Everyday AI Podcast – An AI and ChatGPT Podcast - EP 446: Your Complete AI Toolkit: Tools, Rules, and Global Changes (2025 AI Predictions - Vol. 4)

Episode Date: January 23, 2025

After testing every major AI development and seeing what actually works (and what's just expensive snake oil pushed by know-nothings online)… lemme tell you what 2025's bringing to the table:  ↳  Regular humans like Jan are about to become kingmakers. ↳  Not just "I-can-use-Canva" queens. ↳  We're talking "I-just-built-an-entire-software-suite-that-3X’ed-our-sales” friggin royalty.  We're exposing the REAL future of business tech in Vol 4 of our 2025 AI Predictions and roadmap series: Your Complete AI Toolkit.  Miss this one and you'll be the last company still waiting for IT while your competition clones your 2026 roadmap and goes to market in 3 days.  The code barrier is DEAD. Time to learn how to play in a world where everyone's a builder and the rules …. Aren’t really there. 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. Non-techies Gets Smarter2. Reasoning wrappers3. Virtual machines4. Politicization of AI5. Global AI RegulationTimestamps:00:00 Evolving AI Toolkit: Future Changes05:24 Predicting Likely Outcomes Live08:44 Future of App Development Trends12:04 Adapting to Emerging Reasoning Models13:29 "Evolution of AI Models"19:27 Musk's Controversial Government Role23:29 "Stricter Global AI Regulations Emerging"24:39 AI's Impact on Global PowerKeywords:SEO assistant, AI tools, large language models, artificial intelligence, AI toolkit, generative AI, AI predictions, AI influencers, software development, non-technical people, AI applications, GitHub Copilot, Cursor, Windsurfer, Lovable, AI models, reasoning data collection, virtual machines, Windows 365 link, Google's Project Mariner, OpenAI's operator, EU act, AI accountability laws, artificial general intelligence, superintelligence, AI regulation, business applications, natural language processing, structured data, unstructured data.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. You're not comfortable yet, are you?

I'm talking about when it comes to how you're using AI tools, large language models. Are you finally into your groove? Well, that might actually be bad news. Because over the last few weeks and in the next few months, things are going to drastically change. All right. And today's show is part of our ongoing series, but specifically we're going to be tackling your complete AI toolkit,

tools, rules, and global changes. So how you approach generative AI in your work, it's going to change pretty quickly. All right. And I'm excited to talk and continue our ongoing discussion. This is volume four of five of our 2025 AI predictions as we are laying out the landscape for 2025. All right. I'm excited for this one, y'all.

I hope you are too. What's going on? My name's Jordan Wilson. Welcome to Everyday AI. This is for you. It's your daily live stream podcast and free daily newsletter, helping us all learn and leverage generative AI to grow our companies and our careers. Is that what you're trying to do?

You're trying to be the smartest person in your company or your department when it comes to large language models, AI, well, welcome. This is your home now. Your other home, Your EverydayAI.com. On there, more than 400 episodes, you can go back, no matter what you care about. They're all separated by category. Learn from the world's leading experts for free. So maybe this is your first time listening.

Thank you. So go sign up for the Daily Newsletter. We're going to be recapping today's show and everything else that you need to know. But more importantly, we are in the middle. actually now toward the end, this is volume four of five. We've been tackling the biggest AI predictions all week. So you might be wondering, like, why am I doing this? Well, I've talked to hundreds of the world's smartest people. I've interviewed them. I've have offline conversations. Big companies, you know, pay us to train their employees on generative AI. And I'm lucky enough

to learn from some of the smartest people. And, you know, sometimes I'm up late at night, connecting the dots, right? And putting all these pieces together from hundreds of conversations. And that's where we're at today. So last year, you know, you might be like, all right, this is random. Well, last year is actually kind of on point, right? Even did a mid-year checkup for my crazy bold AI predictions for last year. And essentially, they all came true.

But they weren't these soft predictions. They were somewhat bold. All right. So that's the good thing about doing this every day. You can go back and check the receipt. So that's, I'm telling you that because I do not make these predictions lightly. I'm trying to lay out the roadwork. All right.

