Jorge: 0:00

Before we start running off to implement efficiencies, it might be good to take a step back and ask, “What are we doing here? What is this in service of? What’s the mission here? And how might the technology help us do things that might not have been possible before?” I think that’s really exciting, Harry.

Narrator: 0:23

You’re listening to Traction Heroes: Digging in to Get Results with Harry Max and Jorge Arango.

Jorge: 0:33

Hey, Harry. Good to see you again.

Harry: 0:34

Hey, Jorge. I don’t even think about commenting on the fact that I’m wearing a red shirt again.

Jorge: 0:40

Well, I’m just sorry that I’m not wearing one of those red shirts. We’re gonna have to do something about this.

Harry: 0:45

I’m working on it.

Jorge: 0:46

I did promise that I would tell you my size, and I haven’t, so it’s my fault. Um,

Harry: 0:54

I’ll show you just again, just in case you have forgotten what was on it.

Jorge: 0:57

Well, and for folks listening in, what Harry’s talking about is, he has made a beautiful, bright, red T-shirt with the Traction Heroes logo on it, and I’m feeling very jealous because it looks really good. Um, but that we’re not here to talk fashion, Harry.

Harry: 1:20

Yeah, a T-shirt is fashion. Go ahead.

Jorge: 1:25

Yeah. I, you know, as is our norm here, I’ve brought a reading.

Harry: 1:30

Excellent. I can’t wait.

Jorge: 1:31

All right, here we go.“The vision of a fully automated future is hard to resist. It’s a story written into our cultural imagination from science fiction, the promise of lower costs, higher productivity, and competitive advantage. These promises feel closer than ever as automation-enabling technologies advance at an unprecedented pace and scale. At present, this push is largely driven by generative AI development. Generative AI systems are increasingly able to perform tasks that were once considered exclusive to humans. With a few natural language inputs, generative models create everything from high-fidelity prototypes to working code to your headshot in the style of Studio Ghibli. These systems are only getting better, potentially a lot better.” I’m gonna skip ahead a little bit. “Beneath this momentum is a rarely examined assumption. Human thinking is a process that can be separated from the thinker and reproduced by a machine. If that assumption is true, then full automation is just an engineering problem on a timeline. If it isn’t”—and then there’s a parenthesis here—“decades of cognitive neuroscience research suggest it isn’t,” close parenthesis, then the destination of “automate everything” needs rethinking.” And automate everything is in quotes there.“AI’s momentum is also no longer confined to software or knowledge work. It’s rapidly taking physical form. Stroll down any street in San Francisco, and there’s a good chance you’ll see more than one autonomous vehicle shuttling its passengers to their destination. Companies like Figure AI, valued at forty billion dollars in 2025, are designing humanoid robots for both industrial and domestic tasks. At the same time, Fei-Fei Li’s World Labs is building, quote, “spatial intelligence,” end quote, the capacity for AI systems to perceive and interact with the three-dimensional world, much like LLMs interpret and generate text. Meanwhile, companies such as ElliQ are creating AI companions to support aging populations. These advances make fully autonomous systems feel not just possible but inevitable for technology companies. If you’re not building agentic workflows or humanoid robots toward this end goal, you clearly haven’t gotten the memo. Under this pressure, quote, “automate everything,” unquote, often becomes the organization’s focal point rather than one of many possible ways to execute on strategy. This is how market pressures to automate can crystallize into a reactive vision of a fully automated future.”

Harry: 4:32

Okay. That is not a book I have read.

Jorge: 4:40

I would be very surprised if you had read it because it’s a book that is not out yet, so I cheated here. Uh, y-

Harry: 4:48

Very clever.

Jorge: 4:50

Yes. Well, but it might be out or close to coming out by the time this episode airs. I did read a preprint of this book. It’s a book that I’m excited about because I think it’s very relevant and very timely, and I wanted to bring it to our conversation because it does bring up some interesting points here. It should be out in November 2026. So again, it won’t be long after this episode airs, and I believe it can be pre-ordered now.

Harry: 5:26

I’m wondering if I’ve read a couple of chapters in it, but haven’t read the whole book.

Jorge: 5:31

Well, you might because it’s being published by our shared publisher, Rosenfeld Media, and it is called What Your Machines Should Do by Stefanie Hutka.

