Subscribe
Share
Share
Embed
What happens when Ai gets better at doing the work, but relying on it too much starts weakening the human skills that make your work valuable?
In this episode of Capability Amplifier, I sit down with my good friend Steven Kotler, bestselling author and founder of the Flow Research Collective, for a fascinating conversation about Ai, creativity, flow, and human performance.
Steven has spent decades studying the brain at its best, and he has some strong opinions about how we're using Ai.
His biggest rule is simple: the human starts first.
Steven believes you should think, learn, create, and develop your own point of view before bringing Ai in for feedback and refinement. Otherwise, you risk handing over some of the mental work that develops judgment, memory, creativity, and original thinking.
Steven and I also dig into why so much Ai-generated communication feels the same, how Ai changes purpose and mastery inside organizations, and why taste, storytelling, attention, and intuition may become even more valuable as the technology improves.
In this episode, Steven and I break down:
- Why the human should start the thinking before bringing in Ai
- Where Ai works best as a feedback and iteration partner
- Why predictable Ai-generated writing can become easier to ignore
- How Ai can affect purpose, identity, and mastery inside organizations
- How I use Ai to rapidly prototype ideas and possible solutions
- Why expertise and human judgment still matter
- Why taste, creativity, attention, and storytelling are becoming more valuable
One of my biggest takeaways from this conversation is that the better Ai gets, the more valuable human judgment becomes.
Ai can help you move faster, explore possibilities, and get feedback almost instantly. But you still need the expertise, taste, creativity, and intuition to recognize what's worth pursuing.
That's where I believe the real opportunity is.
Want more from Steven? His Instagram is the best place to keep up with everything he's working on: @stevenkotler. And if you want to go deeper into flow and learn how to access it on demand, check out his Flow for the Many program here: flow.stevenkotler.com
TIME STAMPS
[00:00:00] Why Ai-Generated Communication Can Become Predictable
[00:03:08] The Biggest Misconceptions About Ai
[00:09:01] Steven's Rules for Using Ai
[00:12:20] Why the Human Should Start First
[00:17:03] Why Ai Writing Tends Toward the Average
[00:25:12] Bringing Ai Into Organizations
[00:28:02] Purpose, Identity, and Mastery
[00:33:46] Rapid Prototyping With Ai
[00:38:56] Where Ai Prototyping Can Break Down
[00:45:30] The Limits of Ai With Complex Information
[00:50:39] Why Human Skills Matter More
[00:58:29] The Future of Flow Research
PS – When you’re ready, here’s how I can help:
Want to find the hidden revenue in your business? Grab a Cup of Coffee with me: AiAccelerator.com/1k- Ready to reinvent yourself, your business, and your brand, and create “Your Next Act”? Watch this.
Discover More
📘 Ai Accelerator Book: https://www.MikeKoenigs.com/AiBookFree
🎉 Referral Party: https://www.MikeKoenigs.com/RPFree
🌍 Your Next Act: https://www.MikeKoenigs.com/YNAFree
🚀 Private Experiences: https://www.MikeKoenigs.com/PEFree
🌟 Follow Mike
🎙️ Podcast: http://www.MrBz.com/CA
📺 YouTube: https://www.youtube.com/koenigsmike
💼 LinkedIn: https://www.linkedin.com/in/mikekoenigs
📘 Facebook: https://www.facebook.com/koenigs
📸 Instagram: https://www.instagram.com/mikekoenigs
🚀 X: https://X.com/mikekoenigsreal
Creators and Guests
Host
Dan Sullivan of Strategic Coach
Dan Sullivan is founder and president of The Strategic Coach Inc. A visionary, an innovator, and a gifted conceptual thinker, Dan has over 40 years’ experience as a highly regarded speaker, consultant, strategic planner, and coach to entrepreneurial individuals and groups.
What is Capability Amplifier?
Join the eternally curious, interested, and interesting hosts, Mike Koenigs of the SuperPower Accelerator and Dan Sullivan of Strategic Coach®, to amplify your capabilities, value, status, and authority on the Capability Amplifier podcast. Ever episode focuses on a new mindset, shortcut or deep thinking exercise that will improve your performance and lifespan. Learn more at: https://www.CapabilityAmplifier.com
Steven Kotler [00:00:00]:
The myth is that if we get better training data, if you just trained an AI on the greatest writers in history and blah, like all that stuff, it still won't get better. It's an average machine.
Mike Koenigs [00:00:12]:
Yeah.
Steven Kotler [00:00:13]:
What's happening is you're reading AI-generated copy, your brain knows it's AI-generated at a really deep level, and it tunes out the familiar patterns. Yeah. So you've just now lost all ability to communicate or persuade.
Mike Koenigs [00:00:25]:
Probably over the past 6 months, what are the mindsets that matter most? It's being able to see between the cracks. Not what's straight in front of you. Everything is all about knowing how to get attention, stand out, differentiate, be a great storyteller, and synthesize data and information really, really quickly and be able to develop a strong intuitive sense that is sensitive towards humans.
Steven Kotler [00:00:53]:
I am personally not a fan of agentic AI, and I think anybody, any leader who's trying to learn how to use agentic AI right now is stupid.
Mike Koenigs [00:01:00]:
And the reason is My guest today is my good friend Stephen Kotler. You might recognize these books, We Are as Gods, Bold, Abundance. He's co-authored with Peter Diamandis. He's written 17 books, 13 bestsellers. He's also had 3 Pulitzer nominations. and he runs the Flow Research Collective. So when you hear the word flow, that's Steven. Now, he is one of my favorite people in the world.
Mike Koenigs [00:01:40]:
He's also a terrible, terrible influence on me. But we talk today about the biggest and most effective ways to use AI from a mindset, brain flow, and team perspective, and where everyone wastes the most time that cuts into their mental health, and how to avoid having AI train your brain negatively, something he calls metacognition. Cognitive Degradation. Now, his frame for this entire episode is what are the soft skills of the 20th century— attention, taste, purpose, creativity, flow— how they're the hard skills of the 21st century. But his angle, the way he thinks about these things, are what makes this interview so fascinating. So if you're using AI heavily in your business, this one's going to sting a little bit because it's supposed to. That's why I asked Steven Kotler to be here. And here we go.
Mike Koenigs [00:02:32]:
All right, Steven Kotler, my buddy, thank you for being here. I just wanna dive right into the big lies that have been sold. You're a journalist, you know how narratives get created and where the public's been actively misled. So I would be curious what your perspective on, on the number one misled idea or notion dealing with the future in AI that is active and in motion right now.
Steven Kotler [00:03:08]:
So I think it's the AI apocalypse, that AI is going to come for our jobs. I think that's the— I, I don't know if that's the biggest lie. That's a— like, people are making outrageous utopia claims about AI that I think are, are very silly. Um, the one that I actually think is, is actually probably dangerous for business, possibly for policymakers, for government Okay.
Mike Koenigs [00:03:42]:
And next, you've spent over a decade, you've done a partnership with Peter, written books. The most recent is We Are as Gods. I am curious, when you look back at Abundance and Bold, We Are as Gods, where do you think You have been wrong looking back.
