When AI Does the Building: Innovation, Ideation, and the New Creative Advantage

00:00:11:01 - 00:00:32:10 Michael Welcome to disambiguation. I'm your host, Michael Fauscette. Each week we interview experts in artificial intelligence, generative AI, and business automation to help business leaders understand how to use these tools for the biggest business impact. 00:00:32:13 - 00:00:44:07 Michael I show today is when AI does the innovation, ideation, and the new creative advantage.

I'm joined by Doctor Alex Mehr, founder and CEO of Famous Labs. Alex, welcome. 00:00:44:09 - 00:00:46:02 Alex Thank you for having me. 00:00:46:04 - 00:01:10:15 Michael I really appreciate you joining.

And I think this is going to be a fun conversation to. But just to get us started, because you have a very interesting background and kind of what I'd say an extraordinary path from where you were to where we are. So, so I'd love to know a little bit more about it. I mean, obviously PhD in mechanical engineering and you've worked at NASA Ames Research Center.

00:01:10:18 - 00:01:24:16 Michael You were co-founder, which I think a lot of people will recognize. And and now, you know, founding Famous Labs. How does that journey unfold for you and what drew you into the AI space? 00:01:24:19 - 00:01:52:13 Alex Yeah.

So growing up, I never really wanted to be. Never thought about even becoming an entrepreneur. My family was very academic oriented. So my goal was to grow up, become a physicist and become a scientist and become a university professor.

So it wasn't really like early 20s in grad school that the thought even crossed my mind. Believe it or not. 00:01:52:14 - 00:02:22:06 Alex And then when I was working at NASA Ames Research Center in, in California, near San Francisco, that's when I kind of through osmosis, I kind of started picking up entrepreneurship language and seeing people that are doing sort of some stuff like that. And I decided to try my hand at it.

And, and it's it's like crossing the Rubicon. 00:02:22:07 - 00:02:48:25 Alex Once you becoming an entrepreneur, you've crossed the Rubicon. And now all of a sudden there's no going back. Actually, I even continued publishing papers because, you know, everybody was telling me, you know, your, you know, research, academic career is going to be jeopardized, blah, blah, blah.

So I even like the first 2 or 3 years of Zeus, I kept publishing research papers or on my website just so I have the option of going back to academia. 00:02:48:25 - 00:02:55:12 Alex But you know how it is. Was he going to stop going back? 00:02:55:14 - 00:03:20:18 Michael It is very difficult to come back from from a lot of reasons, I guess.

But if you're if it's something that excites you, it really excites you. And it's definitely hard to get that out of your blood. I think. Yeah.

Well, so when we were doing our prep conversation, you said something that really stuck with me. He said that, you know, as an engineering major, you always thought of engineering as this superior field of study. 00:03:20:18 - 00:03:40:21 Michael But now, because AI can handle a lot of the execution, taste, creativity, knowing what works in the market are probably much matter, much more than than just the straight engineering piece. Could you explain that a little bit?

I mean, what does that mean for how businesses should think about their competitive advantage? 00:03:40:24 - 00:04:26:04 Alex Yeah, it is a massive shift. And I can tell you as a tech entrepreneur, I've been a tech entrepreneur since 2007 and I had a large engineering team. You know, 2010, 1112.

I wouldn't say engineering is no longer useful. I think engineering thinking and scientific thinking is useful, but but what actually moves the needle is no longer your ability to code or be super technically detailed in a specific aspects of engineering than it is to have a high level understanding. 00:04:26:06 - 00:04:51:19 Alex Plus, this is the key difference what I call taste. Meaning.

To give you a specific example, when I use to hire engineers or when I hired engineers in the past, I would give them technical questions. Right now we don't do that at all. Essentially, we just make sure that the person has actually has some understanding of engineering principles. 00:04:51:19 - 00:05:15:10 Alex But then the real questions are things like, okay, what have you built?

Let me look at it. And then it is like, why does it look that way? Why did this work? Does this work flow?

Why did you do it that way versus this way. So that is a that is not a technically like a technical question. That is a judgment taste question. 00:05:15:12 - 00:05:46:07 Alex So Engineering's job has completely shifted from oh, I know how to code in react.

And I know all the libraries and all that to I can make I understand what a product should look like, what it should do for the user, what's a value proposition to the user? Those kind of things. It's a massive shift. I know it sounds minor, but it's actually massive.

00:05:46:07 - 00:06:14:14 Alex And if someone who's listening to this, they are engineers. I am I am confident this is a trend that is going to continue. In other words, try to hone your common sense, your judgment versus oh, I know exactly how to write this piece of code that does something is very specific. Yeah, yeah.

