version of interactive apps. So everything is really moving to front end large language models. And you've heard me talk about this a lot. But right now, even if you aren't doing this, your employees probably are. Right? A recent study showed that 78%, 78% of professionals now bring their preferred AI tools to work regardless of company policy. So if you are a decision maker right now and you're like, oh no, well, we just use our version, right? Of whatever AI tool. No, people are using whatever version of AI that they feel comfortable with. And a lot of times many different versions, different models. And I think that if you are a decision maker at your company and you haven't already kind of turned that switch, it's time to turn that switch. Right? Work is AI first. The second way that AI is shaping how we work. And this is not fun to talk about, right? The traditional organizational structure is flattening fast. And I think this is happening in two ways. But the way that I want to talk about here is I think middle management over the next five years is really going to get flattened out. All right? And I'm not saying that to, you know, if, if you are in middle management or aspiring to be in middle management. I'm not saying that to dampen your career aspirations. That's not what I'm trying to do here. I'm looking at the stats and the facts and the writing on the wall and this is what ha. What's happening. So Gartner predicted that by this year 20% of organizations are going to use AI to flatten structure and eliminate half of middle management. And US employers right now are advertising 42% fewer middle management positions. Well, this is a stat from 2025 and I think by 2026 it's actually going to be worse. Right? So the number of middle management positions open, opening up are going down drastically. And research shows that organization, well, that's what they're using AI for. Because whether we want to admit this or not, right. I've been in middle management myself for many years. In some instances it's not needed. And it is kind of, it can be kind of this unneeded bureaucracy. And I think that a lot of the work that middle management does, it is synthesizing and personalizing communication two ways, right? From maybe frontline workers who are hands on keyboard doing the work, your entry level workers and then, you know, upper management, senior management, your, your, your C suite workers, right. A lot of times middle management is kind of organizing the chaos on both ends, right? But a lot of it is just sometimes project management and synthesizing and personalizing information or putting out Fires. And I think that as we see improved agentic traceability, observability with your everyday large language model systems, bringing in all of, you know, companies data into large language models where you can see everything happening before your eyes, I think eventually we are going to see a decreased need for middle management positions. Number three, yeah. People think that AI implementation is a tool problem, but it is a operations problem. And we're to talk about this more in a little bit. But 80% of the value that companies get from AI is just redesigning how they work. It's not choosing the right tool, it's not choosing the right model. It is changing the way that you work. And I think that that's one of the reasons why. And it's shocking to me. This is probably if I had to pick one thing in 2026 that I am most shocked about, right. Two of them would maybe be. One would be just the default agentic nature of today's models and how they can think and reason like a human and just their level of intelligence out of the box. But number two is just the adoption gap is mind boggling to me, right. That you still have large enterprise organizations that haven't fully leveraged AI, maybe because of the bureaucracy, because of the yellow tape, because of lack of training, because of lack of education. Right. But those who have and those who have not, that gap is so wide. And it's something that is continuing to surprise me every single day. But that is changing how we work. Number four, the bottlenecks, right? This is really impacting the flows. And I think that 2026 marks a tipping point where AI agents are significantly starting to outnumber humans in enterprise environments with exponentially more permissions. Right. I actually saw the piece of software that I've used there, the unicorn company called ClickUp. So they do project management. Fraudsters used to need fake documents or stolen credentials to scam a business. Now they just need a few clicks to get your CEO to say anything they want. Voice deepfakes rose more than 680% last year. And most fraud detection systems only flag suspicious transactions after the money is already gone. They're not actually listening to the call where the scam is happening. Modulate is. Modulate's new Velma model analyzes live conversations for the signals that give fraudsters away stress patterns that don't match the story they're telling. Urgency that sounds performed instead of real. Voices that are synthetic instead of human. All detected for your business in real time, not in a report days later. Modulate's Velma model was trained on 21 billion minutes of real audio and is trusted by Fortune 500 companies. It outperforms voice models from leading AI labs and it's 100 times more cost effective. So go see Velma catch what your current tools miss at Modulate AI. And their CEO said yesterday, I believe that they have more than 3200 AI agents working alongside 1300 humans. So humans are almost outnumbered two to one and I'm sure that that number will grow. Now, obviously they're selling an AI agent product, so you know, you would expect them to put that kind of messaging out there. But I've talked to many people, both on the record, on this show and off the record, who have confirmed as much that they have more AI agents working in their organization