Subscribe
Share
Share
Embed
Explore your curiosity - Interesting people with fascinating stories.
Life consists of three things:
-The stories we tell others
-The stories others tell us
-The stories we tell ourselves.
Matt, welcome back to the show.
Speaker 2:Heck yeah, man. And thanks for having me back, dude. These are great times. I really appreciate it.
Speaker 1:Let's start by talking a little bit about how we used to, or how many are still developing code today. So just give a broad high level picture for our audience.
Speaker 2:So as it pertains to software development, and when we get into this, you're going to see the analog, the analogous nature of what's happening in software development and how it's going to affect our business process hierarchy and the way we literally function as a society. So typical software development is exactly what it is. It's called the software development life cycle, and it consists of a series of sequential steps. Normally we start out with something like requirements gathering. Hey, what type of tool do you need built?
Speaker 2:Because that's really what software is. It is a tool made of ones and zeros to help people do a thing. Okay. That's software. All right.
Speaker 2:So requirements, designing the solution, developing the solution where your actual ones and zeros are assembled by highly skilled people. And then now they're computer assistants. You go into a testing phase, deployment, and then overall maintenance where the operations and maintenance O and M phase, where that takes you out all the way to the system decommissioning. Right? And then you have feature updates and improvements to the tool, right?
Speaker 2:That, that occur throughout a life cycle. So that in general is your software development life cycle as it traditionally is taught and traditionally executed.
Speaker 1:And just to add some color to that, we have different teams doing different things here, right? So you have your, your product managers who are writing in essence a document or working with designers to create a mock up, which is like a picture of what the software should look like. And then you have the engineers looking at this document created by product and UI, like okay let's build this and then you have a whole different team quality assurance that are like well what the product manager wanted is that actually what the developers built and then you have yet another third team who are responsible deploying this to the cloud or whatever which is our dev operations team. Correct. So you've got all these teams working together and passing information between one another.
Speaker 1:And it's an ineffective process, obviously, have people talking to each other, always understanding each other, reading long documents. All this is about to change.
Speaker 2:Absolutely. Yes. So with the process that you're really talking about, it's like, yes, you have different teams and in order to project manage or program manage that level of software development, I'd liken it to, I don't know, something like conducting an orchestra and through Agentic AI, we get what's called orchestration, right? Where we have now automated those human, those naturally, those, previously human tasks of orchestrating the different functions, right? So just like you had to, you had elaborated, right?
Speaker 2:When you have that design where you're, you know, you're, you're drawing out the solution, right? Well, that leads right into your development activities, which are then, need to be bounced up against your QAQC, your quality assurance and control. Also verification with any type of information assurance or cybersecurity concerns that cause those teams are always involved as well. But through AgenTik AI, what we've done is we've collapsed the stack. What that means is that we have, by removing that human element in that orchestration, we have sped up those communication flows.
Speaker 2:We have reduced misunderstandings, right. And reduced, cogitation time. Like people don't have to read those documents anymore. The computer's doing it. Right.
Speaker 2:People don't have to determine, you know, do the resource allocation for people's time, right? Or compute time or, or storage or any of those other considerations that you burn through as you're developing software. Now the computer does that. So by removing that human element that will take you minutes to hours and sometimes even days or weeks to resolve these issues is now done milliseconds to minutes. Right.
Speaker 2:And so that's why we, you know, have the now that the more streamlined and continuous process.
Speaker 1:So this is, this is a wonderful high level explanation, but what we're doing is we're completely changing the structure of our teams. We used to know what it looked like, We had engineering, quality, DevOps. What are the teams doing now? Like somebody needs to verify that this is all working and going in the direction. How do you structure teams in this new world?
Speaker 2:Sure. So really I found success and cause again, and we're all, we as in humanity, We're all trying figure Yeah. We're all trying to figure it out. So we are in the messy middle. And quite frankly, I kind of love it because that's where the magic happens.
Speaker 2:Right?
Speaker 1:So
Speaker 2:it's really about by transforming the processes, which is what AgenTic AI does, it also transforms the roles. Okay. And if it's transforming the roles, that means it's transforming the skill sets that an individual needs to be successful in those roles. Okay. So let's take developers for example.
Speaker 2:Right? So traditionally, developers develop. Okay. And what that means in layman's terms is they write code. Okay.
Speaker 2:On that, when you think of the matrix and you think all those ones and zeros, that's a software developer in very, very simple form. Okay. But now instead of doing that, because again, that's all their, you know, that's their human brain bandwidth right there. They're now acting more as system architects, really taking time to plural the design requirements, taking time to validate against QA, right? The quality assurance and, and, and being, and essentially elevating themselves in a role from a software developer that's worried about this single feature or this single bug fix that they've got to go through.