00:06:14:14 - 00:06:39:19 Michael That's interesting. I mean, last summer, I was I was talking with the CTO of one of my larger software clients, and, and they had been investing heavily in, you know, growing AI capabilities in their tech team, developed development teams. And I and I asked them, I thought, you know, so with this, are you looking at a larger reduction in the number of engineers or, you know, how do you think about that? 00:06:39:20 - 00:07:08:21 Michael Where's the real value in this for you?

And and what about those younger engineers, you know, how do we get them in the pipeline and have them be senior engineers someday? And his answer, I thought, was really good. He said that he didn't believe that the number of engineers should change at all. He just he said his his thought was we need to retrain them and refocus them so that, in fact, what they are is providing much more of a conceptual view.

00:07:08:22 - 00:07:35:25 Michael What do you want this to do? How do you want this to do something? And and, you know, almost in an architectural role too, right. And I thought that was really interesting because obviously we've heard so much noise about, oh, it's going to take, you know, all the entry level development jobs and and of course, you know, if you step back from that and you go, yeah, well if you do that then there's no more senior developers down the road, right.

00:07:35:26 - 00:07:55:14 Michael Because eventually they're going to die. Yeah. It's, you know, it's it's it doesn't make sense. So I thought that was really interesting.

Their approach was very different. It was more about how do you them in a different way or in a new role. So they understand that what they know still applies, but it needs to be applied in a slightly different way. 00:07:55:15 - 00:08:03:01 Michael And judgment I think, is a great word to use their because that that does definitely tie the human piece of it in.

00:08:03:02 - 00:08:40:14 Alex Right? Yeah. So I think there's going to be a short term gap in the pipeline, in the sense that there are still people that are going through, let's say, computer science school or whatever, engineering school, and they're not fully trained on how to leverage AI. And then therefore, they're going to have a hard time coming into engineering organizations because they're looking for hiring senior people that they can be augmented with AI instead of junior engineers.

00:08:40:15 - 00:09:04:10 Alex Does that make sense? But I think that's very short term, and it probably is starting to change anyways because the younger generation is already picking up AI coding tools. Right? And by the time they're ready for the job market, they're already can work like a senior engineer because augmented with AI, they can whatever they don't know they can they can do anyways.

00:09:04:12 - 00:09:18:15 Alex So I think that gap might be real. I understand the concern, but I think it's much smaller than people realize and it probably is already dissolving itself. Does that make sense? Yeah.

00:09:18:16 - 00:09:37:16 Michael Yeah, it does make sense. I mean, I think I think you're right. And from what I've seen, you know, in general, I would say that, you know, the even places where where they did lay off some, you know, some some junior engineers or perhaps didn't hire some people that they had planned. You see them backtrack with that after a period of time.

00:09:37:18 - 00:09:57:08 Michael Right. Oh, we had to hire these back or we had to hire different people. Whatever. Which I think is, is is a testament to that.

And I guess the only the only reservation and you kind of hinted at it too, is that in the, in the education pipeline, if you're not embracing it, that you know, that is a problem. 00:09:57:09 - 00:10:05:19 Michael Yes. And and a lot of it ends up being individual. Right.

Because some institutions now are still kind of burying their head in the sand, as it were. 00:10:05:21 - 00:10:32:14 Alex Yeah. And I tell you something, I think, you know, you would find me as like, middle of the ground. I'm not super like, cheerleader, optimistic person.

But also I think a lot of concerns are overblown. For example, people look at layoffs at large organizations and like, oh, look, it's already happening. And like, you know, whatever company, large company laid off 10% of their employees and whatever. 00:10:32:16 - 00:11:00:24 Alex And I'm like, look, go back in history, layoffs having happened to companies grow and a lot of times unchecked.

And then they look for excuses to lay off the underperformers. And that excuse has a different flavor every few years. This is just that it's a little bit of adjustment. The companies are taking the opportunity to lay off and right size themselves, and the excuse they use is AI.

00:11:01:00 - 00:11:11:03 Alex Again, I'm not being like a super. I'm not a cheerleader. I'm very pragmatic person. Right.

But I think it's not all disguise falling type scenario. 00:11:11:05 - 00:11:33:28 Michael Yeah, yeah. I mean I think that makes sense to me. If you think about the history of tech in general, you're right.

The regularity of layoffs is not something that, you know, this doesn't this doesn't throw things off. It doesn't make me believe, oh, all of a sudden, you know, we're we're we're seeing this apocalypse come when when in fact, we know that, you know, layoffs is a part of it. 00:11:34:00 - 00:11:55:26 Michael And in fact, in the sales world, you know, it's kind of common knowledge that layoffs are just a regular thing because there are, you know, underperformers that you need to get out of the organization.

And, you know, so so I think that makes a lot of sense. And the other thing is if you look at history, you know, we had the Industrial revolution to good analog to this. 00:11:55:26 - 00:12:11:27 Michael And if you look at it in detail, you'll see that, you know, a lot of a lot of people did lose their jobs, but they either retrained or they just simply fell out of the workforce completely because they just chose not to, to do anything about their fact that their skills were just antiquated at the time.