than humans. And a lot of times the only reason that that number of, you know, 2x the number of agents 5x 10x, it's the bottleneck of permissions. And that is actually becoming a huge roadblock and a huge area of focus. So yes, on the back end, you know, we have to talk about observability, traceability, all of those things, right? Expert driven loops. But on the front end, it's actually changing how we work because it's causing, I think, agentic progress to slow down. So the fifth thing that is changing how we work well, unstructured data like meetings is becoming a huge data pipeline. So Studies show that AI meeting transcription market is projected to grow up to $30 billion by 2034. About 10x of what it is, nearly 10x, about 9x of what it is right now. And well, you might be thinking, why? And shouldn't we be with AI, shouldn't we be doing fewer and fewer meetings? I've always said absolutely not. I think we actually need more meetings. But we need more meetings in a very smart and intelligent way. Because I think that a lot of the work that the majority of knowledge workers, so if you sit in front of a computer, you know, use the Internet and create business value, which is what most of us do. A lot of the work that we've been doing from, you know, 2020 to 2025. Now, AI agents, if they have the right scaffolding, can do way better than us, right? So you might be thinking then, okay, well then what do humans do? And unfortunately, well, fortunately, unfortunately, depending on how you view it, it is more human and social interaction. But I think what we have to do and think about how AI is changing how we work, it is recording and turning all of that into first party company gold, right? So I've been talking about this for a long time, I swear, I swear it's going to be a multi hundred billion dollar industry as soon as people figure it out, right? Someone please steal this idea, let me know how it goes, give me 1% equity or something like that, right? But transformer models, right, they, they really benefited from having rag retrieval augmented generation. But for the most part the how companies have fed these models, right, whether they built their own just you know, on the back end fine tuning models via, you know, OpenAI, Anthropic, Google's API, etc. Right, is they bring in their structured data, right? And that has helped these quote unquote old school transformer models be better and be useful and create business value. Where I think it's headed and this is a little longer because unstructured data is harder to monetize than structured data sometimes, right? But I think that's where we're headed. It is the company's decision making process and obviously we have the silver tsunami, right? We are losing just millions of baby boomers who have this institutional knowledge that some of it may die, right? So I think that a big focus on how AI is changing how we work is well, recording these meetings and transcribing and getting a better idea of the human expertise that separates your company from everyone else. Because people thought using LLMs would be a differentiator. It's not people thought, you know, connecting their structured data to LLMs is a differentiator. It may be for another year. But the real long term advantage is your first company, first party company reasoning in bringing that into large language models. Number six on our seven ways AI is reshaping work. Well, information is getting repackaged into answers. What do I mean by that? Well, knowledge workers, we're going to stop for the most part using the Internet. And I know that might sound weird, right? But I would say if we looked at, you know, 2025, maybe let's just say, let's say that you went from 90% Internet 10 chatbot in 2024 and then in 2025 you were 50. 50, right. I think it's gonna inverse. I think the smartest in most AI native people are gonna have only 10% on the Internet and 90% inside of large language models. To me there's not really a reason for the most part to use the Internet anymore. Right? Because just about anything can be done inside of these large language models. These front end AI operating systems, we're seeing them, you know, now Bring in data from every single, you know, website that you would use, right? You know, let alone the, you know, the model context protocol servers, being able to bring in anything. But even just by default, these direct integrations that work inside of Google Gemini, that work inside of Claude's anthropic or sorry, interface Claude, that work inside of OpenAI's ChatGPT, right? It's bringing all of your dynamic data to you so you don't have to move, right? So I do think that is a big thing that's going to change how we work. And then number seven, no surprise here, but context engineering is replacing prompt engineering. I think that thinking models do the basics of what 2020, 22, 2022's version of prompt engineering did, right? We did the, you know, the chain of thought prompting and, you know, think step by step, right? And all these prompt engineering techniques that I think, you know, individuals really leaned into heavily and thought, oh, this is the future of work and there's going to be all these prompt engineering roles. FYI, I never said that because I didn't believe it because I, I knew and I started to see the trend that, hey, once these models get smarter, what we've been doing in prompt engineering is no longer relevant, right? That's why even in our quote unquote prompt engineering course, before context engineering was a thing, we've been teaching it, right? We've been teaching it with our Refine Q. And if anyone's taken that, and you can take it for free, by the way, here, I'll give you the secret. If you haven't already, go to start here.com, all right? And you, or, sorry, the start here series.com start here series.com. you know, that takes you to our new Start Here series. But you can also get free access to our Prime Prompt Polish Prompt