Everyday AI Podcast – An AI and ChatGPT Podcast - Ep 840: The Next 12 months of AI: 19 Predictions Every Business Leader Needs to Hear

Episode Date: August 13, 2026

The next 12 months of AI leaked. Kinda. For the past 90ish days, we've been quietly collecting evidence of what's next.1,030 saved posts. 90 Podcasts. Countless conversations. Every model drop, every leak, every quiet product update the big labs hoped you'd scroll past.Then we connected the dots.What came out the other side: 19 calls on where AI goes over the next 12 months. And some of them are uncomfortable.We're walking through all 19. Bring your team's AI roadmap. You'll want to edit it. 👇The Next 12 months of AI: 19 Predictions Every Business Leader Needs to Hear -- An Everyday AI Chat with Jordan WilsonNewsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageToday's Episode on LinkedIn: Thoughts on this? Join the convo on LinkedIn and connect with other AI leaders.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:Reactive Chat Dies, Proactive AI Agents RiseVoice and Mobile Become AI Default InterfaceManager Threads Replace One-Off AI ChatsMultiplayer AI: Humans and Agents CollaborateCompany-Wide Vibe Operations with ChatGPT SitesAgent Native Workflows and Resources StandardizationSkill Reuse as Key Company MetricCompany Reasoning Data as Strategic GoldShift from Public Leaderboards to Private EvalsModel Routing Becomes AI Industry NormCheaper AI Intelligence, Anthropic Competition HeatsFortune 100 AI Token Spend EfficiencyCompute Power as New AI CurrencyLocalized AI Controversies and Election DeepfakesMainstream AI Backlash and Content DetectionMath Benchmarks Solved by Advanced AIToken Maxing Returns with Cost DeclineOpen Agents Crash Risks and CybersecurityRecursive Self Improvement (RSI) in AI DevelopmentTimestamps:00:00 Starting the AI 101 series03:35 Yearly AI predictions roundup07:54 Using full duplex AI assistants09:45 Talking vs. Typing to AI13:45 Breaking down AI silos19:03 Turning processes agent-native22:32 Skill development and reuse in AI24:09 Bringing Slack DMs into Channels29:46 Dealing with AI usage limits30:45 AI startups revolutionizing knowledge work35:46 AI strategy in Fortune 500 companies40:11 AI impact on local politics41:58 Concerns Over AI Watermarking46:52 Experiencing token budget challenges51:05 Sergey Brin prioritizes RSI at Google52:06 Discussing AI model improvements55:24 Closing and subscription reminderKeywords: AI predictions, AI trends, business AI strategy, proactive AI agents, reactive chat, AI operating systems, ChatGPT, Claude, Grokbot, voice and mobile AI control, full duplex agent, AI skills, skill reuse, manager threads, multiplayer AI, agent native, company reasoning data, private AI benchmarks, public leaderboards, private evals, model routing, AI token spend, open source models, compute scarcity, hardware scarcity, AI controversies, local AI data centers, AI deepfakes, AI backlash, AI content detectors, AI in politics, math solved by AI, token maxing, cyber defense, open agents, cybersecurity budget, recursive self improvement, RSI, Fortune 100 AI usage, AI workforce transformation, dashboard automation, AI for dashboards, no-code AI apps, business intelligence AI, automation skills, agent crashes, model overhang, vendor lock in, AI-powered cyberattacks, AI-driven skill creation, AI-enabled workflows, token efficiency, AI local hosting, cost-effective AI models, enterprise AI adoption, AI asset management, company AI metrics.Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Ready for ROI on GenAI? Go to youreverydayai.com/partner 

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. This could randomly be one of the most important shows I've ever recorded. Why? Well, I'm laying out the AI answers for your company for the entire year right here. Let me explain. Usually in December or January, I put on my A. AI predictions hat and tell you what's ahead. And for the past three years, I've done a pretty okay job in these prediction shows at telling you what's next. But the pace of AI is much different now.

I mean, measurably, AI models are literally coming out about two to three times faster today than they were just nine months ago. So we can't wait until another four months in December or January to give you the roadmap ahead. So with the school year starting and with, this upcoming start here series we're about to restart from the beginning, I thought I'd just give you the answers today. Not the syllabus, not the take home test, just the answers. So 19 quick points for you and your team to focus on, whatever the sector, whatever point you're at in your career, whatever type of work that you use AI for. I think today's show can be your answer sheet for the whole year. So you and your company are focusing on what's next, not what already

happen. Let's get into it. If you're new here, welcome to Everyday AI. My name's Jordan Wilson, and well, we do this every day. It's your daily unedited, unscripted, live stream podcast and free daily newsletter helping business leaders like you and me, not just keep up with what's happening in the AI world because it's too much for anyone human. But I help you get ahead to know what's coming so you can grow your company and your career. So it starts here. Make sure you go to our website at your EverydayAI.com. Sign up for the free daily newsletter. We'll be recap, today's show and a whole lot more. So as a reminder, we are taking a kind of break in our normal programming. So since it is that back to school time and all the kids are going back to school,

we're all going to be doing some learning together on everyday AI. So our start here series show that we kicked off earlier this year, we've been doing, you know, one or two shows a week. and we have 30 of them. And they're pretty amazing. And they're some of our most popular shows ever. So we're going to be running them all back, the entire series, starting Monday, with the very first episode. So yeah, a little break from our normal Monday to Friday schedule.