Steven Kotler [00:04:06]:
Oh, that's easy. I can tell you exactly where we've been wrong. We've got a really good, uh, our track record. So let me back into this. I always tell people when it comes to my job, I don't see myself ever as a futurist. That's a very specific thing. Ray Kurzweil is a futurist. He makes predictions about the future based on an algorithm that he developed and a way he thinks about the future.
Steven Kotler [00:04:31]:
That's what Rhea does. I try to be a reporter. I'm very good at spotting trend lines and looking at where trend lines could intersect. Bold was that kind of text a little bit. And that was why I think in We Are as Gods, Peter and I agreed to disagree in public where we normally have disagreements in private, and we try to find a middle path. We're like, no, we're on different sides on population. We're on different sides on AI. We're on different sides on longevity tech.
Steven Kotler [00:05:04]:
Um, not completely. We're— let me just— different sizes may be a strong word. I think we have different timetables for development for the— these sorts of things. So I think the better way of saying it, where we were wrong along the way. First, I think the classic, the, the one you got to point at and laugh, is we opened Abundance, the book that kicks it all off, with Masdar, the city of the future that they were then building in the Middle East, and it never Is that what became NEOM? Pardon me?
Mike Koenigs [00:05:37]:
Is that what became NEOM? Or is that—
Steven Kotler [00:05:40]:
Yeah, ish.
Mike Koenigs [00:05:41]:
Sort of ish.
Steven Kotler [00:05:42]:
Yeah. So that didn't work as planned. The one that, it was funny 'cause we actually had a debate about was biofuels. But we had Craig Vettner, and I remember sitting in his office listening to him talk about biofuels. And when the guy who decoded the human genome tells you we're gonna have biofuels, it's hard to argue with him. And at that point, nobody was arguing with him. The science really looked solid. The progress was there, and it really looked like, hey, we just gonna scale this up a little bit.
Steven Kotler [00:06:19]:
And it turned out that wasn't true at all. To me, the great perpetual puzzle slash disappointment in a sense is VR, 'cause I, it's something I've been involved with since the '90s. You know, and I, I, I knew the Mondo 2000 folks back in the early 1990s.
Mike Koenigs [00:06:39]:
Yeah, you and I have had a big connection with Mondo. Yeah, yeah.
Steven Kotler [00:06:43]:
Even, even Jaron Lanier, at, at that point, the, you know, often called the godfather of VR, he was dating a woman named Kathy Acker, who's a very famous author before she passed away, and I was her intern. So like VR was like really in my world, and that was something— oh, it's just around the corner. And It's been just around the corner, so much so that— and there's a— I just read a phenomenal book on VR and where is it going, but it's— and it was, it was really well done, it was really neat, and nothing in it is technically wrong, but I don't believe it at all. I just don't believe it at all. I, I think I've been fond of saying the metaverse exists nowhere other than Mark Zuckerberg's magic underwear.
Mike Koenigs [00:07:27]:
Right.
Steven Kotler [00:07:28]:
I, and I, and, and the reason I'm so harsh on that is I don't, I really am tired of technology leaders standing on stages and lying for marketing purposes. Um, it's really like, it's real. And I see it. I mean, I, you know, I look at AI and I, my brain goes NFTs, metaverse, right? So I still think a big chunk of what we're looking at falls into that category, which is not to say I don't think the technology's amazing and super useful. And, you know, unlike, say, VR, which— though VR has turned out as a tool for scientific research and a whole bunch of other things to be fantastic, is it widespread? No. Is it specialized? Yes. Like, I remember meeting Skip Rizzo back in the early 2000s. He was at USC, and he was using VR, uh, sort of early work that, like, I've built on a little bit, but they were using VR to treat PTSD in veterans.
Steven Kotler [00:08:28]:
And It was working gangbusters, still does. Really sort of neat, neat work there. So, um, BR, I think, is the one we missed. Biofuels for sure. Um, and, uh, Masdar, which, you know, morphed and changed. And I— and, and— but I still don't— I don't— I still don't think we have this city of the future today. I don't think we have anything You close?
Mike Koenigs [00:08:59]:
No.
Steven Kotler [00:08:59]:
So I think those are the misses.
Mike Koenigs [00:09:01]:
Okay, big ones. All right, um, I'm gonna do a microcosm macrocosm question. Um, so if you remember, I think it was probably a year and a half ago, you and I were at Joe Polish's house and we were hanging out and we started doing a whole bunch of AI stuff. Um, I was demonstrating my methodology for using it and creating. We had a pretty deep connection of just like, oh, a workflow. I want to know what your behaviors are, your own AI behaviors. If you walked me through a yesterday, a last week, last month where you are using AI and then where you deliberately refuse and either you are all analog— And what your decision matrix is or your first principles for where you've maximized it and also where you've got hard rules where you don't use it because it breaks or it breaks you.
Steven Kotler [00:10:13]:
So, Let me start where I love it, where I dislike it, where I never use it. You want me to start on the negative?
Mike Koenigs [00:10:25]:
Sure.
Steven Kotler [00:10:26]:
So the first thing, our work at the Flow Research Collective, one of our 4 lines of research is in human-AI collaboration. And this has been true for 10 years. And in a sense, when I say AI isn't coming for our jobs, when we started this conversation, one of the reasons I feel like I can make that statement, whereas there are a lot of statements about AI where I don't feel like I have any expertise, is when I started the Flow Research Collective, our mission, one of the things I wanted to use was flow for worker reskilling in the face of AI. That was— so we started, we started there. I originally started— the first problem I was interested in solving, and it, and I'm still interested in solving the same problem, is most truck drivers in America, the largest blue-collar workforce, are going to lose their jobs over the next 15 years to autonomous trucks. Takes about 15 years for a truck fleet to go extinct. Autonomous trucks are starting to be purchased now. So 2045, 2050 or so, We have to reskill the largest blue-collar workforce in America.
Steven Kotler [00:11:41]:
And it is clearly going to require flow because flow amplifies learning and performance and productivity, all the things you would need. So we have been looking at this puzzle for a long time. And one of the most important things about how the brain works and how memory works is— and we all know this— it's associative. We have an associative cortex, right? One thing links to the next, links to the next, links to the next. That's how memory works. That's how your intelligence works. That's how your creativity works. So I'll give you this high level.
Steven Kotler [00:12:20]:
We'll go drill into creativity later if you want to. But high level is whenever I start a project, whether it's science, writing, a slide deck, whatever it is, the human always starts the project. Yeah.
Mike Koenigs [00:12:35]:
Mm-hmm.
Steven Kotler [00:12:36]:
I work for a while, long before I bring the AI in. Because if you let the AI start the thinking, you've locked your own associative cortex out of it. You're starting with whatever idea the AI comes up with, and suddenly you're locked out of the equation. And you are— and the AI might be getting smarter, but you're not actually getting smarter because even if you learn stuff from the AI, it doesn't tag into your memory as easily. So you've really hurt yourself. And this is doubly important if you're writing, building a slide deck, or you're going to be teaching people, anything like that. If you don't do the hard work of the writing yourself— writing is how we organize information in the brain, and it's how we expose the gaps in our thinking. So if you don't do that yourself, you're not going to see the gaps in your thinking.