engineering course. But we've been teaching context engineering since before it was a thing. So in our Refine Q, go check out F and I of that Refine Q. It's an acronym and you'll see exactly what I mean. But Gardner is even telling leaders right now that prompt engineering is not enough. And context engineering is how teams get reliable results at scale. So, yeah, what the human does to get the most out of the model I think is going to become less and less important. I think a really clear illustration of this is like Mid journey, right? So if you ever used Mid Journey and AI image generator, you know, one of the most popular early on, it's like you almost had to speak a Mid Journey language to it and if you spoke in natural language, stuff just like didn't work right. And kind of prompt engineering has changed a little bit as well. If you talk to models from 2022 and 2023 in a certain way and you had a certain prompting technique, you've got way better results than everyone else. Now it's not the case because these models, well, they're smarter than us. All right, so now it's all about bringing your business context, your personal context before the model gets to work. All right, let's get into section two. So we went over the seven ways that AI is reshaping how we work. Now let's get into the 10 AI workflows that actually deliver ROI. And yes, I am going to be going a little fast through these. Don't worry what I'll probably do because I actually had a bunch of things that didn't make the list. But I really wanted to and I had to make some tough decisions. So if you repost this, I'll just share my entire notes file obviously in a very well put together interactive website. So if you want, you know, all the details, I'm not sharing everything that I have on screen and all the things that didn't make the cut. Just make sure to go repost this episode on LinkedIn and I'll send it all to you. So let's get to section two, the 10 AI workflows that actually deliver ROI. So let me just tell you a little secret about roi. Companies are getting it right. If you believed. Sorry, I'm going to be a little harsh here. If you believed that MIT piece of marketing that they called this study, that 95% of Gen AI pilots failed and didn't provide ROI, that means that you didn't take the time to read it. Right? Because that was based on 52 informal conversations, which is not an actual study. Right. It was a piece of marketing. They were selling something. So I think people have this wrong viewpoint when it comes to getting ROI on generative AI. But I'll tell you this, if you're using generative AI, you are 100% getting ROI. There's literally no way around it. That's like saying like if you take a cross country flight that you're not saving time than if you were to walk. All right? It's the same thing. A generative AI and large language models goes at the speed of hyper jets breaking the sound barrier versus walking or skipping on one foot backwards. So anyone out there that's like, oh well, we can't prove roi well, that means you have a human measurement problem more than anything else. All right, so all of these workflows I think are just dead simple. And I think one of the biggest mistakes that companies make is they try to over engineer their, their AI, their AI implementations, right? They, they try to make it this, this grandiose, you know, huge undertaking when it's like, no, like keep it simple, stupid. Right? Kiss. And I think, you know, the, the, the boring stuff, the unsexy stuff, that's where you're going to get the biggest roi. And the biggest roi, you're probably not going to know about it because employees are just pocketing it, right? I think especially in remote or work from home capacities. Right. Employees are sometimes completely automating their job. Right. We talked about the, you know, bring your own AI in 2023. I called it second computer AI. Yeah, this is rampant. Everyone's doing it. So that's where your ROI is getting. So I could almost guarantee and please don't do this, but if you were to actually like look over the shoulder or put cameras on all your employees or you know, use screen monitoring, please don't do that. That's terrible. But if companies actually did that and did it at scale for employees that have been properly trained on roi, they'd be straight up shocked at the amount that they're getting. All right, and to go, well, it is Tuesday, so maybe I'll go on a little rant here. I don't hate it. I don't hate that employees are pocketing their saved time because a lot of this too in larger organizations, it's corporate greed, right? You have these huge enterprise companies that are showing records, profits and they're still just laying people off in mass like they're, like they're not making money. So I can't necessarily blame the everyday employee that is using AI pocketing the save time, but that's where your ROI is, FYI. All right, I got that rant out of the way. It is Tuesday. Thanks for allowing me to do that. So let's get into 10 AI workflows that actually deliver ROI. All right, so number one, it's meeting discussion to tasks, right? This is huge. I think one example, if you've properly enabled Microsoft copilot in teams and can use fabric to bring in all of your data, that's enormous. Right? And studies have shown that, I think it's for every dollar that companies invest in Gen AI, if you measure it out correctly, they get $3.70 on a return. And I think one of the reasons is this going from meeting transcriptions to decisions, that is one of the easiest things to do. And you can, you know, whether you're using Google Meet Zoom, you know, teams, etc. If you have everything set up in permissions, this is usually automatically done. So this is, you know, I'm not saying this is replacing the traditional role of project managers, but it's really helping cut down the amount of time that people should be spending doing meeting, follow up, back and forth emails. Right? It's, it's