But I guarantee you if you just stick with the series from Volume 1, listen all the way to Volume 30, whether you listen to one, some, or none of them. You're going to know more about AI than like 99% of the other people at your company. So the series cuts through the jargon, cuts through the fluff, whether you're a beginner, you around AI all the time. Just it's simple. No computer science degree required, but class starts Monday. Don't be late.

All right, but let's now get into it and talk about those 19. Yeah, we have 19 of them. So these are the things that I think your company needs to be focusing on. Yeah. So for the school year here. Right. So yeah, classes in session. That's why we have the little yellow, the pencil yellow accent color here for my slides. So these are no real order, right?

I was just kind of jotting down some notes. You know, sometimes I just am always keeping, not like predictions, but a lot of my observations. Because from doing this for three and a half years, literally spending my entire day every day, reading about AI, exploring AI, talking to other very smart people about AI, right? I have a, what I would say is a pretty accurate notebook of random ideas and thoughts. And I've been spending some time putting them all together. I came up with 19, right? So, you know, for my yearly prediction shows, I always do, you know, for 2024, it was 24,

then 25, then 26. Well, we got 19. All right. So I don't know. Maybe that's roughly how far we are through the year. But this is, this is, we're lining this up at the school year, right? we're starting in August. I think these should all still be fairly relevant come, you know,

when May, when school gets out. But let's start with number 19. So this is when reactive chat is dying and proactive agents are taking over. All right. And I also want you to think, right, where are you at on the AI usage scale? Because maybe what I said is a shock. Maybe you're like, duh. it's been this way for a while, right?

So as I rattle these off, right, keep in mind, our audience is all over the place, right? We literally have AI leaders at the labs listening to this. And then we have people that have stumbled upon Chedgedyptych for the very first time and have never used it. All right. So keep that in mind as I walk through these. So maybe you've heard me mention some of these things before.

Maybe you don't know what some of them are. So I'm going to try to keep it simple. But I think over the next year, reactive chat. is going to start dying, right? It's the combination of not just agents, but just being able to very easily save and reuse skills. That is huge, right?

I think that chat GPT work as an example is really indicative of this. It's no different necessarily than Codex, right? And Codex was kind of the, I would say, one of the birthplaces, of the hardworking around the clock agent, that as well as Claude Code and Claude Co-work. But I think the new normal is just going to be, you know, waking up and reviewing what your AI already shipped overnight. And that piece is extremely important.

All right. So if you're prompting less than what are you doing. All right. That's number 18. You're talking more. I think that number 18 kind of prediction 18 here, I think voice and mobile are already becoming the default AI controls. So if you haven't used Chad GPT's new feature yet, my gosh, chat GPT voice is outstanding. What's crazy is this has been out for almost a month. And, you know, normally the people that are quote unquote early talking about AI are starting to talk about it literally now. But I think, you know, to summarize this, I did a show on it a couple of weeks ago when it did first come out.

But this is essentially kind of the combination of voice and mobile becoming the default AI control. Yeah, you can go out and touch grass. And in my experience, I get a lot more done now. when I take my iPhone out, right? And this is another, you know, OpenAI slash chat. GBT. And I think, you know, Claude has made some improvements to theirs. It's still not very good.

But I assume everyone is going to have to follow suit because this is so, so powerful that you just have to have it. Right. Even I think we saw a glimpse of it with, you know, Grock, Grock bot that just came out. But to be able to talk to a. full duplex agent. That means an agent that can listen and talk in at the same time, but one that also has access to your data and can look up information. It's literally faster than talking to a coworker, right? Because I can talk back and forth and it will respond to me, but then also while it's responding to me, it's going to, you know, personalize everything

through the data that it has connected as well as go and look things up. you know, for me, this is like, you know, working with someone that's smarter than the smartest coworker I've ever worked with, right? But I think that this has to become eventually the main surface fairly soon just because it is that good, right? And so you're like, okay, what does this mean? You know, just isn't that just mobile? Not necessarily because the desktop is obviously much more powerful than what you can accomplish on any mobile phone. But the way that this works, least in chat GPT voice is, well, you can open the chat GPT app, click on remote, and then literally control your entire computer, right? So this is the Jarvis of AI where it can see what's

on your computer. It can click things. It can open, save file, shut, run commands. It can do anything. But I think that that is, if you and your company have not already started exploring this, you should probably have your eyes on this because it is a much more fluid and natural. old way to work. I like to say it's kind of like a flow state. Right. I don't talk about this a lot, but sometimes, are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on Gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft, and Nvidia have partnered with us because they

trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for chat GPT training for thousands or just need help building your front end AI strategy, you can partner with us too, just like some of the biggest companies in the world do. Go to your everyday AI.com slash partner to get in contact with our team or you can just click on the partner section of our website.