Steven Kotler [00:13:30]:
You're not going to get get where there's not enough detail where the AI could actually really help you, you're going to lock yourself out of the process and it's going to take probably 50 times longer than you think it is. Because you, you know, even if you don't know what you want, you have a vibe, something of a target already as a general rule, and the AI is going to miss completely. So, and when I say you have to start I also mean you have to know your subject, 'cause AI works best when you have a generalist. I say generalist, that's—
Mike Koenigs [00:14:06]:
Yeah.
Steven Kotler [00:14:06]:
A loose term. What I really mean is AI works best, best if you are trying to become a polymath. You have expertise in 5 or 6 domains. That's who AI is really great for, um, 'cause it allows you then to bring in a bunch of things the AI doesn't have to the equation. So that you have to start by Learning the stuff yourself, like really learning it, then generating the thing you're trying to generate. I only use AI for feedback. I don't use AI to create ever. It's a feedback and it's phenomenal for feedback within a— if you understand how tight that constraint window is, how dumb the machine is, how much it can't remember a damn thing, and how much because it wants to please you.
Steven Kotler [00:14:57]:
It's going to hide all those things from you. It's going to cover up and try to pretend those things don't exist. Those, I think that's my first and foremost use. And to pull back, I will say I have found that when you're inside of semi-bounded knowledge bases, math, coding for sure, Mm-hmm. Some of the sciences, depending on what you're trying to do, AI is fantastic, but it's truly a terrible writer. Like, it's truly a terrible writer. It cannot write at all. And we could go into, like, the second biggest lie people are living under right now is that AI is good at communication or persuasion or, like, should be used as— Like, marketers are who Not surprisingly, are doing this so wrong.
Steven Kotler [00:15:59]:
It's just, it almost cracks me up. But I sit in meetings with, with people who do info marketing sometimes, and I listen to their ideas and I think, wow, you really don't know a thing about how the human brain works and you're doing everything wrong. But like, I'm like, a part of me doesn't even want to say anything because I'm like, I'll take the advantage. Like, you guys just keep doing it wrong. It's fine.
Mike Koenigs [00:16:18]:
Yeah. Yeah, I think that—
Steven Kotler [00:16:21]:
And we come back to what I think that is, but I'll just pause because I've been talking for a little bit and see where you want to go.
Mike Koenigs [00:16:27]:
Well, I think what we've seen, especially in marketing and communications, is the level of gray that has erupted. So, you know, if you— let's just assume that this conspiracy is true. that the AIs have been training on a huge body of work, a lot of mediocrity, and a bunch of synthetic garbage that's based on the mediocrity and not weighted properly, right?
Steven Kotler [00:17:03]:
Like, I gotta tell you, that's, that's a— you're, that you're perpetuating a lie here about AI, which is— I mean, yes, all those things are, are probably true, but that's not the problem. And that will never be the problem. This is why LLMs will always fail here. So very simply, if anybody who's using AI in marketing and business, you're doing it for one of two reasons. You're doing it to communicate or you're doing it to persuade, right? Those are the two reasons you're using AI, right, for you as general rule. And AI is an averaging mechanism.
Mike Koenigs [00:17:43]:
Right.
Steven Kotler [00:17:44]:
It's LLM optimization. It, like, it standardizes output. It runs towards the middle. So a great human sentence, and I'll give you examples in a second, is absolutely predictable, absolutely predictable, absolutely predictable, absolutely predictable, totally surprising. So the most famous sentence in 20th century literature could be, and a screaming comes across the sky. It's the opening sentence of Gravity's Rainbow. An AI could never write it. Because they would never— screaming doesn't come across the sky.
Steven Kotler [00:18:14]:
That's the shocking word. That's why it's in great piece writing. I'll give you another one.
Mike Koenigs [00:18:20]:
I swear to you, by the way, I thought I said the exact same thing you're saying in a different way, which—
Steven Kotler [00:18:26]:
Oh, that's funny.
Mike Koenigs [00:18:26]:
Yeah, yeah.
Steven Kotler [00:18:27]:
Oh no. Yeah, it's the—
Mike Koenigs [00:18:28]:
So I didn't articulate it the way you are with case studies and examples. Yeah.
Steven Kotler [00:18:34]:
What starts to happen is one, you lose the ability to generate that great writing. Here's the worst part though, and this is what people miss. So your brain is a giant pattern recognition engine. That's what neurons do at a really basic level. They recognize patterns. AI, LLMs, any machine that's built to go towards the middle is gonna write in predictable patterns.
Mike Koenigs [00:18:58]:
Yep.
Steven Kotler [00:18:58]:
The famous one that we now recognize as AI slop all over the place is not this Not X, but Y.
Mike Koenigs [00:19:05]:
Yes.
Steven Kotler [00:19:05]:
Right? Not X, but Y is a standard AI construction 'cause it's everywhere in language. It's very easy for the AI to write that way. And so now when you see that, you go, oh, that's an AI. That's not a human. The thing that most people don't realize is your subconscious recognizes those patterns long before your conscious mind. So if you're familiar with this very famous test in neuroscience, the Wisconsin Card Sorting Task, I don't have to go into it, but like Antonio Damasio Massio, like, made his name on this experiment. It basically shows your subconscious can spot patterns like 50% of the time faster than your conscious mind can. So what's happening is you're reading AI-generated copy, your brain knows it's AI-generated at a really deep level, and it tunes out the familiar patterns.
Mike Koenigs [00:19:56]:
Yeah.
Steven Kotler [00:19:56]:
So you've just now lost all ability to communicate or persuade because your brain is tuning out all of the copy. So if that's what you were trying to say, then I just did you in other language.
Mike Koenigs [00:20:07]:
Yeah, it's, it's, um, I think what I was effectively saying is the— there's a massive level of gray, which is the undifferentiated slop, which is an average of not just mediocrity and average, but now a pool of synthetic average, which weighs it, right?
Steven Kotler [00:20:31]:
The myth is what I jumped at, Mike, Okay. The myth is that if we get better training data, if you just trained an AI on the greatest writers in history and blah, like all that stuff, it still won't get better. It's an average machine.
Mike Koenigs [00:20:46]:
Yeah.
Steven Kotler [00:20:47]:
It's an averaging machine, right?
Mike Koenigs [00:20:48]:
Yeah.
Steven Kotler [00:20:48]:
Like it doesn't, it's not gonna, it's not gonna get— this is also one of the reasons LLM progress is slowing and, um, we'll never go from an LLM to superintelligence or even a general-purpose AI. Like, if that gap is not gonna close, it just widens over time, which is what we're seeing.
Mike Koenigs [00:21:10]:
Yeah.
Steven Kotler [00:21:11]:
So I think there's a lot of things that come off of this that make me suspicious about some of the AI narratives we're being fed.