Harry: 5:45

I’ve read a couple of chapters of this book, but I did not read that.

Jorge: 5:50

Well, I picked that particular passage because first of all, I think it’s one that kind of captures the spirit of the book and what it’s about. And also because I think it captures a very common sentiment among business leaders. I think it’s becoming tempered, but certainly over the past couple of years, the drive toward implementing AI-powered solutions has felt like it’s being driven by FOMO. You know? Like there’s this almost paranoid urgency to get there before our competitors do. Organizations have been experimenting with technology, If I was being unkind, I would use the word flailing with technology, because it’s very clear that this new technology is really powerful. It has tremendous potential, and we must automate all the things. And as that passage from the book suggests, this has been done at the expense of really thinking through how the technology might benefit us strategically. It’s almost like let’s put the technology forward and then see what happens, as opposed to let’s be really mindful here and think about what the technology might make possible for us. So that was... That’s kind of my opening salvo here, and I’m wondering if you have thoughts about that.

Harry: 7:44

Well, I mean, it’s funny, the thing that jumped... I had a number of thoughts, but the thing that jumped into my little brain here is the Scroll back to November, December of 1994, right? I had a full head of hair. There are lots of people that are probably listening to this podcast that had not been born. The internet was the internet and not the World Wide Web yet. It was just starting to form at that point, and I was building Virtual Vineyards, which became wine.com. Which turned out, in retrospect, to have been the first online secure shopping experience, which launched in January of 1995. That was seven months before Amazon, to give you a sense of timing. And the criteria for automating anything at that point was if I had to do it by hand three times, I had to figure out how to automate it, because the number of things that were growing that I was having to do by hand was increasing at a rate that it became very clear that if I didn’t stop at three times and figure out how to automate something, pretty soon all of my time and energy was gonna be executing manual tasks, and I would not have time to problem solve and to problem identify and to innovate and to create the things that ultimately became a world-class shopping experience. So the thought that came to mind in, in this frame of automate everything is I think everything needs to be defined because there’s everything that’s known, there’s everything that is known that is reproducible, there’s everything that’s known that’s reproducible that has low variance. There’s... What does everything mean? Because I think it’s fine to say,“Okay, I have... There is an activity that is predictable, reasonably consistent, and it requires a fair amount of manual intervention.” Machines are better at that, right? Automate it. But as you start increasing contact with the real world and other human beings and other unpredictable aspects of the environment, that notion of everything starts to look like a world of very high variance that requires much more creativity, much more situational awareness, and a much more adaptive approach to interacting with it to determine if there is, in fact, something that can be repeatable and is consistent enough and has to be... and is going to be automatable in a way that will reduce the effort on the humans who are having to engage and increase the quality of the experience and the outcomes that derive from that. So that was really the thing that jumped out. It’s that classic case of, you know, best practices are all about things that happened before. They’re not about things that are going to happen in the future because these practices are... Best practices are the best of the practices, right? And that means that we’ve experienced them. But innovating against the incredible unfurling of possibilities as we contact the real world is you-- there, there is no best version of that. And even the practices themselves have such high variance, it’s hard to know which ones ought to get our attention. And those are not easily automated well.

Jorge: 11:58

Yeah, I don’t wanna speak for Stef, but I suspect she would agree with you. Of this book is about what might be worth automating and understanding the role of automation in an organization. Very much to your point there, not everything is worth automating. And it might even be worth questioning the premise of as a goal in itself,

Harry: 12:33

Mhm.

Jorge: 12:33

right? And the book has a sidebar. It actually happens just before the passage I read, where Stef cites Shoshana Zuboff, who we’ve talked about before on the podcast. Zuboff makes this distinction between automating and informating, which is a word I had not heard before. And it’s a different use of technology where, you know, if you can... gonna probably not do it justice, but the idea as I understand it is automating is about replacing the labor of humans, whereas informating is about using the technology to give the humans kind of superhuman abilities. So maybe exposing the workings of the system, giving them greater control, greater agency. And when we— you know, when faced with new technologies our natural response is to want to project them onto the ways that we’ve worked in the past.

Harry: 13:45

Mm-hmm.