We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on Gen AI. And maybe this is because I'm a former journalist, but sometimes when I'm a hands-on keyboard, maybe I'm trying to get things a little too right. You know, sometimes I don't just like to blurt out on the keyboard. And that might interrupt kind of the natural flow state of what I actually want to communicate to an AI or to an agent. Obviously, when I'm talking, it's a little easier, I think, ultimately. And I think that holds true for a lot of people. There's actually some

studies that show that you can much more effectively communicate your point when talking to AI agents while you're talking versus when you're texting or typing. Because when you're typing, you're already putting, you know, multiple limiting filters on it because you know that you can only type so fast and your fingers cannot keep up with your brain. your mouth can do a little bit better of a job. But I think number 18, that is definitely the next control. Number 17, this is the manager of threads. So we've already seen this in Codex and Chad ChbT work.

We've seen a smaller version of this that was actually announced. I'm losing track of time. I think it was this week from Claude, right? But what does this mean? I think we were trained for better or worse when working, whether inside, you know, co-pilot, Gemini, Chad, GPD, Claude, etc. You know, that any conversation had its home, right? So, hey, I need to go back into that chat, right? And this was especially more important in the earlier days of prompt engineering.

This is something I taught to thousands of students that took our free PPP course because this was the constraint, you know, probably up until a year ago. is the context really just lived in that single chat thread. It's not like that anymore, right? So now I think you have to think of manager threads replaces those one-off chats. So you could call those like a chief of staff, a chief of staff thread, right? So that's the way I think a lot of people set it up. And in a project, you can have, you know, an agent's MD file in any project. And you can create an unlimited amount of threads or chats in there.

But then any thread in, you know, Codex and ChadGBT can control any other threat. And again, that is an interface that all, you know, Google, Claude, Grock, everyone else, you know, has either, you know, taken that as well or they will have to soon because that is, you know, the next step in terms of communication. And I think that's, that's a smart way to do it, right? That's how it is with humans, right? If you're sitting with your team, you can talk to them at any point about anything that you all know. You don't have to wait for the marketing meeting. If you're sitting in the marketing team is there, you can just talk about it. So this is a much more natural way to interface with AI agents.

All right. And once those threads are managing threads, the next step is obvious. Well, you put people and agents in the same room. So that is number 16 on the list here. Multiplayer AI with humans and agents sharing the same workspace. And over the past two weeks, we've seen some great examples of that. I already mentioned Grock Bot. So if you haven't seen this, it's only been out for like not even 72 hours yet.

Right. But very similar to Buzz from Block, you know, Jack Dorsey of Twitter fame. But the same thing. So we've seen these two recent interfaces come out that are multiplayer by default. So where you and your entire team can come in and chat with the same agent. Right. So we've seen glimpses of this, right? There was a Google leak about bringing some multiplayer capabilities.

Obviously, you know, Open AI inside chat chvety. You know, they've had these group chats. So presumably a lot of these. these things are going to be moving over. But I do think that is the next interface. And why is that important for you and your business? Because I think siloed thinking when it comes to AI has been so deeply ingrained in our thought. And it is going to change quickly because what you can accomplish with a team in this environment

is going to be so much better because, y'all, I've done training for large organizations, right? And there's always such a disconnect between kind of the AI champion. and everyone else. And one of the things limiting, I think a lot of teams and keeping them from moving faster is things like sharing skills, right? It's these permissions things. So when you start to break down those walls, that's where I think the gains are going to be

compounding faster than they have so far. And we've seen it now, right? This is two strikes in the past two weeks. And I think the other big companies will be following suits. So number 16, multiplayer AI, you should be investing in it, right? So I'm not saying, you know, drop Claude and go to Grockbot. And I'm not saying, you know, drop Chad GBT and go try out buzz. I'm not saying that, but you should be exploring in there.

So then when your AI operating system of choice brings something like this, your team is ready to go. All right. So now that the whole team's in one room. Yeah, look what they can build. Number 15, vibe operations, y'all. Again, I keep leaning on, you know, some chat GPT and open AI examples because I think, you know, probably in late 2025, early 2026,

I think it was, you know, anthropic that was innovating. And I think ever since it's been open AI. But, you know, let me just give them one example. I think chat GPT sites is probably the most. under hyped and underused thing out there right now. Like I said, I think chat GPT voice, now enough people are realizing like how good it is, still no one is talking about chat chbt sites. And let me tell you what I mean by vibe operations, but this is when the whole company starts

building. So if you don't know what chat chitpt sites are, think of it like a lovable or a replet. right so kind of a vibe platform um that you can you know create simple apps and dashboards and all of these things but the difference is well you can change them and they live so it's it's back and off it's databases all of these things you don't have to know anything right i actually sent something to my wife uh when was it i think it was today and she's like wow this is really cool you know it's chat you vd sites because all this information you know it had photos and directories and listings and all these, you know, the ability to save these things. It is like, oh, wow,

what is this? Oh, you built, yeah, Chad GPT sites, right? But, you know, that is available within teams. So, uh, I, I'm not saying that teams are going to, you know, replace their, you know, expensive software subscriptions. I'm not saying that, but I think that they're going to start filling in the gaps. And, you know, instead, I think like dashboards are probably the simplest, um, explanation of that, right? So, you know, a lot of times, you would have to, you know, get your request in line with someone in, you know, BI working in business intelligence who puts together the dashboards. And even then, you know, it's a lot of back and forth.

Well, now you can literally just drop your files in there. Your team can collaborate. You don't have to know backend off servers, anything. It's just there. And you have live databases that you can collaborate on in it natural language. So I do think that we're going to see a shift eventually toward vibe operations. It's one of those things I'm like literally still pulling my hair. about like, why are people not using chat GPD sites more?