So this is volume four. So let's very, very quickly recap where we've been so far. So if you've listened to every single one, you can just, you know, two acts for the next minute. I'm going to give you numbers 25 through 10 as we're now knocking on 10 through 5. All right. So you can go back and listen to all of these episodes. Again, this is volume four. So volume one, we went over and I'm going to give you the numbers.

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. 21, UBI becomes a household conversation. Then in volume two, we went over number 20. Open source searches. Open large language models will temporarily overtake proprietary models. 19 Chinese AI will dominate and cause confusion. 18 perplexity will pivot, get acquired, or get squashed.

17 API prices are going to drop like they're hot. 16 embodied AI will be an exploding sector. Then volume three, we went over number 15. AI video tools will one shot five minute HD videos and advanced personalized media. 14, the future of traditional internet will come into question. 13, social media makes deep fake AI problems way worse. 12, first big copyright case is decided. And 11, AI influencers start killing off human UGC content. All right. So that brings us up.

So you can go listen to those episodes. Here we go, volume four, no more chit chat. Here's the predictions in the roadmap ahead. Number 10, non-techies will build on the fly software. Number nine, reasoner rappers will hit the scene. Number eight, virtual machines will become all the rage again. Seven, AI becomes overly political. And number six, global regulation around the world.

Titans just not in the U.S. when it comes to AI regulation. All right. As always, live stream audience been doing this all week. Go ahead. Let me know what is the most likely one, think is going to come true. You can just give me the number, say, eight most likely or 10 most likely, whatever you think, and then tell me which one is the least likely, because I want to see,

right? I want to see if some of my predictions here as I'm laying out the roadmap for you all the head is garbage or if it's actually good. So let me know. And without further ado, we're jumping straight into it. I'm also in the theme of this 20, 25, 25 AI predictions, trying to keep all five of these shows under 25 minutes, which if you listen a lot, you know, that's really hard for me to do. I don't know. Apparently, I talk a lot. All right.

Take a sip of the smoothie. Let's go. Number 10, non-techies will build apps on the fly, on the fly software. All right. So, non-technical people. I would consider myself somewhat non-technical, right? Yeah, I've, you know, been in and out of web development, sure. But I've always been more of a duct tape, a digital duct tape, web developer.

if that makes sense, right? I'm not hand-coding anything. You know, I know how things work, but I would say, well, maybe I'm a little more on the technical side of the non-technical people, but I'm saying you are going to be building your own software in 2025. It may not be until later in the year. It may be a small little thing.

You're going to be doing it because it is so freaking easy and so freaking powerful. I actually had a good chat with the Microsoft. off to GitHub team at the Ignite conference. I'll probably have them on the show soon, and you'll see what I mean. All right. Like, it's so easy now. So it's anticipated by this year that 70% of new business applications will use no

code technology. We're not even talking low code when it comes to being able to use these AI tools. So just building apps with natural language processing, all right? So being able to talk to these apps, to build apps, or just typing natural text prompts. So you've probably heard of some of these AI tools that help you build AI tools or not AI tools, right? So GitHub co-pilot, cursor, windsurfer, lovable, et cetera. There's dozens.

I probably forgot some popular ones. So, yeah, let me know if I'm missing one. you know, I always leave in the show notes. If you're listening to on the podcast, I always have my LinkedIn URL there. Just let me know you're from the, from the podcast because Stranger Danger or our email in there. So it's the reality. I honestly don't know what the future of the software development industry is, right? I think for the first time in 2024 software development jobs, new software development jobs were down.

you had some of the brightest graduates in CS computer science, you know, widespread. And it's not just like small instances, but you heard widespread cases. You know, normally people that five to ten years ago would be in the highest demand can't get jobs, right? We saw the story that Salesforce was no longer going to be hiring software engineers, right? So might sound crazy today in two years when even non-technical people are building apps on the daily, I don't think it's going to be that weird anymore. But that's the way I see it. I've done some of these things for myself already.

I started just using Claude Anthropic and then using their artifacts. And there's some of those that I go back and use pretty regularly. But, you know, I've been experimenting a little bit in GitHub copiope. which, you know, Microsoft made a free version of that. Cursor is obviously fantastic, you know, and being able to work with very powerful models like 01, 01 Pro. Right. So that's the other thing.