archaic looking at it now, but that's the probably one of the biggest ROIS number two inbox triage. This is huge. So yes, make sure you talk with your company about, you know, proper data sensitivity, all that good stuff, right? But I mean ChatGPT, Claude especially I think those two and obviously Google, if you are a Gmail organization, right? But the major three as well as actually a Copilot, even Copilot Online, they have connectors for the other competitors, which I think is great, right? So as an example, even if you're using Copilot, the online version, you can connect your, you know, Google workspace. If you're using, you know, Chat GPT, you can connect your Outlook, you know, etc. But this is huge. This is one thing I use AI for not the most, but probably the most because I can't keep up with my inbox because I get spammed all the time. But I'm just like hey, what are the 10 emails? You know, having a task that's automatically triage triages my emails. Like hey, what are the 10 most important emails that I missed? The 10 most important emails I need to follow up on and then give me suggested replies based on the context that you know about me, right? Make sure that you share the right context with the large language models. But I mean this is huge. And in a study Microsoft 365 copilot users save an average of 30 minutes per week just on emails. For me it's way more than that. Especially I get on these like super long threads that are like 60 emails deep and I forget things because I have like 20 of those going. So I'm just constantly, if I'm being honest, I'm constantly using like Whisper Transcribe and talking to, you know, Claude or Chat GPT and just being like yo, like catch me up on this email. I'm lost. I already forgot what, what do they need from me? What do I need from them? Bullet point it, help me, you know, help me with the draft. I go in There, finesse it and send. All right, number three, 10 AI workflows that actually deliver ROI sales call to objections to tailored follow up. This is amazing, right? And I think this is. I've shared this multiple times on different of our AI at work on Wednesday shows, you know, creating different GPTs or projects that by default, with some custom instructions, you just dump transcript in there and then they are automatically gonna, you know, go through. Objection. Handling, you know, go through making a little, you know, project management piece of software or disposable daily dump of, you know, priorities based on a meeting transcript. Right. This is another big one. And I don't know, maybe in our community, in our free community, the, the inner circle, maybe I should start, you know, sharing these. So, yeah, let me know inner circle people. Let me know if I should. So that's number three. Number four, proposal and RFP first draft grounded in your documents. That's the important thing though. You know, you have to ground it in your source of truth. But I remember back in the day having to work on, you know, RFPs. Working in a nonprofit, they were so dang time consuming because you couldn't just use like a copy and paste draft, really, because so, so much of it had to be personalized and there's always so many different, you know, requirements. But ultimately it was using your company's knowledge in, in a modular fashion. And that's what large language models are great at. So whether you're doing proposals, RFPs, right. First drafts, that's a huge time savings. But again, make sure that you ground it using context engineering best practices in your data so you're not just getting a bunch of generic hallucinations. All right, number five, on the 10 AI workflows that actually deliver ROI research brief to executive memo on a weekly schedule. I like this one. Using different deep research tools. All right, I'm not going to go through, you know, how they work in each one, but you can use deep research, which is highly accurate, right? Because it usually takes anywhere from five to 30 minutes to go through and do one of these runs. But a lot of people overlook the fact that in Gemini, Claude and OpenAI's ChatGPT, you can do deep research just with your connected data. So just with your email inbox, just with your calendar, just with your, you know, drive storage, again, assuming that you have the permission to connect those things. That's huge, right? So if you have a huge meeting, a huge presentation, right, Maybe you do something quarterly. You know that you have multiple meetings with multiple teams and you have to put together, maybe something for internal stakeholders, external partners, et cetera. And it's a big part of what you do. Well, running these deep research on your document is a tremendous time saver. And talk about roi, right? And let alone the fact of just how answer engines now are replacing traditional browsing, even the ability to do deep research, but personalized browsing based on your context and your data, Again, just a straight up silly roi. All right, we got more. But before we do, take a very quick break for a word from our partners. Fraudsters used to need fake documents or stolen credentials to scam a business. Now they just need a few clicks to get your CEO to say anything they want. Voice deepfakes rose more than 680% last year. And most fraud detection systems only flag suspicious transactions after the money is already gone. They're not actually listening to the call where the scam is happening. Modulate is. Modulate's new Velma model analyzes live conversations for the signals that give fraudsters away. Stress patterns that don't match the story they're telling. Urgency that sounds performed instead of real, voices that are synthetic instead of human. All detected for your business in real time, not in a report days later. Modulate's Velma model was trained on 21 billion hours of real audio and is trusted by Fortune 500 companies. It outperforms voice models from leading AI labs, and it's a hundred