Because there's no real true competitor, a single competitor, right? The closest would be, like I said, replant and lovable, but you don't have the same power of users in data and connection, especially at the multiplayer level that you have with like chat GBT teams or chat chbt enterprise that can get these sites up and running for their organization. All right, next. Agent Native. All right.

This is, I think, more than a buzzword. All right. I've been thinking a lot about this and what it means to be agent native. And, you know, I keep thinking back to, you know, certain things in my background, working in marketing, you know, things like, you know, mobile first or mobile responsive, all these, right, there's this huge push. push to get since people were using the internet more on their phones, you know, 15 or so years ago, there's this big push. It took 10 years, right? But eventually the entire web became mobile responsive

by default because the majority of people are accessing the internet on their phones and apps and everything, your business, everything had to be mobile responsive. The same thing is going to happen, but going agent native. All right. So I'm not just talking internally. I'm talking externally as well. So what does that mean? So resources that your team makes, they need to get repurpose in an agent-native format. So, you know, all that prompting that you go through and do, you and your team, all right, all of it needs to turn into a skill because that is a skill that, or that is you, taking the outcome that you came to in making that outcome in the process that you got there, agent-native. In the same way that, you know, teams would get together and build

SOPs, right? That's what you should be doing here, but doing this for agents, according to the successful processes that you've brought into different AI operating systems. So that's one example. Another example is, well, I think all websites should be marked out, right? Like there should be a button, right, aside from having an agent first web, which I think is already happening in the background, right? I think many companies are going to have two versions of their websites. The second version will probably be made automatically or it just converts everything to mark down.

So it's easier for agents to go through without having to be a token inefficient AF and spend billions of tokens just to understand what's going on in the website. But I think in the same thing, simple user interfaces where you can say, you know, save this as marked down or convert this to markdown, you know, being able to easily turn your, you know, your company's websites or anything that you publish online. So an agent can more easily access it. So just two quick little examples there of being agent native. All right.

That's number 14. So every once everything is kind of agent native, well, you need a new way to keep score. And I kind of already referenced this. But I think that skill use is going to boom and become a company metric. What do I mean by a company metric? Right. There's all these things, I think through the course of AI that have been like weird metrics, right? So like early earlier on, right, when AI was really hitting software development hard, it was all about the lock, right?

It's all about the lines of code. Right. Everyone's like, oh, I wrote 87 billion lines of code today. So, you know, I'm awesome. And then, you know, late 2025, early 2026, it was the token leader boards, right? Those two things say absolutely nothing. I think one of the most important measurables in AI is going to be skill, re-reaching. use. All right. I talked about this on the show last week, right? Kind of how I AI, right? We did the

entire thing. I think it's like 6,000. I can actually bring it up here. Bring up my codex. Let's see. Okay. So I've used skills 6,700 times. So it's a lot of skill use, right? It's not, could be more. But I think that skill reuse is actually going to become one of the most important company metrics. Because when I said, it is so easy to turn a repeatable process into a skill. Let's say that you're working with a large language model. You're going back and forth. You're sharing data.

You're doing all these things. And you finally get an outcome or an output or deliverable that really works for, you know, you, your company, etc. Well, what do you do with it? Maybe you might turn it into a project. Okay, that's cool. But you need to turn that into a skill. not just because you can invoke it or an agent can invoke it on its own without you even asking it, right, which is usually helpful.

Sometimes it's not helpful. But the other thing is that is how you actually build AIIP for your company because skills are transferable, right? So if you're just stuffing everything, you know, into a project inside, you know, Google or co-pilot or Chad GPD or Claude or whatever, it's not really an asset, right? At that point, that is a vendor lock-in. That's all you're doing at that point. So, you know, skill reuse is such an important part of, I think, any measurement of a good, someone that is becoming AI-native is how much and how often you are reusing skills.

Because in theory, even if that skill is not always leading to a improved output or deliverable, it is undeniably saving an extraordinary amount of time, even versus normal AI use, let alone doing things the quote unquote analog way, right? Working on the internet, you analog. All right. So let's move on. So skills capture how works gets done. Number 12 is how you capture decisions getting made.

And you know what? Finally, I think I had this in my prediction, like two years ago. ago and I was too early, but we're finally starting to see this. Number 12 is company reasoning data finally surfacing as the goal that it is. So you're now seeing big companies like Zapier was one of the first that I saw talking about this about the concept of bringing like Slack DMs out of the DMs and into the open. And why and what does that mean and what is it signal? So, you know, essentially when you have an agent in your Slack, most,

agents cannot see DMs, right, aside from your own personal agents. But when you start moving important conversations from individual DMs or groups of DMs into channels, right, not only does that help the decision making, help others maybe see the decision making progress, but most importantly, it allows agents to understand how your company or how your individuals think and make decisions. Yes, that's obviously out in the open in normal channels. But the concept of, you know, and I actually talked with the Slack CMO about this on the very show about how important it is to get kind of that human reasoning, that one-on-one

human reasoning, a lot of times, that is what the agent so sorely needs. But we're not getting it. We're just throwing more spreadsheets, you know, more skills and all these other things. When really it just needs to be able to see the nuance of a conversation between, you know, two, three, four decision. makers that was maybe private. So that's just one small example of company reasoning data finally being serviced as gold. But I think that we are going to see that become bigger and bigger as agents are more easily

to go out and crawl and understand different contexts, right? People are like, wait, you know, even things like just yapping, right? Kind of going back to my thought of, you know, voice and mobile being the surface that is going to be the most important. You know, the amount that I can speak something versus type, there's so much more nuance and even reasoning that I can fit in there when I am just naturally speaking what's on my mind versus what just gets typed out. All right, let's keep going. Number 11, the public leaderboards, I'm not saying they're going to die completely, but they're going to begin to die.