Speaker 2:Now you're worried about the entire end to end software stack, Right. And you're overseeing that, which means, so what does that mean for developers? If I may just really quick, that means instead of coming into the workforce and being concerned with just the technical acumen of code validation, writing the correct code, etcetera. What you're now concerned with are, wait a minute, what are my people skills? Because now the time and the opportunity for communication has been condensed.
Speaker 2:Right? So now we, that communication becomes even more important. What about managerial skills, right? For my allotment of time, my personal boundaries, right? For what tasks are important?
Speaker 2:And all of these, these managerial supervisor skills become much more important to your junior developers as we advance.
Speaker 1:Let's put a pin in the impact on the human relationships. I want to come back to that, but you said something incredibly important, right? These skills of these engineers that used to need to know how to code just one type of thing, front end, like HTML, whatever, or back end, right. That is changing where they need to know the whole architecture. So if I had to give one recommendation to current and future, you know, software developers who identify themselves as back end or front end is like that's not going to fly in the new world.
Speaker 1:You have to be an architect. So go and learn AWS, go and learn front end and back end. You don't need to know it at the coding level because you know what, you're not going to be writing the code yourself but you need to understand the architectures, the design patterns, you need to understand how they work together otherwise you're going to find yourself obsoleted pretty quick and obviously the simple example of what you need to not need to understand how to use AgenTic but you're not going to be successful unless you understand architecture. So these are the things that I think software engineers of the future need to study.
Speaker 2:Go ahead, Matt. So traditionally, like, think about that. So now we have an AgenTic, so you're a software developer, right? And we're an agentically orchestrated SDLC. Okay.
Speaker 2:So what happens now? That means that errors and bug fixes, well, they occur nearly automatically and they occur asynchrony, async, or I can talk. It's been a long day. Basically with the actual feature coding and encoding into this final product. So as an arc, so as a developer, well, you're just concentrating on that one bug fix.
Speaker 2:But now as a developer with what Ari said, taking that architect visible approach, Hey, wait a minute. Is this even above? Okay. Let's look at that. Let's, let's talk about that first.
Speaker 2:Is it fixed and does it fit within the overall architecture or purpose of the toolsoftware that you're developing?
Speaker 1:Brilliant. So people are asking themselves at this stage, are humans even going be in the loop? And maybe the answer at some stage is going to be no, but right now, absolutely. Because there's hallucinations, it doesn't really work. It builds something that you don't expect it to build.
Speaker 1:So you need you know, adjust it. So, so that's, that's one aspect of it. If you're building a team from scratch, right? And you're building it within a agentic architecture, who are you hiring?
Speaker 2:Who am I hiring?
Speaker 1:What's the skillset? Is it just the old developers or suddenly because of this new world, you can hire new people, right? That maybe couldn't do this job in the past.
Speaker 2:Well, correct. So, so let's talk about skill sets that I'm looking for when I hire, you know, again, these new developers. I'm looking for process. I am looking for process design. Because again, what we're talking about is that architect role where you're overseeing the process.
Speaker 2:You're not in the process. So I found a traditional process design, such as Six Sigma, such as TQM, such as the Kaizen philosophy go really, really well within the new AgenTic driven software development teams. Simply because again, you have so many people at that architecture level where you have to pin all of these processes together, right? And you have to be able to vision that. Again
Speaker 1:Let me, let me stop you right there. All those verbiages that you just used, they're all manufacturing words. And what those manufacturing words mean, it's different ways to stop manufacturing in the middle of its process if something goes wrong. It's ways to basically orchestrate the manufacturing process. But I think the most interesting thing here that I want to kind of highlight for the audience, you're using manufacturing processes in software because you're basically arguing, I'm asking more than saying, is that you're turning software development into a manufacturing process.
Speaker 1:That's what you're saying.
Speaker 2:Well, that's exactly what's happening, Ari. So what's happening with knowledge workers through agentic AI software development is exactly what happened to manual labor during the industrial revolution. Exactly. Right. Because artificial intelligence isn't a, isn't practical effect on our everyday lives, nothing but democratized intelligence.
Speaker 2:Okay. So, so guess what? Like you now, like what we had is the time saving and effort saving new tools that we physically use. Now we have the intellectual capacity, time, and effort saving knowledge tools of the ones and zeros, which falls under information signs. So as we move forward, understanding the principles of manufacturing is directly analogous to understanding the principles of software development.
Speaker 2:Absolutely. And software development has always been a manufacturing process anyway. People just don't under ever really realize it. So instead of building a car, right, full of nuts and bolts, you're building a video game full of ones and zeros, but they are the same fundamental principles behind each.
Speaker 1:Agiles, you know, agile and all this stuff we think we now associated with software. That's not where it
Speaker 2:was That was Toyota. Right? Like, I mean, we're talking about, was Toyota Six Sigma or TQM? It was one of the two.