times more cost effective. So go see Velma catch what your current tools miss@modulate AI. All right, let's keep it rolling. So in the 10 AI workflows that actually deliver ROI, number six, customer support agent assist. So to draft, categorize, and go on to your next step. So as an example, you know, using Microsoft copilot to, you know, go into your knowledge base, get recommendations on how to reply on customer support. That's an easy one, right? I've helped companies in the past right before generative AI, but you know, to work with, you know, kind of online chatbots and, you know, having customer support have to go through and manually look in their knowledge base and find the right answer and customize it. It's something that AI is honestly just a lot faster at right now. Danfoss actually said that they automated 80% of transactional decisions with AI taking the response time from 42 hours to near real time. So there you go. That should tell you everything you need to know. All right, number seven, finance variance explanations that are board ready. So this is a great example. As for another use case. So having executive Teams, you know, increasingly want the narrative layer, not raw tables. So AI can accelerate kind of that first draft significantly. So organizations report 50 to 70% reduction in data preparation efforts after they implement AI assisted engineering platforms. So being able to create narratives out of a lot of boring structured data like finance reports and being able to present that in a digestible way to executive teams is huge. Right. There's obviously people whose entire job this is, right. They're working with the finance department and they need to get, you know, marketing on board and they need to sell this to the C suite, et cetera. Right. So being able to use AI and lean into AI for that is a huge roi. Number eight, policy and compliance Q and A with internal only answers. So this is something I think is great. That's great for something like Clog Projects, ChatGPT Projects, or even Notebook LM, right? Because when you're talking about policy and compliance, you don't always need the creativity and the brainstorming prowess of a, you know, of a Claude Opez 4.5 or a Gemini 3 Pro as an example. Maybe you can just use something that is rooted in your organization's data, like Notebook lm, especially when you're talking about policy and compliance. But this directly reduces hallucination risk by forcing your answers to stay inside approved documents only. This is always one of those kind of demos whenever I go out, you know, every once in a while I'll go do some, some keynotes and some in person workshops when, when I have a little bit of time, which isn't a lot, this is one that always really shocks people. Right. You know, I'll get copies of their old compliance policies and then to just be able to show people that, oh, you can ask questions of a thousand pages of documents at once and get cited sourced answers that aren't hallucinated. You know, especially HR departments, legal departments, you know, it's to, to, to see the look on people's faces. It's, it's kind of wild. All right, number nine, content repurposing. You already know this, right? But this is great and something I practice all the time, right. So as soon as I'm done with this podcast, you know, someone from our team's gonna upload this into a program and it's automatically gonna transcribe the podcast and it can spit it out in hundreds of different formats. Right. We have prepackaged prompts that this goes to and then that can send it via Zapier to anything. Right. So if I really wanted to, I could make a Hundred videos, a hundred websites, a hundred different, you know, graphical animations just from this, without really doing anything, without putting any work in. Right. So content repurposing it is so good. Now it is on autopilot, I think, especially since we've seen models like Google Gemini's Nano Banana 2.5, Nano Banana 2, Nano Banana Pro 2, ChatGPT's GPT Image 1.5. Right. And then being able to work with those systems on the API and just having it do all automatically. Right. A lot of my previous career was kind of manual content repurposing and where we're at now, I mean, it's. You can't even really measure the return on investment because the capabilities are kind of endless. All right, and then number 10, schedulized personalized, scheduled, personalized research. This is one that I use all the time, like chat GPT tasks. I have a handful of tasks that get run every day, but they're usually just personalized research. And then I make sure that each day it's not looking at the previous day's information. So if you are someone that has to constantly stay abreast of industry movements, competitors, you know, just news that impacts, Right. If you're in finance, if you're in spaces that change often, I don't know, crisis communication. Right. There's so many different spaces that change often. Using scheduled tasks to go out and personalize research and personalize report for you is huge. Right. And this is separate than, you know, triaging your email and all those things. This is just synthesizing and personalizing up to date information by default. All right, trying to go fast here. All right, I was promising myself, I'm like, all right, even though I'm going to go 7, 10, 10, this isn't going to be a 50 minute podcast. All right, so we're going to try to go through these next 10 in 10 minutes. Let's see if I can do it. All right, so let's talk section three. And this is the 10 AI skills and every professional needs now. All right, I'm cutting it to you straight again. This isn't. These aren't just hot takes for me. These are through conversations with hundreds of really smart people that are putting this to work in enterprises, small businesses, AI startups, etc. But the AI skill gap that I'm talking about here is not what you think. Because if you think about, oh, AI skills that I