Right? 00:12:11:28 - 00:12:33:18 Alex Yes. Yeah. Yeah.

That's right. And then I think that parallel to industrial revolution is very I mean, it's a good analogy now. It's not exactly I want to say would it turn out exactly. No.

But you know, they say history repeats itself. What it does, it rhymes but it doesn't repeat it. So it kind of rhymes. 00:12:33:25 - 00:12:36:01 Michael It does.

Right. Yeah, it does definitely does rhyme. 00:12:36:04 - 00:12:36:14 Alex Yeah. 00:12:36:20 - 00:13:04:04 Michael Yeah, yeah.

Well, so at Famous Labs, you know, you described it as a research lab that, you know, can constantly launches new products. And, and so you have AI for marketing content, famous AI for e-commerce, Heisenberg for drug discovery. That's a wide portfolio. I mean, how do you know what ties that all together?

And how do you think about the innovation process inside your own company? 00:13:04:07 - 00:13:14:24 Alex Yeah. So one of the changes, if people are listening for people who are entrepreneurs and listening to this. 00:13:14:26 - 00:13:52:19 Alex I think they are Revolution will enable idea people like idea machines, I call them.

And and it's again, I don't want to talk in extremes, but in the balance between, oh, I have a lot of ideas and let me pursue a lot of them versus do one thing and focus on that one thing. It used to be that it was very like in business specifically, it was very focused on one thing and do one thing oriented. 00:13:52:19 - 00:14:22:15 Alex I think the needle has moved a little bit towards the middle, meaning it actually because execution is a lot easier. People who have a lot of ideas, they will they are able to execute multiple ideas because execution with AI has become easier for some people.

It's good news for some people's bad news for people who are like, you know, constantly full of ideas and and all that. 00:14:22:15 - 00:14:40:07 Alex It is an enabler enabled, enable enablement tool for people who are super focused on one thing and they're afraid of change. I highly encourage them to. 00:14:40:09 - 00:15:11:07 Alex Basically revive that playful inner child again and start playing with new ideas.

Because if you sit still on one thing in a rapidly shifting market, the chances of your business not becoming relevant is just now low. Right? So so it's, you know, software is no longer it used to be that you could build a SaaS company and, you know, and your moat was your software. 00:15:11:12 - 00:15:26:21 Alex Well no more no.

Yeah. So so you better get moving. And like I said, for some people that's good news. For some bad news, my overall philosophy is this.

00:15:26:24 - 00:15:54:15 Alex We remain the ultimate playground, okay. For really smart people that I've assembled in my team, the ultimate playground. Everything is on the table. If I can build a product that helps people in some way or in some way or form, we will attempt it and try to build it.

I have a few things on what kind of ideas to pursue. 00:15:54:16 - 00:16:27:09 Alex I have to see that it's useful to people. I also try to do products that are, I call them human centric, meaning it is it. It adds value without taking value from another human.

So it's like it's essentially like ads. I mean, obviously with AI you can do things that removes humans, but I try to do things that are additive, and if it has those, then it's fair game and we play with it. 00:16:27:09 - 00:16:30:01 Alex And if it makes sense, we continue playing with it. 00:16:30:06 - 00:16:54:19 Michael Yeah, yeah.

I mean, again, that that is something that I've seen my own, you know, use of the tools and in other companies that I've been working with that, that are implementing a genetic AI across a lot of different functions. And, you know, one of the things to that I learned over time is just sitting with like, Cloud Coworker, for example, which is an excellent tool for for the kind of work I do. 00:16:54:21 - 00:17:15:22 Michael You know, people will see something that I put out. They'll go like, what was the prompt that you use to get this?

And often I have to go there. There's not a prompt per se. There's a long conversation that took, you know, two hours to go through to get to the end result, but there's no prompt. Right?

Because it's not it's different than generative. 00:17:15:24 - 00:17:25:26 Michael When you think of I'm going to tell it, you know, tell A and B and they're going to get C out of it. That's just not the way this works. It's a much more iterative interactive process.

00:17:25:27 - 00:17:55:08 Alex Yeah. And I think that's actually one of the what you just talked about that is one of the misconceptions out there. So I think if you think of AI so people might disagree with me, but I'm at this point in my life I don't care. So AI models are very much similar to human brain juicy even structurally.

So imagine you hired a new person and you wanted to give them a task. 00:17:55:10 - 00:18:25:13 Alex You don't come up with one massive instruction or prompt, and then you give it to them and they come back with a great product. Essentially, what happens after a while, you say, oh, this person has been with us for two years now. They can almost read my mind, okay.