Speaker 1:TQM. It's Toyota quality.
Speaker 2:Sorry. Yeah. I So apologize, But it was TQM, and now we still have that. I believe Motorola might have been Kaizen, right?
Speaker 1:Yeah. I can't remember.
Speaker 2:Or maybe, again, business history. Sorry, but
Speaker 1:Kaizen was the Japanese, you know, pull the lever and then everybody stops. Then you think about what's wrong with the manufacturing. That was, that was the
Speaker 2:Philosophy though. It's a manufacturing philosophy that, that again, allocates your intellectual capital and bandwidth to address the problem in production, which is the exact same thing that's now necessary with ejecting software development. So the lessons that all these manufacturing companies learned in the seventies and eighties, dust off the books folks and just, you know, change the verbiage around. It's the same thing.
Speaker 1:So, so like people who are experts at, you know, software architecture and also manufacturing, I mean, except for you and me, they're not very common, so we're gonna have to hire people that come from different backgrounds So have to work what is that gonna look like? Who are the kind of people that just come out of university? It's no longer only software engineers. It can be all kinds of stuff. So who are we hiring?
Speaker 1:How do we build these teams? Going back to the original question.
Speaker 2:If the success of our operations is contingent upon the intellectual capability of the employee that we hire, And now we're moving from, I'm forgetting the correct term. Someone in the comment section correct me here. We're moving from the static intelligence of rote memorization to now the more fluid intelligence of people skills and interaction and orchestration and, and, systems thinking with stocks and flows, then that's exactly what we have to quantify and measure if we're looking to bring on the right people.
Speaker 1:Brilliant. So hold on, let me put an, this is brilliant. I just want to put a hit this on the head. What you're saying, and I'm adding some words to your mouth, keep me honest here, is that things are so fluid. We don't know what the new technology that's not going to come out next year, it's going to come out next month.
Speaker 1:So what we really want is the natural capability for people to A) be fluid with that, learn new things, think out of the box, basically whatever tool and whatever innovation is going to come, they can fit that into their manufacturing process and adapt to it. So it's no longer about studying computer science and knowing things and memorizing. It's about your ability to be adaptable and to really be predominantly a problem solver in a very technological world.
Speaker 2:All right. All right. I will agree with every word out of your mouth if I may insert this one capstone. You're looking to hire learners.
Speaker 1:Right. Why is that? Why is the learning aspect of it so important?
Speaker 2:Because the dynamics are changing. Like, just like you said, it's not, it's not the tool or the solution that we're looking to begin incorporating next year. Is the tool or solution that no one's thought of. That's going to be here in three months. Right?
Speaker 2:Like, like that's how fast this is moving. So you need to have a continuously learning mindset, which if I can go ahead and use a sports analogy here, Popovich ball. Greg Popovich, head coach San Antonio Spurs, multiple perennial title contender, multiple championships in the NBA. And why? Fundamentals.
Speaker 2:Fundamentals. Fundamentals. Fundamentals. Once you have a learning mindset and once you have an engaged team and you have solid core fundamentals, the rest, the story writes itself.
Speaker 1:Yeah. That, I mean, that's so important because if you take a small example of that understanding fundamentals of how neural networks actually work and the math behind that, that's going to explain to you why they behave in certain ways. That's going to enable you to really orchestrate them in a smart way, so there's different orchestration decisions you can make. Do you do A or do B? Understanding why neural networks are what they are and in essence an agent is just a generative neural network.
Speaker 1:Understanding that is going to give you the power to make these smart decisions. I completely, completely agree. Software developers, I would argue, are a little bit ahead of the curve because they have this architectural knowledge. So they're either going to suffer because they don't adopt fast or they're going to do great because they have this cutting, you know, they have this advantage. But let's look at the rest of the people, right?
Speaker 1:If you're a graphics designer, if you're a journalist, if you're a, you know, a product manager, if you're a sales executive and you have that magic of, you know, learning and technical skill, huge opportunity is opening up for you.
Speaker 2:Right? Absolutely.
Speaker 1:That did not exist in the past because you had to go and study and do computer science for years. What should these people do? What should be the path forward for them, if any?
Speaker 2:So, so when you're talking about, like, like you said, we're, we're looking at these archetype and this new pathway is opening up. Some, and again, I've recently heard it referred to as solopreneurship. I don't agree to that. Right. Because not all the things that, that, that humans engage in right.
Speaker 2:Are entrepreneurial, right. When generating, I look at it as, social augmentation, right. Cause I take myself as that example, like what happens when you take a highly competent generalist who has some serious, you know, like I have some subject matters where I possess a deep understanding of the subject. I'll admit that. Right.
