Everyday AI Podcast – An AI and ChatGPT Podcast - EP 443: The Complete AI Business Revolution: Agents, Wages, and What's Next (2025 AI Predictions - Vol. 1)

Episode Date: January 20, 2025

You know those common good habits and best practices that have worked since the dawn of the digital age?  Kiss all that goodbye in 2025 y’all. Your regularly scheduled reality is evolving. AI Agents aren't just changing HOW we work - they're changing who we ARE and what work even means.  I’ve spent thousands of hours talking about Generative AI, interviewing hundreds of the world’s AI leaders at big companies like Microsoft, NVIDIA, Google, OpenAI, IBM and others.  This week — I’m sharing what’s REALLY happening in our first-ever Podcast Series: 2025 AI Predictions: The Roadmap ahead.  We kick off Monday with Vol 1 — Agents. Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion: Ask Jordan questions on AIUpcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:1. AI Agents in 20252. AI and Businesses in 20253. Discussion on AI Job Market4. Universal Basic Income and AIKeywords:AI agents, professional services, AGI, UBI, AI predictions for 2025, Everyday AI, artificial intelligence, Jordan Wilson, companies, AI education, Microsoft Copilot Studios, Salesforce agent force, AI economy, public companies, AI jobs, agent orchestrators, unstructured data, structured data, reasoning models, AI adoption, AI-based companies, OpenAI, Microsoft, Google, IBM, large language models, generative AI, agent systems, AI reasoning, AI technology.Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Start Here ▶️Not sure where to start when it comes to AI? Start with our Start Here Series. You can listen to the first drop -- Episode 691 -- or get free access to our Inner Cricle community and all episodes: StartHereSeries.com Also, here's a link to the entire series on a Spotify playlist. 

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. Meet Firefly AI Assistant, now live and Adobe Firefly, the All In One Creative AI Studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. Will AI agents be more than a buzzword in 2025?

Will AI drive the price of professional services down? And will AGI lead to an inevitable introduction of UBI? We're 20 seconds in and we already have alphabet soup. Well, I'm going to be talking about those things in a whole lot more. on this first episode of a special edition of Everyday AI. This week for five straight days, I'm going to be bringing you five at a time, but the 25 biggest and boldest AI predictions for 2025.

All right, I hope you're excited to tackle those questions and a lot more with me on Everyday AI. So what's going on, y'all? My name's Jordan Wilson. I'm the host of Everyday AI. This thing's for you. It is a daily live stream, podcast, and free daily newsletter, helping everyday people like you and me, not just learn AI,

but how we can leverage it to grow our companies and our careers. If that sounds like you, you are in for a treat. I have been working on this series technically for like 11 months now. So this is going to be something if you want to understand where the future of AI lies and how that's going to impact your department, your job, your small business, the enterprise company that you work for, you are definitely going to want to tune in today and the rest of the week. You're also going to want to go to your everyday AI.com.

While you're there, you can sign up for our free daily newsletter. Also, there's hundreds, that's hundreds, hundreds of hours. of free AI education interviews, YouTube videos, everything is on the website. It's all there for free by category. All right. Enough. Let's get straight into it. And well, let me talk about why we're doing this AI prediction show.

Well, I'm lucky enough, I do this every day. You know, everyday AI. It's not just a saying. This is what I do every day. I've talked to the smartest people in the world. NAI, from the leaders at, you know, companies like Microsoft, Google, IBM, to fast-growing startups, right? I've had hundreds of guests on the show this year, and I've had another hundred or more conversations, kind of quote-unquote, offline. And this series is a, it is a, not a

celebration, but it's almost just a result of those thousands of hours of conversations that I've had. But I always think to myself late at night, I have so, so much information floating around in my brain from messages I receive from people at these big companies, from, you know, old podcast guests as I go back and listen to them, some super smart people. And I start to connect the dots, right? Yeah. Eventually I do it. until in my brain is all AI one day, right? I stumble upon things and throughout the year I write these things down. I text myself.

I put little like sticky notes. And for the past two months, I've been compiling all these. So like I said, over the next five days, I'm going to be bringing you five AI predictions each day for a total of 25 predictions for 2025. All right. And I did this last year as well. And the reason I'm doing it again is because I had people reaching out to me, even in like October, who were like, yo, I just listened to your show.