What is that? That is essentially a like two years of chat history and feedback that has trained that person's human. 00:18:25:13 - 00:18:43:16 Alex In AI language, the context window is built up, so the person creates the thing that you envision with very few words. If you think about AI that way, then all these things for oh, what's a magic prompt?

It's just I'm like, there's no magic prompt. I just. 00:18:43:18 - 00:19:01:03 Michael Yeah, I mean, and that's hard I think for some people to, to get. But if you, if you use this and you do this for a very, you know, short period of time, you actually to get what you want out of it.

You do have to understand that to me at least, I treat more like a human worker than I do tool. 00:19:01:04 - 00:19:24:27 Michael Right? And I think that was part of my mental shift once I started to think of them as coworkers, then I started to, you know, interact with them in ways that took advantage of that context window that got much bigger memory, which, you know, is a part of that channel history kind of idea. And those things have have changed the way the tools work tremendously.

00:19:24:27 - 00:19:30:25 Michael And certainly that is, you know, the enabler for successful AI, for sure. 00:19:30:26 - 00:19:33:00 Alex Exactly, exactly. 00:19:33:02 - 00:19:59:06 Michael Well, you talked a bit about the ideation to output pipeline, and that's an interesting piece. And that's something that comes up a lot of my conversations with clients about innovation.

Right. How do we how do we go, you know, ideation to output. What should that look like. And and, you know, execution becomes easier with AI.

So bottlenecks have moved, you know, maybe upstream to the ideation part or taste as you called. 00:19:59:08 - 00:20:16:08 Michael Yeah. For a business leader, what are you what's a practical AI powered innovation workflow look like? You know what.

What's what sparks the idea. And then all the way through, you know, producing it or launching it, service or product. 00:20:16:10 - 00:20:45:12 Alex You want to know my workflow, how we come up with ideas? Okay.

It is not for everybody, but I'll tell you guys anyways. So we have about 30 or so people in my company and what I have found, and I've been doing this for over two years now, very different than how I used to do so. But this is how we do it with a group of my team members. 00:20:45:14 - 00:21:16:08 Alex We actually we are actually sort of like a hybrid remote with a small office in Florida kind of setup.

What we do is that you may call it off site, but it is not exactly an off site. We kind of converge in a city and either get multiple hotel rooms or a big Airbnb and, and we just spend like a few days together. 00:21:16:10 - 00:21:46:26 Alex And it's funny how every breakthrough we've had, like in terms of ideas, big things to do pretty much has come from that process. So essentially getting a bunch of really smart people into a literal hotel room and, and spending many hours, maybe TVs running in the background, we talk a little bit about maybe a basketball game is going on or whatever.

00:21:46:28 - 00:22:22:04 Alex And in and we are talking things, it's like, hey, how about this? And then like, I know. And then the other person comes up with another thing. And now that makes no sense, I know.

And then we mock something that we look at it like, no. And then eventually breakthrough happens. So in a world of remote communication and zoom and technology and AI and everything, the seed that gives you that breakthrough idea is a bunch of smart people in a room. 00:22:22:07 - 00:22:29:03 Alex So I think that's really the core.

I literally gave you the secret sauce. 00:22:29:06 - 00:22:48:15 Michael So no, that, that, that is I mean, certainly from an experience, I would say that I've seen that happen many times with, you know, with teams and off sites and just get into kind of neutral ground and having it open to to brainstorm and think through things and talk through problems, that sort of thing. But it makes a huge difference. 00:22:48:16 - 00:23:12:00 Michael Yeah.

I you know, we mentioned Hyson is one of the companies, the ideas that you've been working with and I, you know, not to spend a ton of time digging into how this work. But I'm curious because you've called it, you know, cursor for chemist and, and, you know, quantum informed AI system for small molecule drug discovery. That's a that's a mouthful. 00:23:12:01 - 00:23:25:22 Michael But, you know, that's different from marketing or or operations, right.

Yeah. So what what made you go into biotech and then, you know, how does this actually tell us where, you know, vertical AI is heading perhaps. 00:23:25:24 - 00:23:58:27 Alex Yeah, vertical AI is going to be huge. Just to put in perspective.

And it is huge already. I don't even know. There are many very big AI vertical companies now. So essentially just so people understand AI models like anthropic, OpenAI, they are built, they are general AI, they're trained on a huge data set, and they have a very specific purpose, which is to do a lot of things, even though they might be better in some aspects and others, but in general they're generalized.

00:23:59:00 - 00:24:16:27 Alex But then you have things like, okay, so how would a chemist use AI to do drug discovery? That is not a ChatGPT prompt. 00:24:17:00 - 00:24:19:07 Michael If it is, it's one I've never seen. 00:24:19:09 - 00:24:48:18 Alex It is.

It is essentially a combination of yes, LMS, but also foundation models that are trained on chemistry. They're not even text based. They are like molecular models. And then a specific what's called a harness that is built on top that can keep track of okay, so this is what we need to do.