Jorge: 13:46

and imagine how they might help us do things in the way that we’ve done in the past. And ultimately, over time, people find what the technology is actually good for. And there’s this William Gibson quote that I’m gonna mangle, “The street has its own use— finds its own uses for things,” or something like that, right? Greg Petroff, my partner at Unfinished, and I interviewed Stef about this book. I’m going to include a link to that conversation in the show notes here, where we get into these questions a bit deeper. But the reason I wanted to bring this to our conversation today, and this is very self-serving, it’s because Greg and I have been talking about this, you know, about this notion that the, the first wave of applications for this technology seems to have been almost exclusively focused on efficiencies, on gaining efficiencies, right? And that, you know, when I hear automation, like I’m thinking, you know, do the thing that we were doing before. Like you said, it’s like, you know, I’ve done it three times by hand. It’s like I don’t need to be doing the same thing three times by hand. Let me up a little batch script or something that will do it for me. That’s about efficiencies, right? It’s about doing the same thing that I was doing before, but doing it faster, doing it more cheaply with lower energy or what have you. And That is probably the most boring use of this technology. You know, the real question here is, now that we have something like relatively inexpensive cognition, what new possibilities open up? And I’ll give you another one. Now that we have the ability to make things, software things especially, faster and cheaper than before, what are we going to make, you know? These are not technology questions. These are strategy questions. And it might be that AI can help us with those too, but I do think that before we start running off to implement efficiencies, it might be good to take a step back and ask, “What are we doing here? What is this in service of? What’s the mission here? And how might the technology help us do things that might not have been possible before?” I think that’s really exciting, Harry.

Harry: 16:38

I, I agree with you. In fact, I, you know, I’m working on a number of AI projects right now, and one of them is with a world-class... Interestingly, is with a world-class bookstore. And we’re looking at the application of artificial intelligence and new technologies to the next generation of the bookstore experience. That’s both online and physical store. And this question of automation and efficiency has come up over and over and over, but what’s really interesting about it is this very... This almost gravity-level tendency to head toward efficiencies seems to be, in this case, very much about buying the blue sky time so that the people who are mired in the operational day-to-day can free themselves up enough to even begin to have the mental space to have the conversations about what’s possible in the future. And so for this particular bookstore, because I’m very deeply interested you know, low-hanging fruit from moderate investment, from significant investment, from cost savings and efficiencies to revenue generation and blue sky, so that we could place every project into a grid and look at where the leverage was actually gonna be. And all of the excitement was on the blue sky stuff, but they couldn’t get there without figuring out how to get themselves out of this corner they’d painted themselves into with so much work that they have to do day in and day out. So I think there’s a... I think there’s two things going on, at least, right? One is, let’s make a better saddle for the horse so that we can have a faster horse problem, right? That’s the application of technology to the known and how do we make that better, right? I just want a faster horse. No, the automobile and the jet airplane and all these other things, people couldn’t even imagine them, right? And so there’s that challenge. It’s the imagination challenge. It’s the strategy challenge. And the other challenge is how do you free up the mental space and the creativity, to buy the right to do that, to spend the time to explore when the overall, the overhead of the operation day-to-day is consuming not just the, you know, the capacity that you have to do the work and the capacity that you have to do the thinking, but the excess capacity to burst into having more freedom, having more time to work on these things as they could be, rather than just making better what you have.

Jorge: 20:00

I’m gonna say something good about that approach before I tear it down. Um,

Harry: 20:05

Okay.