Mike Koenigs [00:21:18]:
Yeah, I, uh, I think this is, uh, this is where we start slicing and where the gaps in time actually get wider, meaning— I'll use the fastest parallel, which is, in the early days of the Internet, no one saw Amazon coming and no one saw exactly where credit card processing was going to come from. A lot. The market got confused and looked at one thing and what seemed impossible became possible. In the world of AI, I believe that we'll look back 5 or 10 years and there are going to be some things that seem absolutely obvious that are non-obvious and simple that are massive direction shifters. That's what I anticipate is something unexpected right in plain sight. This is one of those things that's, you know, if you'd say, well, what could we have predicted are going to be the biggest problems? I don't think they're obvious until you're right in the middle of the noise. I don't know if any of that made clear sense, but that's— That's looking back at the last 20, 30 years is, you know, if we all could predict, had a crystal ball, we'd all be filthy rich. And it just doesn't work that way.
Steven Kotler [00:22:57]:
I think you're going to get black swans, that's for sure.
Mike Koenigs [00:23:00]:
Right on. All right. I got the next one for you. And I want to set this up a little bit because I want to set you up and get your perspective because this is a real-life use case. I've got 2 use case studies. One of them is, I want to speak from a corporate perspective. Part of this is, I'm going to be training an organization in a different country that has a lot of employees. I can't be more specific than that because of some NDA business.
Mike Koenigs [00:23:37]:
What I've found in most organizations is there is a sense of, Up on high, we know we need to be using AI and there isn't a choice any longer, but the leadership isn't sure exactly where to put it. Of course, the devil is in the details and you start moving in. Then there's the psychology of implementing AI and it goes beyond the fear of losing a job and displacement. But if you were walking in to do training, what would be the hard truths that you know to be true that you'd want to instill inside the team and the organization? Before you answer that, I'll give you a couple, for example, that I've for sure learned. A, you and I have talked about this before. AI doesn't have good taste for all the reasons you just described. Humans network really well. When we brainstorm together, we create together.
Mike Koenigs [00:24:50]:
AI can be a great creative companion, like you said, a feedback machine. I'm curious, like some of your fundamentals. You walk in and say, here are the hard truths, hard expectations, hard yeses, hard nos from a scientist, a journalist point of view.
Steven Kotler [00:25:12]:
So that's interesting. You know what? We have in We Are as Gods, there were 7 commandments for working with AI. I think I still hold of those. I can go through those if you want. I think the one— the place I would start is— and I'm gonna— I like to— I like to explain this because we all went through this with social media. So do you remember when Facebook showed up?
Mike Koenigs [00:25:45]:
Yeah.
Steven Kotler [00:25:45]:
Might have been MySpace or Friendster for you, but for most people it was Facebook when this happened to them. And for like the first 2 weeks you're on Facebook, you felt amazing. Like you were high as a kite, you were getting likes, you— but it was really like, I'm connected all over to all these people. Oh my God, I will always work again. I will always be able to get a date on Friday night again. Like it was that you were so safe. And 2 weeks later, 3 weeks later, you started to get that ick feeling in your gut. You'd work, you'd be on Facebook and you'd walk away and you'd feel a little crap, a little lonely, a little weird, a little like you ate too much junk food or watched too much television or whatever it was.
Steven Kotler [00:26:26]:
It felt like felt icky. And we all felt it. You're nodding. And we ignored it. We ignored it. And what we were actually ignoring was the start of the largest mental health crisis in history.
Mike Koenigs [00:26:39]:
Yeah.
Steven Kotler [00:26:39]:
That's what we were ignoring. That's what that was. So you get the same signal from AI when you're working on AI, when you've gone a little too far. When you get that signal from AI, it's a sign of one of two things. You are either cognitive offloading. So currently then the machine is thinking for you and taking your reins. Um, or you're cognitively overloaded.
Mike Koenigs [00:27:01]:
Yeah.
Steven Kotler [00:27:01]:
And you're just driving yourself into burnout. So it's funny that my first rule for working with AI, this super machine, but very in line with what Peter and I wrote in We Are as Gods, is you actually have to get really good at interoception. You gotta get really good at your own internal signals, um, and detecting them to work with AI. So that's sort of one of the, one of the things that I want to talk about. And if you're working with corporations, let's just talk about like, or go into like my AI laws, because maybe you can look them up in the book.
Mike Koenigs [00:27:37]:
Yeah, yeah.
Steven Kotler [00:27:38]:
But the things that aren't, we don't talk about in the book, and you probably know this because you've done this work with organizations. One of the first things that happened is, um, if you are used to handling every step in a project, and AI suddenly comes in and is handling every step in between, you're doing less. And doing less can make you feel like less.
Mike Koenigs [00:28:02]:
Mm-hmm.
Steven Kotler [00:28:02]:
Can make you feel a lot less important. That's scope collapse, right? And that's very common. Um, purpose drift. So when AI completes— when we complete a task, we get dopamine reward, right? When an AI completes a task, we don't get the dopamine reward, not at the same level, not the same way. And in a weird way, the brain adds up those little gold or dopamine rewards that add up into meaning and life satisfaction. That's the— and purpose. That's— those are the ingredients. So you start to get purpose drift or identity drift, right? And this— a lot of people are feeling this anymore where they don't— what am I? Who am I in this new world? And what I'm doing doesn't mean what it used to.
Steven Kotler [00:28:44]:
And, and why am I doing this? And then, um, There's a mastery vacuum because we are hardwired to get better over time. And so, with mastery back on, that's actually really easy. When that one shows up, leaders have to remember it's their job. AI is going to create these vacuums. Purpose is going to drift, scope is going to collapse, mastery is going to have a vacuum. It's the leader's job to fill them. And people literally, Because we have loss aversion, we're going to hold on to what we have and stick, right? It's a leader's job to let people know, hey man, the AI handles this stuff now. These are the new skills you have to master.
Steven Kotler [00:29:31]:
So you have to like, and AI did this for you so you can do this. It's a leader's job because these things are going to happen automatically. So This is what we're seeing inside of organizations, even organizations who get the AI right. You get it perfectly, all this happens. There's a lot of ways to get it wrong, make yourself stupid, really like do dumb stuff. The only other one bit of advice that I really like that I, that I just think is so useful and so many leaders don't seem to get this. I am personally not a fan of agentic AI, and I think anybody, any leaders trying to learn how to use agentic AI right now is stupid. And the reason is we know user-friendly interfaces are coming.
Steven Kotler [00:30:19]:
This is when you asked earlier, you were like, what are we gonna look back on? And it's user-friendly interfaces. When a technology goes through the 6 Ds of exponentials, the standard lifecycle exponentials, it goes from digital to deceptive to disruptive.
Mike Koenigs [00:30:34]:
Mm-hmm.
Steven Kotler [00:30:34]:
And what's the hinge point? A user-friendly interface. Internet was around forever. Netflix or Netscape showed up, suddenly we were all on it. Right? Like that was user-friendly interface. Facebook, user-friendly interface for social media. It wasn't the best. It's actually the most astoundingly mediocre piece of crap you could build, period, ever in the history of the universe. It's like if you could bottle beige and turn it into a product, you get Facebook.