This is my target and these are the candidates. 00:24:48:18 - 00:25:16:00 Alex And now let's feed this candidate with reasoning into this model. Try to create more candidates then. Now let's see what happens with that drug candidate.

And then give then run a bunch of other libraries to see. Okay, so if this drug candidate is toxic does it what kind of properties does it have and things like that. So a chemists looking at it, they understand non chemists will not understand. 00:25:16:00 - 00:25:49:04 Alex But chemists looking at it understand what's going on and they can modify it.

And I use the example of cursor for chemists. But I believe the cursor for people who don't know like cloud code cursor, they are coding tools for software engineers. So it's essentially it's a harness that people. It's like a coding environment.

Software engineers can go in, it understands like code repositories and things like that, and allows a software engineer to do their job faster using AI. 00:25:49:07 - 00:26:20:15 Alex Well. So that has already like permeated throughout the software engineering industry. But then I look at it and I'm like, okay, the world is not just software engineers.

There are all sorts of people. There are mechanical engineers, civil engineers, electrical engineers, you know, so aerospace engineers, chemists, physicists. So what is the equivalent of that workflow for them? And it also makes sense why cursor popped in the market first.

00:26:20:15 - 00:26:31:03 Alex Because who are building AI models software engineers. So what's the first tool. They built a software engineering tool because they understand it doesn't make sense. 00:26:31:03 - 00:26:32:14 Michael Right.

Wouldn't you do that if you were. 00:26:32:16 - 00:26:53:13 Alex Yeah. Exactly. Exactly.

Yeah. So so it's like they understand that. But how about the chemist. How about the physicists.

How about the aerospace engineer. So that is I think is the next evolution. You call it vertical AI I totally agree. And I'm going to try to make as as many of them as I can myself to help those people essentially.

00:26:53:13 - 00:27:02:16 Alex So so essentially a chemist should not be going to charge you saying hey or that that just is not built for that, so to speak. Yeah, yeah, yeah. 00:27:02:16 - 00:27:23:04 Michael It's interesting because that was kind of a revelation for me last year as I was using the models more and playing around with different tools and talking to different people that, you know, there are a lot of professions that are very specialized. There are a lot of companies in industries that require a lot of specialty.

Right? The different the approaches are different. 00:27:23:06 - 00:27:35:03 Michael Workflows are different. So I mean, it seems to me that the output from what you're saying really is specialized models, no matter what the size.

Right? It might be a small model, it might. 00:27:35:03 - 00:27:37:21 Alex Be smaller. There are smaller normally.

Yeah. 00:27:37:22 - 00:27:51:10 Michael Yeah. Then then that is you know, that's the future of vertical eyes, the capability to build out these specialty models that can really cater to the needs of a specific population, whether industry or a person. Right.

00:27:51:12 - 00:28:19:07 Alex Yeah. Exactly. Like a, like a, like a model that you give it, for example, the probabilistic material engineers like or chemical engineers, you give it a, a compound, some sort of material, and it gives you other similar things like that. Does that make sense?

How about this one instead of that other one. So so that is a very specific product that is not general. 00:28:19:09 - 00:28:44:18 Alex But then it shouldn't because it's like even in your own brain you have different parts of your brain. They have completely different specialties.

You have like part of your brain that's responsible for hearing vision, you know, motion controls. So it's like there's all sorts of different specialties in our brain. And it's not some monolithic one AI model that rules them all type thing. 00:28:44:25 - 00:28:58:14 Alex So there's a there are huge opportunities there that is just starting to be to be explored.

And I'm actually I think it's very exciting for everyone out there. Yeah, yeah. 00:28:58:15 - 00:29:17:15 Michael And I think it does provide a lot of opportunity and particularly the idea of sizing the model, make it specialized, but sizing it to the use case and, you know, then you get much more, you know, economies of use there and, and that sort of thing as well. 00:29:17:18 - 00:29:43:04 Alex Yeah.

And also reinforcement learning of that is a lot easier because you know, you get you don't have to train on the world. You train on. If it's chemistry, you just make sure the output on like molecules that come out makes sense. Does that make sense?

It is a much smarter in a way. I mean, it's like positive negatives. 00:29:43:07 - 00:29:56:13 Alex It's a much smaller problem set. But the depth you have to go to make it good is much deeper.

Does that make sense? So it's like it needs to be really good at what it does. 00:29:56:14 - 00:30:09:21 Michael Yeah I mean broad models solve general generic or broader, you know, types of questions but very specialized models. Center for specialized Industry and professional questions, whatever it might be.

Right. 00:30:09:22 - 00:30:11:13 Alex Yeah, exactly. Yeah. 00:30:11:15 - 00:30:36:00 Michael So we've had what some people have called the Sasaki lips, which I kind of like better than SaaS apocalypse.