And I think that private evals are going to start to take over. I'll say this, like 90% of benchmarks are just becoming useless. I mean, there's some that are obviously extremely important to look at, regardless of the type of work that you and your company do, right? There's ones that are just great for general intelligence, right? Like GDP Val, the artificial analysis index, deep suite. I think some of those are just like no-brainer. So they're always going to be important.

There's like hundreds of benchmarks. And people spend so much time looking and thinking about benchmarks. And aside from bench maxing, which is where, you know, frontier companies, you know, they overfit a model to perform very well on a specific benchmark, even if maybe humans or real work might not benefit as much. So I think that public leaderboards and some of these benchmarks are going to decrease very quickly. And I think the practical thing that's going to happen is, well, company, companies, internal workbench, essentially, you know, kind of what I'm calling it, that's going to become more important, right?

Companies are going to start in some of the larger companies already have, right? They're actually, you know, creating these as, you know, like public evaluations, right? They're creating these as public benchmarks, things that, they use for internally and then sharing them with others, I think those are going to become more and more important, but more and more common, right? I think it was like DoorDash or something that shared one of theirs. And I was like, okay, this is random.

I'm like, no, this is actually very important because they took the models and it maybe didn't for them, you know, a certain model might have been much better or much worse than even in the corresponding evaluation just because of the type of work that certain people do is so specific. All right. And then next one, kind of related. I think model routing is going to become the norm.

And eventually it is going to become built in. Okay. So here's what I mean by that. Yes. Most companies should stop defaulting to the top model because you don't need it. Right? You literally don't need it for most of your work. All right.

And this also plays into kind of the, you know, model or capability overhang. A year ago, I would have said the opposite. I would say absolutely use the best model available, even if it takes longer because you need it for your work. You don't need it anymore, right? You don't need, you know, Fable 5 Max or GBT 5, 6 sole max to, you know, browse. the web for you. You know, you don't need Opus 5 or, you know, Max or GPD 56 terra to rewrite an email, right? You don't. You know, there's very capable models that cost a lot less. So if you're paying for it on the API side, you've already been through this, right? But even on the subscription side,

right? Now, I think eventually the free ride of AI, even in subscriptions, will become less subsidized. and you're going to have to start dealing, you know, with usage limits and in quotas. So model routing, it's going to have to become a default. And it's actually very easy, right? I've made skills that do this for me, right? I have a skill where, you know, I give a big project to, you know, GPD 56 soul. And it breaks it up and it uses subagents at its discretion, but lower volume one. So, you know, I really just have it working as the orchestrator in many instances.

And, you know, I say if it requires heavier lifting, you can do some of that middle work. But for a lot of those things, well, I'm just using, you know, maybe Seoul in the top 10 and last 10% and then using a bunch of, you know, smaller subagents in between. But I think that's going to become the norm. I think there's going to be some, and there already has been some interesting, you know, AI startups like merge and others that have come out that, This is literally what they do, right? Because I think it's going to save a lot of companies, a lot of money. And that's something that companies are looking to do as the model overhang starts to become larger and larger. And what do I mean by that?

I think that 90% of most knowledge work can be done with using only 10% of that top, top models, And just wait, you know, until we get Fable 5-1 and we get, you know, GPT-6 Astra or GPD 5-7 Astra, right? Whatever it is. Our work as humans is not growing in capabilities at the same rate that the models are going. They're just not any. Yeah, like I said, two years ago, there were models that, you know, the best models weren't capable to do all the work that we do. A year ago, I would say it was maybe in parity, right?

But yeah, you needed that most powerful model to do your hardest work. And now for most people's hardest work, like I said, you know, if you, you know, Sonnet might be good enough, right? Or, you know, Opus might be overkill, you know, let alone Fable, right? You might just need Claude Sonnet or you might just need. you know, GPD 5-6, Luna is plenty for many people. But that's why I think model routing is going to become extremely important.

All right, we're going to pick up the pace so this doesn't turn into a two-hour show. Number nine, intelligence is going to get cheaper, cheaper, cheaper, cheaper, cheaper. And what that means, I think in the short term, at least, is Anthropic is going to get squeezed, right? I've talked about this a lot on the show. If you care about, you know, who's your vendor,

I'm not telling you to choose one or the other, but you have to understand what's going on. And it's been absolutely nutty the last two weeks. So as an example, OpenAI used a sniffling of, you know, recursive self improvement. You know, they use GPD 56 sold to make their other models smarter and cheaper, right? They drop the price of GPD 56 Luda by 80%. So that model is nearly free. That is like nearly sonnet level.