Steven Kotler [00:31:01]:
And that was Zuckerberg's genius. He just, he took all this funkiness and went, no, no, we're going to give the world beige because everybody can agree on beige. Nobody loves it, but we can all agree on beige, right? And like, that's what happened. And even Bezos, by the way, like you said, Bezos, like nobody saw Amazon coming. I think I find that surprising because I thought early internet, like the one, the 3 things you could see in the early days was, oh, this is great pornography distribution.
Mike Koenigs [00:31:29]:
Yep.
Steven Kotler [00:31:29]:
This is some kind of encyclopedia.
Mike Koenigs [00:31:32]:
Yep.
Steven Kotler [00:31:33]:
It would be terrible in the beginning, better over time. Right. And it, we, and we're gonna be able to sell shit mostly cuz our buddy Dave had figured out how to sell wine.
Mike Koenigs [00:31:41]:
Right.
Steven Kotler [00:31:41]:
And, but it was like really clear. that people were going to be able to sell stuff. Bezos just came in and did what Dave Asprey never bothered to do, which was, okay, wine is expensive and hard to ship and it has problems. What's really cheap and easy to mail and everybody wants lots of? Oh, books. Dave knew wine, I think, is why he went into wine. I don't know why he chose that as the first thing to sell online, but I never asked him, but that stuff to me was like, you— I saw that coming. This was a predictable line of development out of like what we saw early on. But it was a user-friendly interface.
Steven Kotler [00:32:27]:
And AI, we think it is a user-friendly interface because ChatGPT, that's just a user-friendly interface for search, basically. User-friendly interface for agentic AI. And Grok just gave us a— I think they just introduced one, you'll probably know this better than I am, like a week ago. That's okay.
Mike Koenigs [00:32:43]:
Mm-hmm.
Steven Kotler [00:32:44]:
Definitely better than doing it from scratch. But that, like, people are building Jarvis. You're gonna be able to talk to the machine in 6 months. So really, if you, if you wanna spend the next year figuring out how to build agentic AI in systems that aren't gonna be around a year from now, and you're gonna like work for 9 months to get what, like a 1-month advantage or 2, like that's where That's another silly one that I see a lot. I don't know. Am I answering your question?
Mike Koenigs [00:33:10]:
Yeah, good enough. There's gold in them thar confetti balls. I kid. I kid. Next one. And this is a practical use case because this has been where I've had the most success with AI. I would like to get your point of view on how you see it working and the why. Again, I'm going to ask you to put on your journalist hat, your scientist hat.
Mike Koenigs [00:33:46]:
I'm going to walk you through a use case of where I've had the most success using AI with teams. We've been calling them super sprints. The basic premise is, I can meet with an organization and have a future-paced conversation. In other words, what would you love for your organization? What would a great future look like? It just opens up with an open-minded future and a little diagnosis. If you could have the world the way you want it, what would it be like? Followed by what— this is very Dan Sullivan-esque— but what are the dangers that are preventing you from getting there? Just make a big-ass list, all the obstacles preventing you, all your superpowers and strengths as an organization, and some evidence of that. In other words, I'm looking for use cases, case studies. Talk about products, services, et cetera. And then what I've been doing is taking that vision and saying, telling AI, I want to make a future-paced video about that better future and name these things.
Mike Koenigs [00:35:10]:
So it's effectively a trailer for the better future. And I have a belief that if you make a movie, and make it real and make it entertaining. I open up all of my shows with a video story. I develop every product with a video story and then move into acquiring data as a generalist, just gathering information, gathering information, and then using AI to prototype multiple solutions simultaneously. For example, one of the use cases I did when I spoke at NASA is I met with NASA and I said, what are the biggest challenges you have? I said, what's your big vision? We want to get back to the moon by 2040. What are all the problems you're having? They said, well, we have a list of 140 of them. I said, send them to me. Great.
Mike Koenigs [00:36:03]:
Then we did a prompt and said, well, pick one and tell me a way we could solve one of these problems. It turned out respiratory problems were one of the key things in space. respiratory system starts to fail over time when you're outside of gravity. And we talked to AI a little bit for a possible solution, and it suggested measuring cough response. So you do like a— and basically what we did is over just a period of about an hour, came up with a core challenge, a vision, made a video that told a story, and it was about a respiratory analysis software product. Then I told AI to design the software, and I fired up 10 different code generators simultaneously, each prototyping a solution. Then we looked at them to pick which one it did the best. One of the great things about code generation now is You've got all these different products.
Mike Koenigs [00:37:12]:
They write code so well. They write user interfaces. They create tools. Then we just looked at them and narrated and described what we loved most about each of them and fed all those instructions back into the AI to make changes to make all of them better. Effectively, it was It was crowdsourcing story, crowdsourcing problem, crowdsourcing an app. But, over a period of around half a day, we had a vision, a story, and a product. Then that also became a training for how to think differently and it got marketing people, bureaucrats, engineers aligned on Wow! There's a different way to solve a problem. That was maybe a gross oversimplification, but that's a big part of what I've been doing lately is just walking into an environment.
Mike Koenigs [00:38:16]:
I don't have to understand the business by asking good questions, generalizing, prototyping, creating a vision, and then refining. It's a matter of iteration. That's where I find AI works really well. I want to reflect back to you because I'm curious from a neuroscientist's point of view, as a creative, as a journalist, what you see there, what's your interpretation, and what about that works? Where does it break? I know it's a really complex question, but I've wanted to ask you about this for a long time.
Steven Kotler [00:38:56]:
You asked me to consult and you're consulting, Mike. And one, I don't actually know if I'm qualified to do that. Two, I wanna— so one, I think the story you just told is a lie.
Mike Koenigs [00:39:13]:
Okay.
Steven Kotler [00:39:13]:
'Cause you don't come out the other end with a product. You come out the other end with a map to a thing that you can call a product when NASA builds it, tests it in space with astronauts, and it comes back as real. That's a, that's a thing.
Mike Koenigs [00:39:29]:
There we go. So yeah, it's an idea. Okay, good.
Steven Kotler [00:39:32]:
Fair enough. So I mean, if you were to ask me where do I think AI is awesome, it's as a scientist. I know it's— so I'll tell you where AI is not awesome as a scientist, where it breaks first, and then I'll tell you, like, if— because they're bounded knowledge bases. So I work, right, on— I'm a computational neuroscientist working on, on flow and altered state and consciousness and things that are at the very edge where we have a line between this is reality and this is total make-believe. But this is reality is a very flexible, dynamic, fast-paced industry. So the AI is not trained on any of that. So when I go to work with AI trying to work on the cutting edge of things where we are, It often doesn't know. It often will be too conservative in a sense.
Steven Kotler [00:40:30]:
Like, it will be like, well, that's not real. And I'm like, no, no, here, give it this paper. Here, give it this paper.
Mike Koenigs [00:40:35]:
But it's—
Steven Kotler [00:40:36]:
the knowledge isn't really in the paper. I can't feed it in one-to-one. So it tends to miss that. And if I wasn't really well-versed and didn't know about that research, right, you'd miss that.
Mike Koenigs [00:40:48]:
Yeah.