And the fact that you said these things like in December or January of last year, insane, right? So yeah, I went back. I even did a midterm report card on my original bold predictions for 2024. And I think I got like 21 of them right. And the other three were just like TBD, right? Or couldn't really give myself a grade on those. So without further ado, let's get into it. This category that we're going to be going over today, these five predictions,

we're calling it the complete AI business revolution, agents, wages, and what's next? All right, I'm not going to make you wait any longer for the five predictions that we're going to be going over today. So number 25, agent orchestrators will be a growing position. 24, public companies will post jobs for AI agents. 23, company reasoning data collection. 22, high-end professional services will go through a pricing crisis. And 21, UBI becomes a household conversation. All right.

Hey, live stream audience, technically, this is a pre-recorded one. If you still want your daily AI news, that's going to be in the newsletter. But tell me right now, out of these numbers on the screen, which one? is going to be the most likely to happen in 2025, and which one is going to be the least likely. I'm always curious, you can just put most, you know, if you think agent orchestrators is going to be the most likely, just put most 25, least 23, right? I want to hear from you guys. I want you guys to fact check me and see how crazy some of these are or how realistic they are. All right, let's start with agent operators.

All right. Here's the thing, y'all. Yeah, we already know. 2025, the year of the AI agents. Yeah, we know. I'm not going to do y'all like that. That's not going to be the first prediction I give you. Everyone knows.

My mom knows. 2025, the year of the agents, whatever, right? We all know that. It's already here, right? It's already here. And we've already seen it with Microsoft co-pilot studios, autonomous agents. We've seen it with Salesforce, their agent force. autonomous agents, we're seeing it with open AIs, and we'll be seeing it with open AIs operator,

their agents. We're seeing it and we'll see it as it continues to roll out with Google's Project Mariner. So this has already started to happen. So I'm not going to make a kind of a weak AI prediction to start off with and be like, AI agents are going to take over. No, agent orchestrators. I'm just going to call them AOs, right? This is going to be a thing.

All right. Recent studies showed that 75% of companies plan to adopt AI agent technologies in the next five years. So it's kind of flipping this dynamic of how we've been working with generative AI in large language models on its head over the last two or three years, right? How this usually works is you get a team of people, you know, and this is how larger enterprise organizations generally start to roll out, you know, AI or large language models. It starts with a few rogue individuals who are maybe using it and they're not supposed to be using it, right? And then it goes to a broader team.

From there, it usually goes to a small team. And then you have a small team usually working with one AI system, right? And they're collaborating and saying, oh, how did you do this? Hey, we're getting hallucinations here. you know, should we, you know, should we look at an external rag system? Should we be using, you know, built in, you know, projects from chat GPT or Claude, whatever, right? But it's usually a group of people working around one AI system or large language model. We're going to see that flipped.

Not only are we going to see that flipped, but we are going to see public companies actually hiring or having positions. open for agent orchestrators. There might be a different name. I'm going to call them AOs. Hopefully that sticks, right? Last year, a lot of people started using a term that I threw out there. It was called like second computer AI or something like that, right? But I think we're going to see agent operators or sorry, agent orchestrators, where one human,

their only job is going to be orchestrating. probably it's going to start with a dozen or so, but eventually thousands, right? I think of it like air traffic controllers, right? Think of that. You have a group of, I don't know, in the movies, right? You see it's like, oh, five to ten people in this room, but they are orchestrating hundreds of flights at the same time, right? Like I'm in Chicago, we have O'Hare Airport, Midway Airport, and I assume there's literally

people up there in the towers, right, like there are in the movies. And they're watching the skies and making sure all of these airplanes, which are largely autonomous, are doing their jobs, right? Humans overseeing autonomous technology to make sure everything is going right. They're orchestrating air traffic. That's what I think we're going to see in actual position. And it's going to be a growing and trending one, right? Like three years ago, it was, you know, prompt engineers, right?

That kind of died off. But I think we're going to actually see this. So in early pilots, they said that there was about roughly seven hours saved per week per knowledge worker who was using multiple agents. This is, you know, it's all unofficial. It's all early. All this, you know, multi-agentic kind of workflow in scenarios is brand new. All right, we're moving on in the next one. I'm going to try to do this all five in less than 25 minutes.

So let's see if I can actually do this. Y'all know I'm long-winded, but that's why I'm breaking these 25 predictions up into five different shows coming back for you all week. All right. So next, 24, public companies will post jobs for AI agents. Yes, these two are sort of related. So data is already showing, right, 65% of organizations surveyed in the U.S. are already using gender of AI, which is kind of baffling to me because it's like, are the other 35% lying to themselves or are they just going to be going out of business soon, right?