These are essentially IDE's. So if you don't know what that is, don't worry, you're not alone. So IDE is an integrated development environment. So think of that as it's like a software application that helps, programmers write code, right? So these are these IDs, but powered by large language models, where you can just talk and then it just builds you an app. 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.adobie.com. In one shot and you're like, oh, wow, put a screenshot of something or, you know, say, hey, I want a clone of this social media app and it just works, right? Depends on how your prompting skills are, but you've seen examples. of this and we'll probably do one live here on the show where it's like literally like one prompt.

It's wild. All right. Next. Number nine. This one. I haven't heard anyone else talk about this. Company reasoning data collection. All right.

I think we're going to see reasoner wrappers. So I think this is going to be a big focus of the most forward thinking tech and management consulting companies in 2025. So right now, business decisions. and data reasoning, it requires both structured and unstructured data. So we've spent, you know, the big buzzwords of 2023 and 2024 were rag, right? You got to get your rag pipeline. You got to get your company's rag in there, right?

But what is that? I mean, is that more helpful? Yes, obviously, when you combine a large language model with your company's data, right? Duh, that makes sense. But, okay, you are still at the whim of whoever is steering the vehicle, right? So I think if you've used some of these new reasoner models, right? So essentially think of it like this, you're quote unquote traditional, you know, large language models, let's just call them the transformer class of models, right?

Your GPT, your Gemina, your Claude, et cetera. Then you have your reasoners. Okay. So these are essentially models that take much longer, more inference time and, you know, require more compute. In general, they're a little more costly, all right? And they do this kind of chain of thought reasoning or thinking while you wait. So sometimes I use 01 Pro a lot.

I love it. I mean, I get that thing to think for like 10 minutes, right? It's almost like a little agent. I give it a super hard task. Leave it there. Come back and check on it in 10 minutes. But it's still generic, right? So this same iteration, this same metamorphosis, so to speak, that we went over.

through the last two years when we were essentially, you know, pairing our company's data with the transformer models. We have to do the same thing with these reasoners, right, with these reasoning models, the 0-1s of the world, 03 when it comes out, Gemini has their flash thinking model, Amazon Nova has a reasoning model as well. So we have to take the same thing. But where rag pipelines, for the most part, or in many instances, are more structured data. All right. Reasoning is not. This is exciting for me.

I think there's going to be a big push of, let's just say health care, right? Think if there was an 0-1 model, but it just, the chain of thought or the reasoning was all health care related, right? And it always went through kind of that almost like a triage, right? underneath the hood. But now think of that if it went through your company's decision-making process, right? I don't know how, you know, don't talk to me on the feasibility of this, but I know it's coming, right? Two types of models, two very different classes.

And, I mean, technically, it hasn't just been since chat GPT, right? Because there was rag pipelines and large language models before chat GPT, right? So it's not a new sceptor. for, you know, half a decade or more, or at least popularized for half a decade or more. So we're going to start to see, I think, the first iterations of reasoning wrappers or tying in your first, I like to call it first company data, but on the reasoning side. So it's all about how do your business leaders make decisions. What is that thought tree that maybe isn't, you know, written down?

it's going to be a big, a big boon of business in 2025. All right, let's keep going. Number eight, virtual machines are going to become all the rage again. All right. You might be like, there's no way. Yeah. I think you'd be surprised. I don't think it's also a coincidence.

We saw Microsoft release the Windows 365 link. So it's not a technically a virtual machine because there is a, think of, you know, I'm trying to think of the size of something, right? Think of like a Mac mini, right? But think of like an extremely powerful computer that would normally be like 10 times the size. So, you know, the Windows 365 link, it's essentially a virtual machine, but you have a physical box, right? And then you get to tap into, you know, via the web, the interwebs. You tap into a computer that would be, like, in theory, 10 times the size, right? I think we're going to see the same, but just completely virtual.