But but they both work, I guess, you know, and that's over the last six months or so. Certainly the markets have reacted to, to this narrative that, you know, SaaS is being replaced. It's, you know, companies aren't going to buy it anymore. 00:30:36:01 - 00:31:00:20 Michael They're going to, I don't know, go in a closet over the weekend and use cloud code to build a new ARP system or something, and you push back on that really hard.

And so, you know, you you pointed out, you know, nobody's going to go there a role, which I. Yes, yes, that makes sense. And when you, you know, where do you see the real opportunities for disruption versus like where the existing enterprise software still makes sense. 00:31:00:24 - 00:31:01:12 Michael Yeah.

00:31:01:13 - 00:31:28:10 Alex So I think that is another thing. It's just it wouldn't change. Like this happens. People tend to go to extremes.

Like it's like I'm like it's neither this or that. It's in the middle. So the truth is there is so SaaS solutions inside companies. A lot of them have a moat.

That is not just because, oh, somebody coded it. 00:31:28:12 - 00:31:58:19 Alex It is testing compliance, integration with existing systems. Like you don't want to take a chance on your payroll system. You don't want to take a chance on your inventory management system.

You don't want to take a chance on like, oh, let me just go and wipe code and electronic medical record and put my patients in there so I don't pay for EMR. 00:31:58:21 - 00:32:21:24 Alex First of all, you will not be compliant to begin with. So it takes a. Lot to to to making sure the software is compliant and it works and it's tested.

Also because these software companies have a lot of clients, they are constantly they have engineers and they deal with bugs across a lot of clients to make sure their software works. 00:32:21:24 - 00:32:43:26 Alex Now, obviously there's always there are flaws in that too. But it's you can you can pretty much rest assured that your payroll system is not going to make a mistake and deduct the right amount of taxes, and your EMR is not going to put, you know, the x ray for one patient into another patient's file and stuff like that, right?

00:32:43:27 - 00:33:17:01 Alex So so so so those things I don't see even I call it reason for existence. I don't even see a reason why a hospital system will check, try to wipe code their EMR. I just don't and and or or a large retail store with multiple like warehouses, shipping things with trucks all over the place or locations. All of a sudden they're saying, you know what, we don't want ERP we're going to use.

00:33:17:02 - 00:33:27:10 Alex We go, we are going to unleash our engineers. They come up with a whole routing and trucking and inventory management system. It just it that company will collapse. 00:33:27:13 - 00:33:27:28 Michael Yes.

00:33:27:29 - 00:33:52:18 Alex So so so so it's like it's like I'm like okay. Now are there other companies, SaaS companies that are like completely cosmetic? Yes, there are some. Okay.

So and also the other thing now the good what's the upside of this is actually for SaaS companies. And I have no dog in this game because I'm an AI company. But there are SaaS companies. 00:33:52:25 - 00:34:22:02 Alex The upside for the SaaS companies is now they can implement AI to make their product better, to make it integrate with other solutions.

With AI it must. So data consolidation becomes a lot easier. So there's actually upside for a lot of SaaS companies. And there are some companies that are going to be destroyed because they added no value, had no reason to exist to begin with anyway.

00:34:22:03 - 00:34:25:04 Alex So the truth. Like I said, in my opinion, somewhere in the middle. 00:34:25:08 - 00:34:44:02 Michael Yeah. I mean, I think that that to me, as I've thought about this and looked at it more, you know, I can see the pricing problem because that kind of jumps out at you, right?

Like maybe per seat pricing doesn't really work in the world. Okay. But but when you think about I'm going to go, you know vibe code in Europe system. 00:34:44:03 - 00:35:03:07 Michael No thank you.

Like no thanks. Because the problem is the mode is not your ability to to to to to approach that or even understand kind of what you're trying to do with it. The problem lies underneath it. Right now the data is one problem.

And and you've got to certainly get that in a, in a usable format. 00:35:03:07 - 00:35:30:03 Michael But but you can you know, there are ways around that, right? I mean, you can federate the data, you can move it all into a data lake or whatever. But but when you move above that one level, the business logic, you're not you can't recreate business logic automatically, magically.

Right? It doesn't it doesn't do that. If there's a moat there, that to me that jumps out as a real a real technical challenge that I'm not willing to to try. 00:35:30:04 - 00:35:49:27 Michael I mean, there are certainly simple little programs and projects that I would take on that don't require me to rebuild the, you know, you rebuild the the Empire State Building with matchsticks.

I mean, that's, you know, no, thanks. But but but at the same time, a lot of those things are, for me, very protected. Yeah. For now anyway.