Steven Kotler [00:40:48]:
So it's a little wonky when you're at the cutting edge of problem solving. And I don't, That may be more user error than machine error. I also want, like, on that front, like, I— we— but we get phenomenal, like, phenomenal results. Um, as a scientist, I love working with AI because it's, it's amazing. Um, I get nervous when it starts to spin up ideas because there's just gaps, and it's really good at hiding the gaps. So— Uh, that, that is there. I don't— I would say, just listening to this, I'm just— I'm all over the place. You've asked me where I think the holes are.
Steven Kotler [00:41:30]:
I think the movie as a whole. And the reason I think the movie as a whole is— by the way, if you would've asked me 6 months ago, I thought I would've told you the movie is a strong suit. But now, like, I— yesterday, I'm sure you've done this recently, gone to Peter's website and seen all the trailers for all the new Abundance Forward. I don't know if you know he's doing this, but it's right—
Mike Koenigs [00:41:51]:
Yeah, the film competition.
Steven Kotler [00:41:53]:
Right. So, and you know, they're very— a lot of the trailers are very good in terms of, you know, they— but they all are AI-generated, and they look a certain way, and they all feel bloodless, and they creep me out. I don't wanna watch them. I don't like watching them. They don't— like, there's nothing sort of there that speaks to me. It's— I'm really turned off by it already. Kids are turned off by it already, right? Like, it's— we're— the uncanny valley really shows up a lot.
Mike Koenigs [00:42:26]:
Oh, huh.
Steven Kotler [00:42:26]:
There. So the whiz-bang out of the movie is gone because we've all seen it. And I like— now, having given people a way to see their future and vision into their future, super important, right? Like, we all— like, that's really important. So I'm not criticizing the methodology. The methodology basically taught us how to do it. to me. I'm just wondering, is the movie as effective today as it was a little while ago? Because, right, so this isn't like— the methodology is right, but the movie, I'm like, you asked me to poke holes at it, and so I'm poking holes at it. Um, and I don't know if I'm right, but this is what I would look at.
Mike Koenigs [00:43:07]:
Yeah, yeah.
Steven Kotler [00:43:08]:
I think, um, um, the, uh, the other, the other, and the other thing again is like, but this is If you don't realize how fast you can invent with AI and iterate and cycle, yes, you're absolutely missing out. So if companies, to me, I mean, it might've been you, it might've been Michael Mintz, whoever it was when I first started playing with AI, they literally, somebody came to me and was just like, would you just sign up and play? Just play.
Mike Koenigs [00:43:42]:
Yeah.
Steven Kotler [00:43:42]:
Boom. You, somebody, I mean, like, it was very early on. And as soon as I did, I was like, oh yeah. And that was the funny thing, 'cause I, I laughed. I was like, so prompt engineering is actually a, a, it's a fake job for people who never bothered to open up an AI.
Mike Koenigs [00:43:59]:
Mm-hmm.
Steven Kotler [00:44:00]:
Right? Like prompt engineering is this, this fake thing I can sell you until you bother to open up and type a sentence into AI. And then that job doesn't exist anymore. it's a joke. Yeah, right. And you saw that sort of right away. Well, we—
Mike Koenigs [00:44:14]:
if you— I don't know if you remember this, um, it was right around the time we sat down at Joe's and—
Steven Kotler [00:44:22]:
I think it was— yeah, but I don't think it was— well, I think it was actually like 3, because by the time we sat down at Joe's, I had been playing with AI for a while. And so I think it would— I think it was— I went to a couple of Joe's events right in a row.
Mike Koenigs [00:44:36]:
Okay.
Steven Kotler [00:44:36]:
Um, and I think it was the event prior to that, like 6 months prior, where you were like, just play. Because we had been using it, machine learning, so like as AI in the— on this, as a scientist.
Mike Koenigs [00:44:49]:
Yeah.
Steven Kotler [00:44:50]:
I just hadn't bothered. I just didn't know it was anywhere near ready for prime time in terms of the LLMs. And you were like, just play. And you were right, right? I was like, that was— you, you play your way into knowledge. The other thing that I always sort of want to— I don't know how this shows up in what you just walked us through. The last thing I'll say is this. One of the reasons I think people don't understand how limited systems are is because they're not routinely doing what I'm doing, which is working on a 500-page document.
Mike Koenigs [00:45:29]:
Yeah.
Steven Kotler [00:45:30]:
with tremendous amount of detail. And if you work on a document that big, the AI's blind spots— what it can remember, what it can't remember, where the holes are, what it's going to try to cover, what it will lie about— like, all that stuff becomes incredibly visible. The advantage of, like, doing it with a giant book that you know that you've written, where you actually— is it's, it's sort of like having a bounded knowledge set. So I can I know when I go in as a scientist that, you know, it, it's got 500 pages of science that it's, you know what I mean? But it's gonna have the same limits it has. It's just that I have such intimate knowledge of the 500 pages that I've written that I can see it. And I, and maybe I got this, like, I got it the hard way because I have a giant novel. It's 1,000 pages. And I haven't, you can't use the AI to ...to really write fiction at all.
Steven Kotler [00:46:23]:
It's really sort of terrible at that. But I was just trying to use it to edit it. Like, is this plotline— you know, if you're writing thrillers, every plot is— all the details are really important. And being able to figure out, like, when you're starting to move things, when you're editing something that big, you're like, oh, if I change— if Mrs. Lincoln stabs him with a fork instead of a knife in scene 2, what does it change at scene 7?
Mike Koenigs [00:46:47]:
Right.
Steven Kotler [00:46:48]:
And, and that you have to track that stuff and know that stuff. And I went in thinking, oh, the AI's got this. It's like, this is what it's built for. And it doesn't have it at all.
Mike Koenigs [00:46:56]:
Yeah.
Steven Kotler [00:46:57]:
So I write, and that's like, that's not a normal experience for most people. Most people don't hold 500 pages worth of content in their head and then move it around. So it gives me, and probably every other author, a view of what it— the tech can do with the knowledge base and what it can't do. That's the other hole in what you said. And I don't— I don't know how, other than saying what I just said out loud, I can't prove it to you. And you've heard me literally on Moonshots, on Peter's podcast, like, have this argument with those guys. But I was like, literally, are you using a different system than me?
Mike Koenigs [00:47:38]:
Mm-hmm.
Steven Kotler [00:47:39]:
And it's what I'm seeing that they're not seeing is that.
Mike Koenigs [00:47:42]:
Yeah, yeah. Ultimately, um, everything you described is a matter of, of context, priority, and weighting or weighing the data. And that, that I think is what all these folks have learned. Yeah, are some massive— there's huge limitations and we've got to make orders of magnitude shifts. And just compressing the data isn't the way to get there because you lose detail. You lose the detail. And what your brain does, what you've tuned your brain over your history as a journalist to do is know what's important, what isn't, what can be— Well, let me put—
Steven Kotler [00:48:29]:
And I know we're getting so damn nerdy right now that That's, uh, no, but let me add, let me add one more thing to this to go actually one level farther, because I think it's worth saying, because you— any people can have this experience. So I— there are certain science textbooks, all of Csikszentmihalyi's fundamental work on flow, for example, a number of like 3 or 4 or 5 different kind of foundational neuroscience textbooks, whatever, where I've read the, the works Multiple times, right? I don't know how many times I've read Csikszentmihalyi's book Creativity, but— or his, uh, uh, Flow and the Future of Adult Development. Like, I've read those books a lot. And when you read a text, any text, the first time through, your brain focuses on really what's in the center of the paragraphs because that's where the facts are. And the first time through, you're learning the facts. And usually it takes like 1 or 2 repetitions till you actually, oh, I've got all this in memory already. Scientists hide all their wild-ass ideas in the transition between the facts. It's like what'd be in between paragraphs there though.