Here's why. Let's talk about Claude's computer use. It's clunky, right? But it uses a virtual machine. Granted, that virtual machine looks like it's from like 1997. But if you use, this is kind of the first popularized agentic AI, question mark, right? at least from one of the big, either tech companies or the big AI companies, super clunky, but like I've said on the show before, you got to tip your hat, tip your hat to Anthropic

for actually releasing it. But it was the first kind of taste, the mainstream taste of agentic AI. Or when you can tell a large language model, hey, go use the computer, go to all of these different things, you know, go to this website, browse this, then put it in a word document, then go ahead and make a spreadsheet and save that spreadsheet, right? stringing together multiple tasks that a computer needs to, or sorry, that an AI needs to have access to a full working computer. They're coming back.

So it's not just Claude, Claude's computer use, but Google's Project Mariner requires an active tab. So that previously, it was called Google's Jarvis. I told you all they weren't going to, you know, be able to use that name, the whole Marvel thing. So Project Mariner requires an active Chrome window, right? So that's going to be their first, Google's first agentic AI product. I haven't even seen it released. It was supposed to go out to like tested users or whatever. Who knows if we'll even see it in the first quarter.

I'm guessing it's going to be the latter part of 2025. But the same thing. It requires active use of your browser, right? And then we have, should be released any day now, Open AIs operator. Presumably, it's probably going to require a little bit of your current screen or your current ChatGBTGBT account as well. So, I mean, all these things are adding up, right?

All of these more agentic workflows, everything's agentic in 2025. All the big companies, Microsoft, Google, Open AI, Claude, Salesforce, everyone. Literally everyone is all in on agentic AI. And so many of these systems, right? They emulate. a computer, right? Or you can run them locally, right? There's going to be local models, too, but they're going to need your computer. So I think people having virtual machines is going to come back, you know, remote PCs, you know, because you can't really do anything else if you're doing

those things, right? I guess, luckily, this is where my computer hoarding comes into play. I think I have, what do I have? I have four Macs and a PC that I use. So maybe that's where that comes in handy for me, I guess. But it's going to be virtual machines are definitely going to come back. All right. Two more. Look at this, y'all. We might actually finish an episode in 25 minutes.

Let me get a little hydration there. All right, here we go. Yeah, if this is your first time, this whole thing's, it's unedited, unscripted. This is just live. All right. Number seven, I'm going to be careful. so I don't get any hate mail, all right, from you, Elon lovers. But AI is about to become very political.

Unfortunately, all right. I already told you one of the reasons right there, Elon Musk, right? So former co-founder of OpenAI, now founder of, you know, I don't know, 50 companies, a couple of them losing a lot of money from Twitter. All right, but, you know, XAI. So technically one of Open AIs main competitors. So you might be saying like, okay, well, what's getting political? Well, Elon Musk has this pseudo official role in the U.S. government.

All right. Dodge, the department of, oh, I think it's D OGE, not D O'DGE. My screen here is wrong. I did that wrong. I'm a human. I make mistakes. What do you think I am a robot? I don't sleep.

All right. So the Department of Government Efficiency, which is where essentially Donald Trump, President-elect Donald Trump, has tasked Elon Musk with this unofficial but official but unofficial government position where he's supposed to go find all the waste in government, right? But he's already throwing his weight around in a way that I would say is extremely inappropriate, right? because he just, there's just some news stories that came out that he said he is urging attorneys general in multiple states to auction off the nonprofit arm of open AI. Can we think about that? So someone who's literally, at least from a public standpoint, right, Trump saying, hey, Elon, come in and make government smaller, guess who you don't want to piss off, that guy?

All right. If you know Elon Musk, he's all over the place. All right. You know, he like went in when he bought Twitter. I forgot, but it seemed like he fired like half of Twitter, right? And also I think he bought it for $44 billion in the last numbers I saw as it was worth like, I don't know, $10 billion. Anyways, I'll leave that there. You can do what you want with that.

He's urging, literally, government employees. to auction off the nonprofit arm of Open AI in multiple states. I was hoping that the artificial intelligence field would never become a political place, but it's already going there. All right. So Zuckerberg, Mark Zuckerberg, is already now getting more vocal and a little more political. The New York Times called this Meta's mega makeover, as he's been, you know, aligning with President-elect Trump on a lot of recent issues.