All right. But it's not just them. I mean, GROC 4, 6. that just came out so good, right? At least, you know, first looks and cost per task. You have Meta's Muse Spark 1.2, Kimmy K3, Quinn 3.8. All of these models deliver 95 to 99% of the intelligence of the anthropic models, but at 5 to 30% of the cost.

Those aren't random numbers. Those are the real numbers there. So what do I mean by that? Well, intelligence is going to get a lot cheaper, right? It used to be the thing where, you know, you could just choose your, you know, choose your horse and you would know that your horse would always, you know, stay competitive. I don't know why it looks like Anthropic is not going to follow the trend here of, you know, making intelligence cheaper and cheaper at the same time stronger and stronger.

So that's going to be something to definitely keep an eye on. But I think as the school year drags on, I think, intelligence is only going to get cheaper and cheaper. So your provider that you're sticking with, if they're not dropping their price, price per performance, yeah, might want to look another way, just saying. All right. Number eight, and this kind of all goes together, but I think Fortune 100, token spend is going to slow while usage explodes.

And let me explain what I mean by that. Similarly to, you know, I was talking about anthropic just selling tokens. I think Fortune 100s, at least right now, are the only ones truly in the position to be able to, well, have the CAPEX or the facilities to be able to run local models at scale. Because right now, you know, again, unless my joke is, unless you have a data center sitting in your basement, you can't run these open models, right? The ones that, you know, oh, you know, everyone's like, oh my gosh, open source models are here. all the big companies, what are they going to sell anymore? Well, yeah, you can't run the open source models on consumer hardware. You need a couple million dollars worth of GPUs to run them.

But that's where it's like, okay, well, the Fortune 100 companies, they've already started to do it. They're going to start to shift a lot of that lower hanging work, I think, to open models that they're running on-prem. Is that going to trickle over to the rest of the Fortune 500? You know, your average everyday enterprise this year, absolutely not. But I mean, I think that you're going to see token usage go up at these companies, but their token spend or the amount of tokens that they're buying from a provider go down. They're going to stop, you know, renting AI and they're going to start owning it. You know, and I think that that is going to become more and more common over the years.

But I think at least for this year, if you are listening at one of those Fortune 100s, you're probably already having those type of conversations. If you are, you know, maybe in the Fortune, you know, 101 to 500, right? These are just loose numbers. It's something that you should probably start paying attention to and start crunching the numbers because I know a lot of companies that are spending nine figures, you know, literally nine figures on AI, you know, so at a certain point, you have to say, okay, how much of that, you know, let's look at our model routing. Are we routing models at all? How much are we

overpaying, right? How many people are using, you know, a fable five to rewrite their emails? Or how many people are using, you know, Google Jevon-I, you know, 3.6 flash to, you know, classify something when you could be using a much more cost-efficient model that is still, you know, far capable to do the job. All right. Next, number seven, compute becomes currency. All right. You know, Nvidia CEO Jensen Wong said something like this. Obviously, this is what his company does.

But I think it's the truth. You know, there was a famous, you know, Dario Modi interview, the CEO of Anthropic, I think it was from 2023, where, you know, he was being asked about compute and investing in these data centers. And, you know, he essentially, said, you know, we're going to take it, we're going to be cautious because we don't want to

overspend. And what if the AI demand is in there? And now obviously, Anthropic is having to kind of pay for that underestimation of AI because they're having to overpay, you know, and have all these other partnerships. And they just aren't able to keep up with demand, which is why their servers and services are down more than anyone else in the industry. You know, on the flip side of that, Open AI invested heavily early. on and they're able to serve things at a much higher uptime, uptime rate. But ultimately, I think as more and more workflows, right, we're still early in the game, right?

If you are listening to this, if you're using AI every single day, we are in that 1%. We are in the bubble. Most people are not there yet. So as the rest of the world catches up and says, oh, wait, this AI thing is pretty essential, right? It's more, you know, it's as essential as, you know, the internet. It's as essential as electricity. As the rest of the business world, you know, stumbles across this over the next few years, I think scarcity, right, hardware scarcity becomes a real thing, you know, but the same time, I think at that point you say, oh, okay, we understand it now. It doesn't matter how good your model is if you can't

run it, right? And I'm saying, you know, compute in the same way, the companies like Anthropic, Open AI, Microsoft, Google, Amazon, meta, et cetera. But in the same way, compute to run things locally, right? That does become the new currency. Because if you, you know, are, you know, a tech, you know, if you work in a big tech company and, oh my gosh, well, we actually have servers, well, all of a sudden, you have the ability to print money, right? Because everyone wants that compute.

All right. Let's keep going. Number six, kind of the prediction here. for the rest of the school year. AI controversies are going to hit your backyard. They're going very local. So between local data centers and AI deepfakes that I think you're going to hit really hard in the elections, I think AI controversy is going hyper local.

We've already started to see it with the data centers. And let me just put this out there. There's a lot of bad information out there about data centers. I'm not going to rally on that. I think there's plenty of people having those conversations. I'm just trying to more focus on the practical. of AI, but I think it is going to become a dinner time conversation at some point this year between, you know, the AI data centers and, you know, certain communities, you know,

saying they don't want them, certain communities welcoming them in. It's going to create, I think, you know, income disparities in places that maybe didn't see it before, swinging both ways. But I think ultimately, where it's going to hit hardest is local in state politics, y'all. This is my previous life. I was a, you know, a politics reporter for the Chicago Sun Times. At the big level, right, you know, presidential race, everyone knows when, you know, AI is being used.