Speaker 2:But then you equip me with the ability to, I don't need to check with QA right off the bat. Like I can reduce that. I don't need to check with these people. I can pull it right here. The information is at my fingertips.
Speaker 2:And as long as I understand the systems, as long as I understand, again, here it is, the architecture, and I'm able to link those processes and procedures together. It allows you to get so far ahead of the curve, it's insane.
Speaker 1:And really what you're talking about is the ability to tool build, right? I'm in sales. I'm going and doing, you know, old school prediction of sales in Excel sheets, but now I can pull the data into Claude. I can pipe the data into different systems. I can create pipelines.
Speaker 1:If I learn how to do this, I'm going to be able to do my sales job of, you know, forecasting 10 times better. If I'm you know a marketing person and I'm all about creating leads, if I understand how all the systems work together and I'm given the empowerment to actually access this data and systems in smart ways which is really going to depend on the CTO CIO to make that happen. Suddenly every single employee in your company becomes a tool builder and they've now X themselves. They're not dependent on IT to come and build those tools through them. That's a huge paradigm shift right there.
Speaker 2:It is. But but and again, it's it's a paradigm shift individually, and it's a paradigm shift for the team. Because just like you said, not only do you have to have the individuals that are willing to go out and experiment. Right? And you're absolutely right.
Speaker 2:You want to give these people the knowledge and the tools, but you do not want to micromanage them or give them the individual guidance. Right? And the reason is, and this is something I tell clients and people all the time, you are the subject matter expert on yourself. No one knows how to do your job better than you do. Cause you do it every day.
Speaker 2:Right? So, Hey, managers. Hey, supervisors. Train them up, empower them. I mean, something that's been, we've been, humanity has been harping on for the past X number of decades and get out of their way and let them solve their problems.
Speaker 2:Right. But again, there has to be something in it for them because if I may take a quick segue, as we go down this road to 10 X ing and a 100 X ing that productivity, right? Well, those efficiency gains, as it stands right now in the market are going back to the company. That's why you see all these AI driven layoffs, right? Where they need to be going is to the individual so we can have these three, two day work weeks.
Speaker 2:Right? Like it completely changes the culture because, you know, like, I don't know about you Ari, but I wasn't put on this earth to stare at this little rectangle right here for fourteen hours a day. You
Speaker 1:know that prices are sticky. Exactly. We know that corporations are selfish. So corporations being influenced by shifting supply and demand, it's going to take time. Going to take time for corporations figure out that there's, to oversimplify, there's two types of employees.
Speaker 1:There's the type that leverages AI is in 10 times X. And then there's the employees that honestly you're not going to hire or they're going to have a much lower salary than these 10 X employees. So, and it's going to take time for the demand for those employees to be there. Their price goes up significantly and those extra revenues and efficiencies goes to that point. That's going to take time because of that fundamental element of prices are sticky, meaning that prices don't change quickly.
Speaker 1:They kind of stay where they are until enough market forces happen. So just, you know, that's a little optimistic on the, on the, what we call the short run. Long run, completely agree with you.
Speaker 2:Oh yeah, absolutely. It's, it's going to be, and it'll be a messy transition. It'll take decades just like every other. And again, I'm not a doomer. I'm not a boo hoo.
Speaker 2:Like, look at our human history. Every technological transition has been messy. Had to come together, write new laws that figure out how it works. And we've had to incorporate it in our society. Right?
Speaker 1:We were protesting the railroads, right? Like that was, it's trains guys, it's trains and yet we didn't want trains, okay? We didn't want railroads. We had huge fights and we still do once in a while about the internet being free, right? Net neutrality.
Speaker 1:Like it's always messy, But, know, twenty years later, we look back and we're like, oh, that was simple.
Speaker 2:How did
Speaker 1:it do? It wasn't.
Speaker 2:What what were we do what were we thinking? We were so ignorant. Yes. It's the same conversation, brother. We are we are functionally and biologically no different, man than the individuals that walked the savannah fifteen thousand years ago into, you know, into the Middle East, right?
Speaker 2:Like we're the same.
Speaker 1:I heard a really interesting quote. We always say history repeats itself. And somebody told me, no, that's not true. History doesn't repeat itself. Human beings repeat their behavior.
Speaker 1:It's the people, not the history, so that's probably a good place to summarize. Matt, this has been such a joy. Feel like, you know, every couple months when we come together, something has happened that we just have to talk about, so you're probably gonna be the first guest ever to come back for the fourth time.
Speaker 2:I'm rocking it, dude. Let's go. We're gonna
Speaker 1:do it. Amazing. Matthew, thank you so much for your time. It's been an absolute pleasure.
Speaker 2:As always, I appreciate you, Ari. Thanks, man.