Jorge: 20:06

I think that the good thing about that approach is that by focusing on efficiencies to free up time for... You used blue sky thinking. I’m gonna use it in the terms I used to really think about where we’re going, right? Like the strategic thinking about what is it that we really wanna do here? Focusing on efficiencies to free up time to do that is a really good way to learn about the capabilities and constraints of the technology because you’re not thinking about it in the abstract. You’re actually applying it and learning what it can and cannot do well. The downside to that is that say that you succeed in designing and building the factory that produces the smarter horse, now you have a business that is producing smarter horses when the world has moved on, you know? There’s a metaphor that I had heard before. Stef brings it up in the book, which is with the introduction of electric motors when electricity first started becoming more widespread. Before electricity, factories were powered with steam engines, and steam engines are bulky, expensive, dirty, hot, relatively dirty. They’re large, especially at that scale. Like you can’t really have a small steam engine and have it, you know, produce the necessary energy. So the way that those factories were structured was there were big central engines that were then their energy output was decentralized using belts and stuff that would convey energy elsewhere. When electricity first came online, the instinct was to replace those big steam engines with big electric engines. But electric motors have different capabilities than steam engines. In fact, nowadays, you can get electric motors in a toothbrush. You know, you could not get a little steam engine to power your toothbrush. And the right answer to that question wasn’t how quickly can we get the electric motor to power the existing factory? The right answer ended up being how can we redesign the factory to account for the fact that now we have the capability to decentralize the production of power, right? My sense and part of the reason why I wanted to bring up this reading is that if you’re operating under a mindset of paranoia and urgency driven by the fear of falling behind your going to devote the time needed to both learn the capabilities and constraints of the technology, which I said it might be good to actually start there, right? But to actually think about what is it that we can really do now that it’s become possible to have these technologies that work unlike anything that we’ve had before, really, you know, at that scale? Yeah, I think that there’s some strategic soul-searching in

Harry: 23:43

I mean, when I was at Silicon Graphics, you know, the company that did the systems that produced the movie Jurassic Park, I was there in the late, you know, in 1989 or early 2000s. Is that right? Yeah, that’s right. God, I know that... I’m a decade off. That’s really scary.’89 and the early ’90s. Whoa, that’s frightening. I learned a phrase that is now part of the way people talk about things in Silicon Valley: violent agreement. That was the first time I ever heard that one. Violent agreement. We had people debating all sorts of stuff in these meetings, and they’re like, “We’re in violent agreement.” Because I think that’s 100% right. We... There’s no question whatsoever that we have to take a step back, and we have to look at what’s possible. And what I would add to what you’re saying is that you were talking about what do we wanna do here? And I wanna emphasize the word do, because I think it’s not just about what we wanna do, I think it’s about who we wanna be. And I think when we start looking through the lens of aspirations and we start expanding the vision of possibility, it opens up the possi- it opens up the question of who do we wanna be, and it’s not just what do we wanna do. And then, then this gets very interesting. So the prioritization model that I shared with you a few minutes ago was really about taking a step back and trying to look more generally at the array of possibilities from trivially easy to incomprehensibly expensive and potentially intractable along the spectrum of the context of that business and its purpose for being. It has operational responsibilities, and it has to operate at a certain level of efficiency, no question. And who do they wanna be in the future? And what does that mean, and what does it imply for what they wanna do and where they could be focused and how they might want to apply these new technologies? So I’m in complete agreement with you. Violent agreement.

Jorge: 25:58

Well, violent. Hey, well, and to lean into the word do there, you know, when we ask who do we want to be, I would append to in the things we put into the world? The products we sell, the experiences we offer, the services we provide, right? So this is not something to be left in a boardroom whiteboard.

Harry: 26:24

Right.

Jorge: 26:24

You know, this is something that needs to be invested in. You need to be making things that move how you show up in the world given the capabilities of the technology now. And the answer to that shouldn’t be, “Hey, we’re now, you know, doing the same thing, but we’ve laid off 30,000 people so we can do it faster.”’Cause it’s like that’s... First of all, that’s good in general, right? But also it’s astonishingly unimaginative, right? Like you’re just imagine what could you do if you empowered those people, you know, the informating versus automating thing. If you empowered those people to do new things, given the capabilities of the technology. I think that that’s much more exciting and interesting a world than the one where we’ve made everything more efficient by taking humans completely out of the loop.

Harry: 27:31

Yeah, I agree. And I, you know, in this bookshop I’m working with, the, owner of the bookshop is all over what’s possible. They know the efficiency stuff matters, and they wanna spend as little time there as they possibly can. And that, I found that very inspiring.

Jorge: 27:49

Well, good for them. And I love reading, and I might last time you recommended a book that I wanna get, so maybe we’ll talk offline and I’ll buy it from them.

Harry: 27:59

Oh, that’d be awesome. All right, man, it was so good seeing you again. Thank you so much for sharing that.

Jorge: 28:04

Well, thank you, and it’s always good seeing you, too. And hopefully sometime soon I’ll be wearing a red shirt.

Harry: 28:12

All right. Talk to you later.

Jorge: 28:13

Okay.