00:35:49:28 - 00:36:22:20 Alex Yeah, that makes sense. And I think you are thinking about it the right way, in the sense that if you're in SaaS, what you should think about business, SaaS, what you should think about is, okay, hopefully, hopefully your SaaS. I mean, there are SaaS companies, like I said, that's going to go away. But if you're one of those core ones, the question you need to be asking yourself is that how can I use AI to make my tool ten times better?

00:36:22:21 - 00:36:44:18 Alex That's really the one question you need to be asking. So if you're if you're in payroll and employee review systems and stuff like that, you're like, how can I? And one way to think about it is like, how can I use AI to make my tool ten times less painful to use for the HR manager in the company? 00:36:44:19 - 00:36:48:21 Alex So that's the question you answer that I think overall you're going to be fine.

Yeah. 00:36:48:24 - 00:37:10:16 Michael Yeah, yeah. And I mean, I see this in a lot of applications where the UI starts to shift towards, you know, an agenda way of approaching it. Right, conversational or, or, you know, other, other methods of interacting and having the agent execute things so I can see change.

But at the same time, yeah, it didn't seem to me like we're we're quite at the apocalypse yet. 00:37:10:19 - 00:37:32:09 Michael Yeah. You know, I mentioned thinking of, of of AI as a digital workforce and, and so take that a little bit further, you know, if I'm, if I'm a company, I have human workers. I now could have digital workers, I put them together.

I have a hybrid workforce. To me, that seems like the, the real concept of where we are right now. 00:37:32:12 - 00:38:02:00 Michael You know, I'm not saying that someday, you know, it might replace everybody or whatever. But for as far as I can see right now, that hybrid workforce makes sense.

When you talk about how do you move businesses forward right now? Yeah. So I mean, given the way you famous labs is, you know, focusing on all these different tools and a lot of it's, you know, specialized, that sort of thing, where do you see the line between, you know, what AI should own and what human should be in the lead for? 00:38:02:02 - 00:38:08:07 Alex Yeah.

Okay. I. 00:38:08:09 - 00:38:09:13 Alex Again it's. 00:38:09:15 - 00:38:10:13 Michael Easy question right.

00:38:10:18 - 00:38:44:12 Alex Yeah. He's a question. Is it question hard. It's a question hard to answer.

So I think essentially if anyone in your company is not effectively leveraging AI. And I say that for knowledge workers. So maybe physical like it's different. But even in those scenarios you need to give them the AI tools.

But knowledge worker not leveraging AI there is a problem. 00:38:44:14 - 00:39:35:20 Alex Okay. So it's like there is a problem. Now you may it may be a small problem that is irrelevant.

Or or for example, if you're a receptionist is not using some sort of a workflow. Let's I'm going to take the like literally like a receptionist at a dental office. Okay. So they need to be enabled using some sort of AI tool to do a better job of following up with patients and giving them like more customized experience versus if it is like before, what is just take a call and then put on a calendar and then just that needs to be replaced to oh, now my receptionist has this tool that it automatically reminds people 00:39:35:20 - 00:39:59:00 Alex to do something if they did is a text message.

If the customer can message the receptionist, AI like replies back and says, hey, I can do this, we can do that. If they have a question, I can answer that. And then the human supervisors, meaning the receptionist becomes a supervisor of that AI agent. If you're not doing that, you're under utilizing your person.

00:39:59:02 - 00:40:36:01 Alex Does that make sense? So now the question is, okay, so does it really matter to your business? I agree, sometimes it doesn't matter. It's like it's fine.

It's like not that big of a line item for my business. I don't want to go through the hassle of doing it. The person has been working here for 20 years, I understand it, that's like practical, but if you do that across your company and and like just ignore enabling people with AI altogether across the company and you are in a competitive business, you will eventually go out of business. 00:40:36:03 - 00:41:05:07 Alex Does that make sense?

So you need to sort of start thinking about how can I enable by people with AI at all times and what other superpower are can I can unlock for them? And you should certainly do that with all the key roles of your company. Now some some businesses are like, okay, so let's say a window manufacturing company in Tennessee has a lock on the market. 00:41:05:07 - 00:41:20:00 Alex And maybe if they don't do anything, they just keep doing the same thing.

Maybe it doesn't make a difference. But but but if you're in a competitive business you have to do that or other people will eat your lunch essentially. Yeah. 00:41:20:02 - 00:41:45:01 Michael Well, so so a lot of the audience are business leaders.

And, you know, part of what we've really tried to do with the show is, you know, distill these concepts down so that it is appropriate and understandable from a business perspective. Right. How do I use this today to to change something? And so I'm curious if you were going to give some advice, particularly around innovation, how, you know, how should executives think about that?

00:41:45:02 - 00:41:55:02 Michael What should they do to plug AI into that innovation process? And you know what? What's the first experiment they should do maybe or how should they think about this? 00:41:55:04 - 00:42:36:10 Alex Yeah, I think I'll answer that in two ways strategically and tactically.