Even AI that looks real or sounds real. Locally, I think, right, so statewide races, I think we're going to see dozens of stories about people using AI and a lot of people aren't going to figure out. until it's too late, but for bad reasons, bad actor AI, right? People, you know, creating fake audio of their competitors, you know, videos, you know, that they're going to use in ads that didn't actually happen. It is going to become commonplace. And I don't think most people will be able to tell the difference, which is why I think the

AI controversies are going to be hyper-local. All right. Similarly, well, I think the AI backlash is going to become mainstream. So as the shifts start stacking up, I think it's going to start with the work slop avalanche. And I think it's already kind of started. And a couple of recent developments are going to think, I think, exasperate this. So, you know, as an example, Anthropic kind of announced to adhere to some EU regulations that they're going to start putting in some invisible watermarking in text. So even if you copy and paste it and try to do all.

this stuff, you know, but essentially, I think that kind of movement has also led to the resurgence of the, you know, AI content detectors, which I'm not a fan of because most of them are not good. But I think we are going to see this weird paradigm because every single company is rightfully so, you know, trying to implement AI from top to bottom. Yet, you know, now I think more than ever, you're going to see people getting called out for using AI, right? Whether it's, you know, like this clawed watermarking or some other, you know, AI text detection. But there's going to be, I think, a huge backlash against using AI.

I even found myself, right? I was, I was looking for what was it, a shirt or something online. And, you know, I noticed that they were using all, you know, AI models. And it's like, I knew that 95% of people weren't going to be able to tell. It wasn't the best AI images. But, you know, I saw that. I'm just like, nope, next. Right.

So even me, right? And people always think that I'm, oh, AI, everything. And I'm not. Right. It's just like, I think it should always be tastefully done. But I think the AI backlash, because it is more and more accessible and more and more people are going to use it, it's going to hit hard. But I think it's going to go mainstream.

Like I talked about this years ago, but I think we're going to start to see it where, you know, people are going to start putting things out as human made. And that's going to become like a trendy thing. But I think the AI backlash is going to hit hard. Number four, and I have no clue what this means, but one of my predictions is going to be that math is solved, right? I still don't quite understand how there's all these dozens of math theories that, you know, people, you know, the smartest mathematicians in the world has spent their careers on. and they've been unsolved for forever, right? People have been working on these problems for hundreds of years. I think math gets solved, right?

I don't know what it means, right? As an example, the unreleased GPT, whatever number it was, Astra, which is their next family of models on top of Seoul. It knocked out like 10 problems with published proofs. But I think math in general is going to get solved. And that's going to lead to a. lot of things that are above my pay grade of understanding. Obviously, great things, medicine development, protein synthesis, all these other things that can lead to positives. But I also do assume

that when math gets solved, and that's going to be able to, I think, on your open models that can run on consumer devices are going to have math solved probably within the year. So I'm guessing there's going to be some downsides to that as well, right? I don't know. like cracking crypto, right? We're still probably a way off on that. But I think that there's going to be some downsides to math being solved. But in general, I think math benchmarks are no longer going to be a meaningful benchmark. And if anything, this should just tell you how good AI has gotten.

Because three years ago, right, when Chad ChbT kind of, you know, started to become consumer mainstream, it couldn't consistently solve two plus two, right? Now I think it's literally going to solve math. All right. Next three, token maxing coming back. All right. So this kind of ties in with one of mine earlier about intelligence is coming too cheap. But essentially, right, token maxing where it's like, let's spend as many tokens as possible.

Was very much in vogue in, you know, like December, 2025 through like February. And then there was the whiplash of, oh, wait, no, we can't anymore because, whoops, we're spending way more money than we thought. We are trying to encourage employees to use AI, and they were all spending billions of tokens, you know, spinning up useless things. So now there's been this shift toward kind of value maxing or token efficiency. But now I think we're going back, right? Just like the mom jeans, you know, they were hot. They were gone. They're back.

I think token maxing is already going to come back. And, I mean, look at the last week. I mean, how could you not? Look at Luna. I am spending more tokens than ever, right? Because the tokens are getting cheaper. All right. So that's why I think token maxing is actually going to return, right?

Experimenting, running more and more agents. So I think we went through this weird phase of like, you know, three to five months where it was like token budgets are getting tight, right? Spend is getting tight, which was reactionary. It was not looking forward into the future because if you were. looking forward in the future, you see intelligence is going to come down. Right. So this, this, this concept of token maxing or, you know, or sorry, the concept of token efficiency or value maxing still hasn't fully caught on, right? So it's already made a complete cycle. I think now. So get ahead of it. Because I think still this, this, the, the token

reckoning hasn't fully hit outside of like Silicon Valley, right? It has. It has. has in some places. So if you're listening to this get ahead and, you know, if someone's, you know, some smart guy comes in, Bill comes in on Tuesday and he's like, all right, well, I don't know if you guys know this, but tokens are getting out of control. So we have to, you know, talk about efficiency. You can say, all right, Bill, you know, it's no longer February 2026. The tokens are actually cheaper than ever. And the tokens now, compared to what they could do, you know, four months ago, night and day. Token maxing, coming back. book it. Number two, the open agents are going to crash in 2027, right? This is the full school year.