need to know, oh, I need to learn, you know, I don't. I need to learn JSON. I need to learn Python. No, you don't. All right, let's talk about the 10 AI skills that are going to keep you 2026 ready? Number one, change leadership. Huge old winning playbooks are expiring quickly. Companies are finding this out the hard way. Those that especially were slow to adopt to AI, they said, well, we've been profitable year over year, right. We're, we're leading our category. We're crushing competitors in 2024, 2025. We don't need to adopt to AI. Yeah, those companies are going to slowly lose their spot atop the food chain. I think teams with the biggest wins in 2026 are those that have already thrown away successful playbooks. And I think this has to do with just change management and looking at work completely differently. All right, I could talk about this for hours, but I'm not. But I think that, right. Especially if you're mid career right now, we've been able to hang our hat on this, you know, input, output equilibrium, right? If we, if we input hard work, effort, industry top notch skill sets, we're gonna, on the output, we're gonna receive something the same way. It's still the, the truth. But what you have to slide in there is AI native, which is hard, right. Because it's changing every single day. But you can't do what you were doing 10 years ago every single day and expect it to pay off. It's not. Because again, 10 years ago, as an example, let me do the math. Yeah, I was working marketing at a nonprofit, right. Working a lot with Nike and Jordan brand and doing these activations. But a lot of what I was doing is content repurposing. If I was doing the same, even though I was putting out just insane day to day effort, high quality work, if I was doing it the exact same today as I was 10 years ago, I'm getting lapped. I'm not even on the track, right. So you have to completely throw away successful playbooks. The second skill is the ability to not trust AI at all and to go through the proper verification process. So a new Gartner study predicted that 50% of organizations will require AI free skills assessments by this year due to critical thinking atrophy. Right. I, I, I struggle with this all the time. If I'm being, if I'm being honest, this is why I read chain of thought. So like chain of thought summaries all day because I need it to stay on top of my skill sets. I think sometimes the more that we hand things off blindly to AI and we don't verify or trust anything. Not only does that increase the risk of hallucinations, right, Which I think are becoming less and less of a factor as the models get better, but what it actually does is the. The atrophy, right? Just our. Our human abilities and our human skill sets. If we're not practicing them, if we're not riding the bike on a daily basis, even if the bike looks different, the bike is actually, you know, a jetpack. Well, you still got to ride it, right, to get your repetitions in. All right? Skill set number three that you need is you have to be fluent in. In multi. Multimodality. And I'm not saying, oh, you need to understand, you know, text and image in chat gbt. No, I'm saying you need to understand multiple models, right? You need to practice kind of the concept of being modular, right? And being able to modularly solve your company's problems. And this is why instantly, if I ever see anyone on social media or otherwise, right, say, oh, I canceled my, you know, Gemini subscription and I'm only using Chat GPT. Or I'm. I cancel my Chat GPT subscription and I'm only using Claude. All right? If I'm being honest, those people are probably not gonna make it. And here's why. If you really want to see if someone knows what they're talking about, ask them what's the best model? And if they have a very simple and very definitive. If they're talking, like, in black and white, and they're like, oh, Gemini, the best, hands down, right? Claude chatgpt garbage. No, right? The. The answer is very hard. If you ask me, and if you say, tell me the whole truth, I'm like, well, you better have four hours, because I'm going to tell you the pros and the cons of every single. Every single different type of work. I'm going to tell you the difference in, you know, GPT5.2 Pro versus, you know, GPT5.2 thinking. And how sometimes I still use GPT 4.1. And I'm going to tell you a little bit on the. On the pros and the cons of using Gemini 3 Pro in AI Studio versus using it on the Gemini chat versus using it in the business version of Gemini, right? You have to understand the different models and the pros and the cons. It's no longer, you know, something where it's like, okay, I'm fluent in one, you know, in one model or in one system, right? For the most part, even though I do think, yes, organizations need to Pick an AI operating system of choice. I've said that, but. And you need to move all your day to day processes in there, but that doesn't mean that you should exclusively be using those. Right. Maybe 80% of the time, but the other 20% of the time a different model is probably going to be 2x3x faster and 2x3x better. All right? So getting back to ROI and you have to be able to measure what matters. You got to, you got to know those things. There is not one model. I don't care. Gemini 3 Pro GPT5.2 Pro Opus 4.5. There's no one model that is the best in everything. All right? So if your organization, or even if you personally are wearing a lot of different hats, it's not, it's not always the best practice to just stick with one model. All right? The fourth skill set that you need is process thinking. So redesigning your workflows is always going to be obsessing over AI tools. Another kind of common mistake that I see people make is straight up obsessing over all the different AI tools. It's something I don't do. Right? Yeah, we put the top, you know, three AI tools of the day in the newsletter, but I don't use them. Right. Like, people are always shocked at. Yes, I've, I