Imagine if there was a survey right now that said 65% of companies are using the internet, right? So if you are in that minority of companies that aren't using generative AI, that's weird. anyways, I think we're in the same way we kind of transitioned out of lockdown COVID into hybrid workforce, right? I think employers saw like, hey, to retain, you know, our top employees after going work from home for an extended period of time, we need to have a hybrid workforce. And we're going to have that with agents, right?

But literally, I see in 2025, and maybe they do it just to, you know, make a splash, right? Like how public companies, there was all these studies, like the more time you threw in AI buzzwords into your earnings call. Adobe just introduced an entirely new way to create, bringing the power and precision of its creative suite into one conversational experience. Meet Firefly AI assistant. Now live in the Adobe Firefly app, the all in one creative AI. Studio. Powered by Adobe's creative agent, Firefly AI assistant lets you start with your vision, just describe what you want, and shape the outcome as it takes form with the assistant. The assistant orchestrates multi-step workflows, drawing on 60 plus pro-grade tools across

Adobe Creative Cloud apps, including Photoshop, Illustrator Premiere, Lightroom Express, and more to help bring your ideas to life. You can also get started with creative skills, a growing library of pre-built workflows for common creative tasks, like batch editing photos, creating mood boards, portrait retouching, and creating social variations. Every step the assistant takes is visible so you can refine, redirect, or take over at any time. You stay in the driver's seat as the creative director. Adobe Firefly AI assistant now in public beta.

See it today at firefly.adopi.com. Like there's actually a study that said like your stock price went up higher, right? because, you know, investors love to hear, you know, rag, large language model, agentic, right? But now it's going to be, I do see at least one public company, a lot of enterprise companies, but I do see one public company posting jobs for agents. It is going to be, this is not a job for humans. Humans need not apply. This is for agents, right?

It might sound weird, but I'm not going to come to you with week. predictions, right? Also, analysts expect a 20 to 25% cost reduction using more agentic roles. All right. What do you guys think? It's using agents, it's inevitable, right? Especially if you are a knowledge worker. If you listen to the show, you probably know what that means. But a knowledge worker is anyone that sits in front of their computer, right, bangs away on their keyboard, clicks on their mouse and they get paid to use their brain, stare at information on the computer, process something, and perform an action, right? Those, when we talk about autonomous agents, what that means, it's, I mean, it lives in different formats, right? Sometimes it's pre-programmed,

and we have more narrow applications. And sometimes we have agents that are more programmable in general and can do a lot of different tasks across a variety of applications. But I do think that we are going to see these multi-agent systems not only being common. And like the first prediction, companies hiring for people that will just orchestrate agents, but I do see companies hiring just agents, right? And right now your team, you know, you might have five of you in Chicago, two on the East Coast, in the West Coast, whatever, right? Probably you're going to have literally dedicated agents that your company is paying, right?

Either by the hour, one-time fees, by the task completion. I don't really know. I don't think Salesforce's kind of approach with charging like, I think it's like $2 per conversation for agent force. I don't think charging like per in action is going to work out necessarily well. I think charging per results. But I think that's going to be a huge potential economy in 2025, 2025, 26, and beyond. Right.

Think of something like, you know, Upwork or think of something like Fiverr, right, where you can go on and there's all these different people offering services, right? I already know there's people on those that are doing that all autonomously anyways, right? But I think there's going to be a thriving marketplace for this, right? So whether one of the big companies want to do that or not, we'll see, right? There's already things like the GPT store. So I do see in the future, you know, OpenAI, Microsoft, Google, you know, meta, etc. having agentic stores, right, where people essentially, you know, if you build like an

agentic workflow in Zapier as an example, something like that, you know, I don't know how credentialing will work, right? If you have to give it access to all of your data. So maybe it's something that can only live in something like a Microsoft 365 copilot or something that can live in a chat GPT enterprise setup. And, you know, maybe there's going to be a central file management system where anything that you upload into chat GPT, you know, any agents in this kind of agent ecosystem can access all of that information. So we'll see how. We'll see how. what actually plays out, but that's a big one. I see public companies doing that. Speaking of companies making moves, this is one. I see company reasoning data collection.

Here's what that means. I think some of the companies that benefited most from the generative AI boom, right, even though a lot of people think it started with, you know, chat GPT in November 2022. It started a little bit sooner. I think once Open AI largely made its GPT technology available to many companies in early, I think it was actually late 2020. So about two years, two years before ChetGPT. I think a lot of companies were smart and they started very early on turning all of their unstructured data into structured data or at least collecting all of their unstructured data. data, right? So, you know, emails, meetings, speeches, keynotes, right? Like all these different things. So then as large language models got more capable and as we see reasoning models as well,

they were ready. They had all their data ready and they started converting, right, unstructured data into structured data, right? Turning natural conversations, emails, right? like using large language models to also turn that into structured data, right? large language models do really, really well with structured data. If you know what you're doing, if you know the basics of prompt engineering, you can really squeeze a lot of business value out of unstructured data as well as structured. But I see for prediction, technically the third one, number 23 in our list, as we're counting them down, reasoning data collection.