Steven Kotler [00:49:39]:
You'll get like these throwaway lines where they'll step outside their hard science role and, and, and speculate a little bit, or this might be the case. And when you start reading text 3 or 4 or 5 times, you start noticing that between the paragraphs, which is actually where all the information lies. So the problem with AI and the human brain is also that AI is gonna grab the middle of the paragraph. That's all it's designed to show you at all costs. Unless you tell it to go hunting for the thing that's between paragraphs, it will never— ever find it, right? So that's another sort of limitation here. And what's interesting is I say limitation, I am not talking about the machine. The machine's amazing. The— you— it's the limitation on the human side.
Steven Kotler [00:50:36]:
And you have, you have to change your behavior.
Mike Koenigs [00:50:39]:
Yeah.
Steven Kotler [00:50:39]:
Not the, the machine. The machine is gonna do what the machine does. Um, it's useful, it's not useful. You, you have to really change your behavior. That's what I think is so interesting about the tech, the current tech revolution is the fancier, faster, more whiz-bang it gets, the more we have to really double down on really basic human skills. The way I like to say it, Mike, is that the soft skills of the 20th century have become the hard skills of the 21st. What I mean by that is, so I entered the business world, I got outta outta high school in '85. I got outta grad school in '91 or '92.
Steven Kotler [00:51:20]:
And if I would've gone into a meeting right, right then and there and been like, you know, you know what this company really needs? We need passion, we need purpose, we're gonna need some creativity and flow. And let's talk about attention and emotional regulation. I would've gotten laughed at. And if I would've brought it up again, they would've been like, dude, you're really fricking annoying. And a third time, And depending on the company, I could have gotten fired, right? Like literally that was the boardroom in the '80s and the '90s.
Mike Koenigs [00:51:48]:
Yeah.
Steven Kotler [00:51:48]:
And now those are mandatory because it's not just that these things are needed, it's that AI, accelerated tech, et cetera, they're eroding those very skills.
Mike Koenigs [00:51:59]:
Yeah.
Steven Kotler [00:52:00]:
So if you're not actively training them, you're, you're going backwards at a really fast rate. I think, okay, so here, let me give you, we're already, I told you before we began that I'm working with the head of the UT, University of Texas Law School, some epidemiologists, and we're writing an article about how we think AI should come with warning stickers, 'cause why they can produce depression and anxiety and a road condition and all that stuff. And I don't think anybody's, that's not much of an argument right now, unless you're an AI company owner. Like, it's not much of an argument for anybody on that one. That's not surprising. I forgot what I was going to say. Forgot the second half of that statement. I had a great lead-in.
Steven Kotler [00:52:49]:
It was all dramatic. Oh, come on, dude. I don't remember what I said. It literally just exited. It'll come back to me in half a second. Ask me another question.
Mike Koenigs [00:52:58]:
All right.
Steven Kotler [00:52:58]:
Before we have to jump out of here.
Mike Koenigs [00:52:59]:
Yeah. Well, I'm going to give you a quick summary because I've got a transcript running. All right? Basically, you really have to change your behavior. That's what the tech revolution is. The fancier, the more whiz-bang it gets. You have to double down on human skills. Then the other thing is, you had been talking about reading between the lines and how all of the important stuff— Before I complete those points, I had another point for you. I can't remember what it was.
Steven Kotler [00:53:30]:
Oh, okay. Yeah, there's one other thing I want to say because you brought it up at the start and I'd like to drill down on it because it's a real cool thing. So I recently did— 2 weeks ago, we did something called the Velocity Weekend where I was really excited because I've spent 20 years training corporations in flow and peak performance and that sort of stuff. Because in the beginning we didn't have as much data as we have now, it was a smorgasbord approach. Like these were all the tools. And, but as over the years I've been able to refine it down, 'cause we've got data and we've got data and we've trained people in 156 countries and 28 industries and we know what works, which tens of thousands of folks at this point. So we did it in a, in a weekend and I built the slides, worked really hard with AI to build the slides. And as I— and people came up to me afterwards and they were like, you know what, this is one of the coolest presentations I've been in.
Steven Kotler [00:54:38]:
What did you do with those slides? That's amazing. How did you do that? And what I thought was funny about it was you couldn't have designed the slides with AI. I agree, you couldn't have designed the slides without AI.
Mike Koenigs [00:54:54]:
Mm-hmm.
Steven Kotler [00:54:54]:
But if the way I designed the slides, I, because I come out of magazines as a journalist, and in the '90s I had the incredible good fortune inadvertently through no luck of my own or through luck, just pure luck, I was in the right place at the right time. I got to study directly under some of the greatest designers, graphic designers, in the 20th century, like giants in the industry. And I got to work with a lot of them, um, often closely, because, you know, in— especially in the early days, I helped start so many magazines, you'd end up working really closely with the design department for layout and everything else like that. And so I got this amazing design education. Amazing. And you call taste, right? When they talk about—
Mike Koenigs [00:55:45]:
Yeah.
Steven Kotler [00:55:46]:
Ah, humans bring taste. Well, you could call it taste. You could say, yes, I have good design taste. Or you could say, I got a fantastic education in design. And when you couple my fantastic design education with what AI can do, if you spend enough time with AI, and it wasn't easy, I spent weeks and weeks and weeks and weeks building this deck. It wasn't faster. It was a lot slower.
Mike Koenigs [00:56:11]:
Yeah.
Steven Kotler [00:56:11]:
But the end result was I gave my audience an actual visceral emotional experience rather than, this is a slide deck. I took 'em on a journey. They had, we had an adventure. They went someplace they'd never been before. And it, right, it was the combination. So I think, you know, the argument for like human skilling up to me goes up as well. It's not right. It's the, those soft skills also, but I think human skills, really start to matter, but they're like gray area skills.
Steven Kotler [00:56:44]:
They're aesthetics, taste. They're things that are not captured. The words that capture them are underweighted in the English language, in my opinion.
Mike Koenigs [00:56:55]:
Yeah, I, I, uh, again, I, I thoroughly agree. That is, those are the— probably over the past 6 months, when I look at like what are the mindsets that matter most, It's— if I'm weaving what you're saying, it's being able to see between the cracks, not what's straight in front of you. This matter of developing taste, and that is, again, being able to see outside of the gray. At the end of the day, everything is all about, in the animal kingdom, Yeah. Knowing how to get attention, stand out, differentiate, be a great storyteller, and synthesize data and information really, really quickly and be able to develop a strong intuitive sense that is sensitive towards humans, none of which is easy to communicate. I don't know if I'm just resaying some of what you're saying. I get it. As we're wrapping this up, I'd like to get your perspective right now.