have used thousands of AI tools over the past four or five years and I continue to try out new ones that are, you know, that are trending, but I'm not making them part of my day to day workflow for the most part. You know, it's certain things. Like I said, I'm working modularly as well. But you know, let's just say I do 80% of, you know, my day to day work inside of one of the, the AI models and then the, you know, 10 in the second, 10% in the third. And I'm not using all these other tools. Right. A lot of people are like, oh, I, I use these 15, you know, tools just for AI writing. And I'm like, okay, don't. Right. Like the amount of duct tape that shiny, shiny AI object syndrome creates takes away from your roi. Right. So like, if I'm being honest, you have to try to ignore a lot of those shiny tools, even though they look really cool and say, no, what are those boring, time consuming things that we can do right now in our AI operating system? Number five, context engineering. Right. Give AI what it needs to be. Right. Like I said, the, the combination of the models getting better and the ability to kind of have like one click mini rag in these tools. Hallucinations are, I'm not saying that they're a thing of the past, but they're not really a huge concern. And it makes it so easy, especially using things like projects in ChatGPT or Claude, Google Gems, etc. Right? It makes it so easy to bring your company's context with these connectors, with these apps. It's a couple clicks. Yes. You still have to verify outputs. You have to understand the chain of thought. You have to be able to scope and test and measure, right? But once you do those things and you do them continually, right? So you have to, that's an ongoing process. But after that it's context engineering. It's making sure the model has the right information and then using the right model in the right mode. All right? The next skill, writing crisp inputs. Yes. Even the best of context engineering, the best models, if you're giving just half hearted inputs, your responses aren't going to be that good, right? I always say use more words. The, the tired old example that I use all the time, right? Right now, large language models don't understand words, right? Even though I communicate or you communicate or we communicate with large language models with words, they don't know them, they convert our words into tokens, then they think and produce in tokens that is converted back to words. And the dumbest example that I give is there's seven different ways the single, the single word just J, U S T. There's seven different ways that that can be tokenized or seven different meanings that a large language model might look at that word. So you have to understand, just like when you were learning to read or maybe learning a new language, how important context is to a single word. Now think of, you know, these things that are able to process thousands of tokens per second. Think of how important the surrounding context in your words, let alone your, your context engineering, but just your words. You have to be extremely clear. And this is why I read chain of thought so often. Because I catch, I catch it all the time myself, falling short and I'm like, oh, I, I use this three word ph days and I probably should have written out two full sentences, right? To explain that a little better. That's another reason why I'm using a lot more voice dictation now because it's, well, it's faster. But sometimes I can explain things a little bit better speaking them than I would if I type them. Because if I type them, sometimes I overthink and oversimplify. Because as A former journalist, I'm used to writing tight, right. And cutting the fat. So just FYI, writing crisp inputs is a huge and one of our 10 AI skill sets that every professional needs now. Number seven, continuous learning and adaptability. You know, it's, it's sometimes called change fitness. So change fitness is now a career requirement. It used to be that you could really master a skill set and ride it out for 5, 10, 15, 20 years. Right. I think for, for the most part, right. Post, I think post Internet maybe changed it slowly and, but it probably took a decade for that to reveal itself. The same thing with social media. So yes, that's been impacted, but I'll say for the most part, I don't know, since the early 2000s, you've been able to become really good at one skill set. You know, kind of keep up, you know, a little bit of polish, but for the most part, you've had people make entire careers of a quarter century just being really good at one thing, right? It's not going to be the case moving in the future because large language models are going to be really good at that one thing and we are going to get domain specific models. So it's no longer, oh, I know something, right? No, it's how you can apply AI to that thing that you know and use it in the right way. But you have to be able to shift because right now the tech is shifting faster than teams can adopt, right? There's literally. I do this every day. I can't keep up, right? I'm a, I'm a small, I'm a small company, small business. I do this every day. I can't keep up because I know even since I started recording this podcast, I can guarantee there's some new tech, new technique that has just been released. All right, Number eight, the automation basics, you have to understand it, including scheduling tasks. That's a huge one that I think most people overlook for whatever reason. You know, Claude Gemini and OpenAI just kind of hide them, hide them. Well, actually it's not fully rolled out in Google Gemini yet. But you know, scheduling tasks is huge. And this is where I think, you know, it's almost like agent creep, right? Like all of a sudden had a great conversation with one of the, the head of AI at Cloudflare and we talked about this. We're like, okay, well if you're scheduling a task and it's an agentic model, technically it's