Okay. So what that means, we've already seen a pretty early on explosion of reasoners, right? So Open AI was the first with 01 preview 01 Mini. Now we have 01 the full model and 01 Pro. Love an 01 Pro. By the way, I use it all the time. But we also have now Gemini 2 Flash Thinking with Amazon Nova. They have a reasoning model.

So essentially, if you're lost and you're like, what the heck is going on, think of large language models. I'm going to oversimplify this. So, you know, if you're a machine learning researcher or if you have your PhD and AI, just allow me to, you know, a little leeway here. Think of large language models, classify them in two different categories. So you have your quote unquote traditional large language models, your transformers. So think of, you know, GPT-40, think of Claw, 35 Sonnet. Think of Gemini 2 or whatever it is, 12-0-226 experimental, right, whatever they call it nowadays. right. So you have your kind of GPT or your transformer variety, and then you have your reasoners, right?

Reasoners is going to be another big buzzword of 2025, so get used to it, a multi-agentic reasoners, right? Maybe we'll stop talking about RAG as much in 2025, right? But you have these reasoning models that essentially do the work. If you had the world's best prompt engineers, right, they're going to spend a lot of time going back and forth and quote unquote prompt engineering. a transformer model to get the most out of it. So reasoning models, they kind of do this by themselves. They kind of think logically behind the scenes. You know, people think it's, oh, level one, level two thinking or level one, level two models, right?

Your standard large language model will spit out thousands of words of text in like three seconds. Reasoning models can sometimes take three minutes before they respond or, you know, I've had it think for 10 minutes, right? So it's different. It can solve complex problems. But I think companies are going to start collecting their decision-making process. So right now, there's no way to tune these reasoning models saying, hey, you know, 01 or maybe when we see, if we see 03, here's how you should think.

Here's how our company thinks. So the same way that companies can fine tune and put in rag in traditional, quote unquote, level one, GPT transformer models, I see an emerging market for doing the same with reasoners, right? Not just fine-tuning them, but teaching them how to think. But before that can become a reality and thinking of like wrapping, like rapper companies, you know, and I always, I always take poohs on companies that are still out there that are just GPT wrappers and really nothing else, right? It's like, oh, here's essentially chat GPT, but in a different interface and you pay, you know,

$50 a month. That's dumb. But I see actually a big market for reasoning wrappers as well. But also, the data collection side is going to be huge. All right. Two more on today's episode of our first five. Here we go. 22.

Live stream audience. Let me know. Most likely, least likely. What do you think of these so far? Are these hot enough? Tomorrow's Tuesday. So maybe I'll crank up the heat. All right.

But high-end performance. professional services will go through a pricing crisis. Y'all aren't going to believe me on this. Buckle up here for a second. Let me take a sit. All right. All right. I didn't put anything on my predictions about big job loss.

Did that last year. We saw it. I don't want to focus on the doom and gloom. All right. Here's what I will say. There's like there was in 2024. I think we're going to see more job loss due to AI than we did in, sorry, we will see more in 2025 than we did in 2024. Here's what happens. It's going to start

to hit the professional services, right? Because I think, you know, when we talk about lawyers, accountants, right? So many of these higher tier professional services, those that adopted early to generative AI, right? I'm saying like 2020, 2020, 2021. on. They're already starting to squash the medium-sized, right? I think smaller, let's just say lawyers. It's easy, right? But think of any, you know, management consulting, IT consulting, right?

Anything that is usually a high-price service that requires, you know, expertise in high-price knowledge, right? Those that adopted to AI early, they're doing very well. I think smaller companies can do very well. They have less overhead and they can innovate a little more quickly. I think of a lot of the middle players have already started to get squashed. We saw a lot of acquisitions in various sectors. We saw some just fade into oblivion, right?

So here's what I think is going to happen. A lot of people are going to lose their jobs, you know, big companies that would generally hire a lot, they're just going to stop hiring, right? Kind of the Clarnah approach, you know, companies, big law firms that might normally hire, you know, 100 employees every single year for the last decades, maybe they're going to start hiring 10 instead. So what's going to happen? These people that are losing their jobs or that are unable to find work, software developers,

it is so hard. Some of the smartest people that were in the most demand, you know, seven, eight, nine, 10 years ago, it is hard for them to get a job. Anyways, lawyers, they're going to get laid off in mass. Sorry, anyone in high-price professional services, any sector, really. when layoffs happen, here's what they're going to do. They're going to build companies that are AI native companies that do the thing that they used to do, right? So I think we're going to see things where you can, as an example, pay like, you know, $12, $15, $20 a month for an AI lawyer, right?