Mike Koenigs [00:58:13]:
What is exciting you more than anything that your research is focusing on, your business? If you were going to highlight a convergence that Yeah.
Steven Kotler [00:58:29]:
So I, God, you know, like we've got a lot going on. There's, to me, the things that are most exciting is, so as you know, the Flow Research Collective, I've transitioned into a nonprofit and we're going after, we're using, we're basically taking, you know, I've spent decades studying the brain at its best. Now we're just taking what we've learned to study the brain at the best to applying it to situations when the brain is at its worst, right? And we're using flow and altered states to go after burnout, depression, anxiety, PTSD, addiction, even neurological disorders like ALS and TBI. And, um, the reason is the breakthrough— and this was sort of our research that sort of led here, but other people have been poking at it too— is that flow restores network flexibility in the brain. And all these conditions are the— or the brain gets locked into certain patterns of behavior. And the patterns are different, but the conditions, right? Depression, basically your default mode network, which is rumination, just spirals, right?
Mike Koenigs [00:59:37]:
Yeah.
Steven Kotler [00:59:37]:
With trauma, PTSD, it's the amygdala, which is fear, and the hippocampus, which is memory, and they just spiral. Those are right. Flow restores network dynamics, so it's useful in all these conditions. That's that work. Plus, like, we just put out a paper 2 weeks ago, 3 weeks ago, in Neural Image. We had 12 new ways to measure metastability, which is this adaptive flexibility in the brain. So it's this combination of we're right here, the research is right here, the tech is facilitating this in really new ways, and it's starting to feel like mental health is a problem, not distributed, but solved.
Mike Koenigs [01:00:19]:
Wow.
Steven Kotler [01:00:19]:
Right? And so that's really incredibly exciting. I really like that work a lot. That's really got a lot of my attention. Similar research, we have a line of research. I said we have 4 lines of research at the Flow Research Collective. We look at mental illness and neurological disorders. There's just a standard line around peak performance, which we've been doing for 20 years. And the human-AI line that we talked about, about.
Steven Kotler [01:00:47]:
The 4th one is we look at consciousness and intuition and synchronicity and those sorts of things. And where that's really exciting is one of the biggest problems in consciousness, the so-called hard problem of consciousness, Yep. which is why is red red? Why does it— why do things feel like things? And we actually got a paper coming out in about a month that has a viable Tentative partial solution to that problem. It's not 100%, but we got like, we moved it significantly. And now we are taking that solution and testing it with both data from the Allen Brain Institute, mouse data and human datasets, and it's holding up. So that's really just crazy because this is an ancient problem in research that a lot of people have banged on, on the fact that like We think we've moved, kicked that down the road. So those are really, those are very exciting things that I— the most exciting thing is probably, uh, last thing I'll say is Flow for the Many. Because I spent, you know, at Zero to Ageless, our core Flow training when the Flow Research Collective was still a training company, um, was great, but it was really aimed at knowledge workers.
Steven Kotler [01:02:03]:
And it was phenomenal for knowledge workers, but it was intensive. It's a lot of work every day. You had to do a lot and whatnot. It wasn't going to solve the problem I started out with. How do you reskill? How do you use Flow to reskill truck drivers or take a single mom with 2 kids working a couple of jobs or any entrepreneur, right? You know how busy entrepreneurs are. It was really hard time-wise to train them. We built Flow for the Many, which is sort of everything distilled down and sort of— it's a 3 to 6 month, depending on how you want to pace it for yourself, but like a little bit, like a couple hours a week worth of work instead of a couple hours worth a day worth of work, um, that anybody can sort of onboard into their life. And that's being rolled out now, um, and it's, uh, it's significantly, significantly, significantly lower cost than anything else we've built.
Steven Kotler [01:02:58]:
able to put together. And that, by the way, only brought to you by AI, right? We took all the data, everything we learned, and churned it and churned it and boiled it down and really looked at that stuff. And so that's really exciting to me because that's always been a real hard problem. It's funny, I'll leave this as a last thought. When I got into this work, I figured the easiest people to train were going to be athletes because they totally got it, right? And fantastic. And it turns out they're the hardest people to train. They don't have any attention span, and they're so used to burying their feelings and hiding the pain that you can't— it's hard to talk to them and coach them one-on-one because they're bottled up. They're so pressed for time.
Steven Kotler [01:03:53]:
And most people I've met who are really great athletes are medicating, self-medicating with athletics for ADHD. So there's attention issues. Um, and that's a really difficult patient population, just a hard population.
Mike Koenigs [01:04:10]:
Yeah.
Steven Kotler [01:04:11]:
So, and it's a population I've been treating now for 20-some years, 30, you know, forever kind of thing, and working with. And we would— which isn't to say we didn't move the needle for people, but like there was a difference between really moving the needle for people and feeling satisfied with whatever, you know what I mean? I was never satisfied because I was always like, wow, this is a big lift. And all my— like it was just so much work when it should have been easy. So we finally, I think, among the many other patient populations that Flow for the Many will work for, I think we finally We tested it along the way on a lot of athletic populations too, 'cause like they were a training population where I really wanted to work with them. And I was thinking about like, get a chance. I've worked a lot, not a lot, I've worked a bit with college athletes. And I don't know if you've ever spent time around college athletes, they don't have any time. Like they don't, they don't have a life.
Steven Kotler [01:05:05]:
It's so full between their classes and being an athlete. Like it's just full on.
Mike Koenigs [01:05:10]:
Yeah.
Steven Kotler [01:05:11]:
And they're the people who would need flow the most. And we had no way to train them because before, the old method was a couple hours a day, an hour a day, 6 months. They, like, they literally aren't going to have that until they get out of college and are done being an athlete, and they need our help now. So that was— yeah, I'm really excited about this because there were a bunch of different problems I was trying to solve, uh, with this training. We finally Like, got in it. So, you know, and that's available. So those are the really— those are the things I'm doing that are really exciting to me right now.
Mike Koenigs [01:05:44]:
Yeah, that's awesome.
Steven Kotler [01:05:45]:
Plus, you know, nothing says Merry Christmas like We Are as Gods.
Mike Koenigs [01:05:52]:
There, shameless plug. Great book. Yeah, yeah, you're welcome. You're welcome. Stephen Kotler, I love how you think. Every time we ever hang out, It starts as, uh, let's hang out for like 40 minutes. It turns out into wee hours into the night, 5, 6, 7 hours. And, uh, I enjoy every minute with you.
Mike Koenigs [01:06:12]:
So thanks a million for showing up and, uh, being you.
Steven Kotler [01:06:15]:
Thanks for having me, Mike. And, uh, you know, call me anytime. I'll have you come get you out of jail.
Mike Koenigs [01:06:21]:
Oh yeah.
Steven Kotler [01:06:22]:
Done it once, I'll do it again.
Mike Koenigs [01:06:24]:
Thank you. Well done.