All right. We're going all the way into May, May or June 2027. Open agents are going to crash everything. I talked about agent crashes on my 2026 AI prediction and roadmap series, which obviously they've been crashing, right? We had all the stories, the open AI hugging face, Kimmy, K, Anthropics models. So the agents are crashing right now, but the open crashes are what's going to really matter. So I had a show on this the other day, if you want to go back and listen to it. But I think essentially, the open models today aren't good enough to cause chaos yet

in bad actors' hands. But the open models that we should be getting around quarter four, they will be because those models are going to be fable level. Those models are going to be, you know, Astra, potentially Astra level, at least GPD 56 soul level. So what does that mean? That means that they're going to be smart enough or conniving enough when bad actors are using them to be able to skirt around the guardrails that are built in. And obviously with open models, once you release them, you can't rein them in. You can't take, you know, Kimmy K3-1 or, you know, Quinn 3-9, you can't take them.

offline, right? Once they're out, they're out. And obviously, because they're open models, you can fork them, you can fine-tune them, you can do what you can to bring the guard reels down. And I think that is going to happen at scale. Just FYI. So what does that mean? I think cyber defense in cybersecurity, especially of the AI variety, it's going to be the fastest growing sector in 2027. And it's not even going to be closed. It is not even going to be close. Assuming, right, that it's going to happen at the rate that it's going to happen, I mean, even small businesses, you know, SMB's smaller enterprises that normally wouldn't

have a strong cybersecurity presence. I'm not saying that they're going to go hire a cybersecurity team of 10, but all of a sudden, it's going to be a budget item. If you're not going to hire people, you're going to have to hire a third party company and it is not going to be cheap. So start penciling that in it now, the everyday average business. right? You're a little, you know, not saying little, but you know, your average company with 200 employees, 500 employees, right, that have maybe never been the victim of a cyber attack. When agents, when open agents are going to be able to find vulnerabilities in your CRM vulnerabilities in your payment systems, they will be able to wreak havoc. You have to be prepared for it now. And then number one, last but not least, and this kind of ties a lot of the stuff together. RSI. is on the horizon. Recursive self-improvement. I also had a recent show about this, but actually some more

recent developments. So Sergei Bryn, one of the most famous people in AI ever, who now seems to have a little bit more, you know, he's always been at the top of the Google AI food chain, but now it seems like, you know, with some recent leadership shakeups at Google, you know, he might have a little bit more, I wouldn't say sway or pull, but it seems like he's going to be a much more consequential figure in Google with some of their other big names gone. But just recently, he just declared that recursive self-improvement is a top priority for Google Gemini and kind of directing and wanting companies to go there. And we've already seen this. You know, we've already seen glimpses of this. I've already mentioned, you know, GVD-5-6 soul as an example, making

it's smaller models, more optimized and cheaper. And we've already seen stories that, you know, companies say that they're models that are coming out now, where, you know, whether going through post-training or, you know, they're not always giving the details. But today's models are starting to help build tomorrow's models. And with recursive self-improvement, that makes everything else. The other 18 on this list go that much faster. All right, that's a wrap. I'm going to go through these one more time, rapid style, so you don't forget.

Number 19, reactive chat is going to start dying and proactive agents are going to take over. Number 18, voice mode and mobile become the default AI interface. Number 17, manager threads are going to replace one-off chats. Number 16, multiplayer AI. Humans and agents are going to share one workspace. Number 15, vibe operations. The whole company is going to start building. Number 14, everything you make goes agent native.

Number 13, skill reuse becomes a company metric. Number 12, company reasoning data finally surfaces as gold. Number 11, public leaderboards die and private evals are going to take over. Number 10, model routing becomes the norm and built in. Number nine, intelligence gets cheap and anthropic gets squeezed. Number eight, Fortune 100s token spend slows while usage explodes. Number seven, compute becomes currency. Number six, AI controversies hit your backyard.

Number five, the AI backlash goes mainstream. Number four, math is solved, whatever that means by AI. Number three, the return of token maxing. Number two, open agents crashed in 2027. And number one, RSI is on the horizon. All right, that is a wrap for today's show. But remember, a quick announcement. Now you're ready.

I gave you the answers. And maybe some of this was a little over your head. And you're like, wait, I'm kind of new here. I need some of the basics. Well, go back to the basics. Because like I said earlier, we're taking a little break from our normal Monday to Friday schedule to replay the entire in restart the start here series. So this is a series made for both beginners and people who are using AI every single day and leaders in their organization. we're to play episodes 1 through 30 in order starting Monday.

So if this got your brain turning and you're like, okay, these are some good things for our company to talk about. Or if you were like, wait, I need to know more about these open models. I need to learn more about, you know, some of these topics on agents that you talked about. Well, in the start here series, we tackled most of these things all in depth at a higher level. detail, more examples, more use cases for your company to grow. So that's it. Make sure to go back to school with us starting Monday. So that's a wrap. I hope this was helpful. If so, please, if you haven't already subscribed to the podcast, then go to your everydayaI.com. Thanks for tuning

in. See you back tomorrow and every day for more everyday AI. Thanks y'all. 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.