So you'll be able to escalate something or I'm sure there's going to be some, you know, some piece, some still human element, right? Maybe you on board with a lawyer. I don't know. Maybe it's not going to be $12 a month. Maybe it'll be $100 a month. But, you know, it's going to be extremely affordable. You'll be able to talk to a human on the front and the back end, but everything else in between will be agentic. Right.

And I think we're like, we're going to have. have more options for these high-priced services, right? So, yeah, management consulting, IT professionals, you know, lawyers, right? These things that might cost, you know, $50, $100, $500 an hour, I think we're going to have very affordable options. And here's the other thing. I think you're top tier. They're going to keep going. They're going to keep crushing all that, right? But they're going to face these kind of cheap AI spin-offs. And here's, here's the other reality, y'all. There's no way the U.S. economy survives more than like three years without this happening, right? Because everyone's like, what happens, right? When, you know,

AI is taking more and more jobs and people have less and less money, well, those people that have less and less money, they're going to create companies that can serve other people. people that have less money, right? And they'll be able to grow and eventually charge more prices, yada, yada, yada. But I do think we have, we will have this cyclical effect of people who lose their jobs potentially to AI or aren't higher, aren't promoted, et cetera. And then they're going to use AI to leverage their expertise and undercut at five to 10 percent of what it would normally cost. Because right now, big companies that are using AI, they're pocketing the difference. Dirty little secret. All right. Our last one for today. UBI becomes a household

conversation. All right. So a new Pew study showed that 45% of U.S. adults favor a proposal of universal basic income or UBI of about $1,000 per month for all adult citizens. Also, we've seen a lot of big tech CEOs who know what's going on discuss UBI. like buffers, right? Because everyone's worried what happens in the future, right? What happens when we achieve artificial general intelligence, right? Where all AI systems are smarter than all humans. Or what happens if we start tiptoeing the line between AGI and ASI, artificial super

intelligence? That's when AI is developing its own AI and we have embodied AI, right? We have robots that can walk, talk, see, feel, right? Sit down at a keyboard if they need to, but they don't need to because, you know, it's all embodied, right? What happens then? I mean, UBI is going to become a conversation. I obviously don't see it happening in 2025, right? I'm a realist, y'all. I know a lot of people in the, you know, in the AI scene, you know, they're just failed at the NFT thing. I'm a former reporter, right? I read more on AI than I'd like to admit. I talk about it every day and sometimes I research three, four,

five, six hours for a single episode. So this is literally all I do. But, I don't think UBI is going to happen in 2025, but I think it's going to become a household conversation, right? In the same way, not everyone was using chat GPT in 2022, but by 2024, it's a household name. Everyone's talking about it, right? ChatGPT has become synonymous with AI. I think UBI is going to have to become a household conversation, right? very few people right now know what UBI is.

It's been around for a very long time. This isn't anything new. This isn't something that is a byproduct of AI's emergence or generative AI's emergence. But the conversation is needed in 2025 with all that's going to happen is going to thrust it to every boardroom, classroom, and dining room table near you. All right, y'all, that is it. at that. I did an episode in less than 30 minutes. My gosh. All right. I hope this was helpful.

If so, please go to your everyday AI.com. Sign up for the free daily newsletter. We're going to be giving you a quick recap, as well as a ton of other stats, facts, and figures that we couldn't fit into today's episode. I wanted to keep these, you know, in the theme of 25, 2025, I wanted to give you guys 25 AI predictions this week. Try to keep the episodes at like 25 minutes or less than 30 minutes. So tomorrow, you're not going to want to miss it. All right. I'll just say that.

I'm excited. We're going to be talking about AI, when AI market forces collide, global powers, prices, and problems. All right. Hope you enjoyed Volume 1. Join us tomorrow for Volume 2. Can't wait to see you tomorrow and every day for more,

everyday AI. Thanks, y'all. Meet Firefly AI assistant. Now live in Adobe Firefly, the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest, orchestrating multi-step workflows across Adobe Creative Cloud apps,

including Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome while the assistant accelerates execution. Stand control with the ability to step in and refine at any time. See it today at firefly.adop.com. 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

your everyday AI.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.