So strategically just know especially knowledge work type companies physical is it a little bit different. But knowledge work type companies strategically what is happening is you're offering a specific product or service. That same customer has needs for other products and services. You can build that other product and service yourself and horizontally expand, meaning sell more things to the same customer.

00:42:36:12 - 00:43:11:09 Alex Okay, if you refuse to do that, just no. It is easier for your competitor to also build your product and service and come to your market. So there's a little bit of a game theory dynamics going on where companies have to expand their product portfolio and offering, because if they don't, other people will come into their territory. To give you a little bit of a like a historical funny tidbit and hopefully I don't I say this one right.

00:43:11:10 - 00:43:42:22 Alex So New Zealand many years ago used to be multiple little kingdoms and and tribes. And the reason that no kingdom or tribe could take over the whole island was because in order to march from one side of the island to the other side, you had to pack a lot of calories. And they didn't. They had like sweet potatoes, if I'm not mistaken, and didn't didn't pack enough calories, so they couldn't.

00:43:42:22 - 00:44:04:25 Alex The truth was, it was like fragmented because you couldn't really get a bunch of men march all the way and the food would be too heavy or not enough of it the moment. Potatoes, if I'm not mistaken. Hopefully I get this right was introduced and potatoes are more calorie dense then you could get. A group of people march across the island and have food rations for them.

00:44:04:26 - 00:44:29:18 Alex Okay, it became one. So there's actually all the warring tribes. It became one. So the reason I say that is because exactly that same thing is happening with AI.

So people can easily go into each other's market now with AI because execution is easy. So it was fragmented and you felt safe with your customer base. Don't feel safe. 00:44:29:22 - 00:44:48:10 Alex It's like other people can go into your territory.

You can't go into their territory. It's the same customer. You expand your portfolio. They you may replace someone else if they expand their portfolio.

It doesn't make sense. I'm not saying it's a zero sum game, but there's a little bit of that dynamic is going on. And strategically just so people know. 00:44:48:12 - 00:44:52:28 Alex Yeah, yeah.

Was that clear by the way. It's like like yeah. 00:44:53:00 - 00:44:59:10 Michael No, no. That makes sense.

I mean if you if it kind of reminds me of that old saying that armies march on their stomachs, right? I mean, that. 00:44:59:10 - 00:45:00:02 Alex Is exactly. 00:45:00:03 - 00:45:04:13 Michael That is in fact the reality of it.

If you can't provision them, they're not going to get very far. 00:45:04:19 - 00:45:35:19 Alex Yeah. And with AI, essentially what has happened is that you can go further, you can take your army further. So that means you have to.

Now from a tactical perspective, what I do is, is sort of simple. I, I sort of ask, okay, so I have a person that is doing this job in a company, how can I make them accomplish more or provide like produce more or provide more value to the customer or reduce the pain of their day to day work? 00:45:35:20 - 00:45:45:25 Alex Okay, with with what AI tools are possible and you have to tactically, systematically ask that question across your company at all times. 00:45:45:27 - 00:46:12:25 Michael Yeah, that that makes a lot of sense to me.

And I think, you know, if I think about how companies are using today that, you know, that ties right in. I, you know, really interesting conversation and I but before I let you go, though, I like to do this at the end of every show. I like to, you know, try to get a recommendation of somebody, give me a thought leader, an author, podcaster, you know, whatever you think would be relevant for the audience. 00:46:12:26 - 00:46:18:02 Michael Well, somebody that you think they would enjoy and they should check out.

00:46:18:04 - 00:46:44:20 Alex I have a lot. But if I had to have have to pick one. Nassim Taleb is probably the person to follow his book. Antifragile is still one of my favorite books, and they cause I like him because the concepts.

He's really good at taking. You're putting a new thought in your head. So it's like. And give it a word like black swan, antifragile.

00:46:44:20 - 00:46:49:13 Alex It's like, give it a word and a new concept that then you can carry with yourself. That's why I recommend it. 00:46:49:14 - 00:47:07:18 Michael Yeah. Yeah.

No that's great. No thank you. That that definitely is interesting. And folks should check that out for sure.

Well Alex, thanks so much for joining. I really appreciate it. And you know hopefully this will generate some interesting conversations. But yeah really really appreciate you joining today.

00:47:07:19 - 00:47:11:16 Alex Thank you for having me. 00:47:11:19 - 00:47:35:21 Michael And that's the show for this week. Thank you all for joining us. Remember to like, share and subscribe to the show.

If you enjoy the show, please leave us a review to help others find us. For more research on AI and other software, check out arionresearch.com. And if you're an expert in AI, generative AI, or business automation, either as a provider or an end user, email your information to disambiguation at arionresearch.

com. Don't forget to join us next week! Disambiguation is an area in Arion Research production. I'm Michael Fauscette and this is the disambiguation podcast.