an agent, right? An agent is going out to do work for you without you even telling it. And I think that you, you have to shift that from saying like, hey, I use AI when I remember to use AI versus AI just runs constantly, always using the updated and most dynamic data. And it's, you know, memory and personalization of our ongoing conversations. Number nine, the ninth skill set, human AI collaboration. You have to work with it, not around it. All right, this one might be uncomfortable, but I think you need to treat AI like a junior teammate you manage that's trying to outwork you and take your next promotion, right? Why do I say that? Well, that's what's happening, right? And I think it does require a little bit more work than traditionally you've had to put in. Because like I said, again, normally you, you, you dig down, you dig deep on that one skill set, right? Oh, I'm a great copywriter, right? So I'm going to dig deeper, deeper, deeper, deep. No, not anymore. Now, if you're a great copywriter, you have to be good at content repurposing. You have to be good at using AI video tools. You have to be good at using, you know, all of these different things. You can't just dig, dig deeper anymore because if you keep digging and spend your whole career, you'll find at the very bottom of you digging down. Oh, that's where all the AI is. They're actually better and they've been down there. So you have to be able to know when it's time to collaborate with an and not compete with it. And that time is now. Because human AI collaborative teams, studies show, demonstrate 60% greater productivity than human only teams in research studies. And again, I think you have to think of augmented intelligence. I think right now, most people, their use of AI is just not doing the thing that they should be doing. So let me use an example. Let's say your data analyst, right? And if you're using AI to analyze your data and you're just kind of copying and pasting everything. Okay, how long? Right again, talking about the skill atrophy. How long until you start to lose some of those data analysis skills? That's why you have to use augmented intelligence. That's why you have to combine the best of you, the human with the best of the AI and push each other to make each other better. So again, if you've taken our free prime prompt polish course, you understand that, all right? And then skill number 10, communication. That drives decisions, not just outputs. This is huge. I talk about this all the time. One of the best skill sets, I think is not leaving gold at the bottom of Your chat when you're done, right? So whether you're using again. Gemini Copilot Claude chatgpt It doesn't matter what is the action, what is the decision, what is the output, right? I have to remind myself of this all the time because I am guilty of this as well. I leave so much gold at the bottom of there, right? It's actually one thing I'm having agents go in and, you know, double check the bottom, you know, reread my chats and, you know, saying like, hey, let's, let's fly these. Let's make a database of things I need to follow up on, or database of decisions with links, right? You have to create next steps in actionable value. Because, you know, I think a lot of times, not that I'm caring about, like wasting tokens, but I think it was Satya Nadella that said we have to stop, you know, shift from just using tokens to creating value. And I think that's this last skill set, and maybe a good one to leave people with, is we have such an unbelievable technology that is crazily affordable, it is insanely capable, and sometimes we're just looking for that one little thing out of there. But what about everything else, right? Sometimes I look back and I'm like, okay, I. I use this entire chat just to make one decision, but look at everything I left on the table. So I think really developing that, that communication skill that drives decisions and not just outputs. All right, so here's what matters, and I'm gonna leave you with this 8020 rule flipped on its head a little bit right now. I think using the right technology is what is gonna give you or your organization 20% of the. Of the value. But redesigning how you work, rebuilding, unlearning, that's the 80%. And I think the winners in 2026 are focusing more of their time on that, more of their time on change management, more of rethinking about the future of work and unlearning and starting from scratch. And there's nothing wrong with that. That's what we do here on this show every day. So if you are doing that, you're ahead. So I hope that this episode was helpful as we went over the seven ways AI is reshaping how we work, the 10 AI workflows that actually deliver ROI and 10 AI skills every professional needs in 2026. Like I said, if this was helpful, go ahead, repost this. I'll send you my complete notes in a nice little interactive canvas document that hopefully can be helpful for you as well. And make sure if you haven't already, go to our website@your everydayai.com we're going to be recapping the highlights from today's show as well as a lot more. So thanks for tuning in. Hope to see you back tomorrow and every day for more Everyday AI. Thanks y'. All. The risk with AI Voice agents isn't that they sound too robotic for your company to use. The real risk is that they can sound too confident while saying something completely wrong to your prospective clients or customers, made up refund policies, promises your company never approved, or discounts that don't even exist. You've got to give your AI Voice agents a trust layer with Modulate. Modulate monitors live voice conversations to flag abuse, false claims, fraud, and user emotions for safer, more empathetic responses. For the guardrail layer you need between your AI agents and your customers, you need Modulate at Modulate AI. And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit youreverydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.