I think it makes good business sense, right? But I think we're, unfortunately, AI is going to become very political. One other piece, right? You got to add all these things up. So David Sachs, you know, popular venture capitalists, also podcast hosts, was named the AI czar. But he's been linked to some controversial opinions in the past. So been linked to some allegedly linked to some.

Russian propaganda and other past controversial statements. So if you thought the world was already political and AI, you know, artificial intelligence and large language models, if you thought that was your safe place, you know, here in the U.S., things get very political, right? If you thought AI was the safe space, you know, you can escape from all the political nonsense. Nope. Looks like it's going to ramp up. And I predict it is going to get ugly.

It's going to get ugly. All right. So we'll see how that actually pans out, but book it, book it. AI is political. I'm not happy about it. I don't want that to happen. I thought we were just great in the last like two years, right? AI is super innovative.

You know, it seems like the outgoing Biden White House, you know, they had, you know, some, their executive order, but it's not like they're overly restrictive in artificial intelligence. So, you know, but it looks like we're making a shift. We'll say that. It looks like some of the leaders in the AI space are going to be cozying up to those in positions of governmental power and it looks like they're going to be throwing around some weight.

All right. And then our last AI prediction as we lay out upcoming trends of 2020. Global regulation around AI Titans. It's not in the U.S., not at all. All right. So we already know the EU Act has introduced some pretty harsh regulations, classifying certain AI applications as high risk in imposing fines for noncompliance. Yeah.

So a lot of these, you know, AI innovations and these new, you know, large language model features from various companies, businesses in the in the EU can't even use a lot of them, right? Similarly, some countries like Canada, Singapore, Australia, etc., have also initiated their own AI accountability laws. Not all of them have gone into effect yet, but it's going to, I think, drive a further divide. Okay. So the EU's, I believe some of it, the EU AI Act, some of it came into effect in about six months ago. And then the rest is going to come into effect in about a year and a half. But regardless, I think there's going to start to be some pretty strict rules and regulations around artificial intelligence. Here's why. The talk as of recently has been all about, artificial general intelligence, and it's been all about actually super intelligence, right? That's what happens next, right? It goes, you know, AI, and then it goes AGI, and then it goes ASI,

and as that global conversation starts to shift, more and more people are going to become worried about their jobs, right? Which means, I think a lot of countries that want to be a little more conservative that want to roll a eye out at a little slower pace, I think we're going to see a couple of things happen. I think that some of these countries are going to start to lose some of their global positioning when it comes to, hey, is this country a global superpower anymore? I think, you know, it's not like a, you know, company is going to go from a tier one to a tier five, but I think if you are too strict and too restrictive on AI and not allowing the

big enterprises in your company to use and fully use them. You're going to see one of two things. Either those companies are going to leave. Yeah, like, we're out, right? Or going to bring in less money. You can't compete. You literally can't compete, right? I wish I had a better analogy, right?

But think of race cars, right? I don't follow NASCAR or Formula One. But think of the very first cars, right? Like, I don't know, 50, 100 years ago, however long that's been around. And then look at them now. They're probably more than twice as fast. All the technology, so much safer, right? That's the difference, right?

Between enterprise companies being able to fully use and fully leverage artificial intelligence versus those that can't. You're driving in an old Formula One car that goes as fast as my 2008 Honda Civic hybrid. It's slow. So you're going to see a great divide there. All right. We made it to the end. All right.

And obviously, y'all, you know I saved the best for last. You know it. My spiciest AI predictions and trends for you to get ahead in 2025 are coming tomorrow. So thank you for tuning in. I appreciate it. Make sure if you haven't already, go to your everyday AI.com. Sign up for the free daily newsletter. We're going to be recapping all of these.

hot take predictions as well as links to the previous four volumes. All right, that was volume four. We're going to have links to volume one, two, and three. You can find that in the podcast as well. If you are listening on the podcast, I hope this is helpful for you. We bust our collective butts here at Everyday AI to make sure you are the smartest person in your company and your department at around AI. And these predictions, they're more than that.

They are trends. They are laying the groundwork for you and your company. to succeed. That's why you've got to subscribe. Please leave us a rating and please join us 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.adobie.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 EverydayAI.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.