“My profession is gone”: how AI agents are changing software engineering | Innovantage Podcast #60

Speaker A: Foreign. Hello there and welcome to another episode of the Innovantage podcast where business meets tech to bring you your competitive edge. As always, I'm your host Max and it is my ongoing job and duty to explore that edge between business and tech and to find something interesting for you to listen to and for me to talk about. Today's topic is going to be on the very intersection of business and tech, coming from somebody who lives on that very intersection between business and tech. Somebody whose career has taken him to the highest of heights of uh, the most successful corporate careers as well as entrepreneurship on their own accord. I'm talking of course about my guest Akos Marwe, who is a fractional CTO and cdo, innovation expert, startup mentor and deep tech strategist from Zurich. Akos, welcome to the show.

Speaker B: Well, thank you for having me.

Speaker A: So tell us a little bit about yourself, tell us a little about your career and where are you right now and what are you doing?

Speaker B: Yeah, so uh, I'm Akos, uh, I'm originally from Hungary, I live in Switzerland now and I uh, come from a technology background. Um, I went to university in a combination of computer science and business studies and uh, and I wanted, went into a technology area quite early on, got uh, into leadership roles and then became uh, uh, retrospectively what I could position as a serial entrepreneur uh for about 13 years, uh, founding service companies, spinoff startups, successful ones, ones that failed, so all sorts of interesting things. And then uh, afterwards I um, relocated to Switzerland uh, and then I kind of revamped my career. I became an employee again at an early stage startup. Then I went to work at big tech. Like I worked at Google, I worked at Meta, uh then uh, as a chief data officer in a scale up company. And now I'm doing mostly advising uh, companies and being an expert in various contexts. So that's what I'm focusing right now.

Speaker A: Okay, thank you very much. So like I said, a wealthy career that having touched the big tech parts and the entrepreneurship and now you're basically a consultant for other companies, right? So tell me a little bit, what is it like, what is it like that switch from entrepreneurship into big tech and from big tech into startups and from startups into consulting on your own.

Speaker B: Uh, I mean these are all major um, major changes in um, agency, uh, and what you can perform, what you can do, what kind of impacts you can have both on your organization itself and also on the market or on the number of customers. Um, and I think these are, this is also extremely educational. Basically if You're a founder of an initiative, then obviously uh, you know, your way of doing things has a significant impact on that organization. Uh, by virtue of, you know, whom you start to work together with. Uh, you know, what kind of people do you select to join your team? And of course they select you at the same time. Uh, so that's a significant impact and is a significant signature of yourself in that organization. Uh, when you join another company they already have that set up in a way, right? So you can kind of relate to it, but you're going to influence that that much. Um, so it's a very different game to do that. Especially if it's a large organization. Then you either like it or don't like it, but basically that's that. So it's a very different uh, kind of uh, uh, relationship with how things are happening and what the culture is and uh, all that stuff. So I think that these changes are uh, they always put you in a different position in terms of agency and ah, what you can change and what is it that you can move forward

Speaker A: and what then is the biggest challenge of making that change of those different positions?

Speaker B: I uh, mean one of them is obviously um, appetite for it. Let's put it that

Speaker A: way.

Speaker B: Of course if you're um, you know, a team member in a larger organization, most probably you know, you have your kind of department or you have your org unit that you're part of, um, you know it gives you a lot of scaffolding, a lot of context, a lot of help that that is already there and you can just rely on and then basically you like move forward with the, with the uh, with the projects and initiatives that you might even come up with. Um, but if this is like an already running company then you already have the clients, you already have products, basically you're kind of iterating and improving on that or introducing new products for uh, uh, kind of an existing client base and things of that sort. As opposed to when you are starting an initiative yourself as a founder, this is all something you have to build up from zero. Um, so you have to have the appetite for that, to pull that off, but also to take the risk, uh, and also to like you know, go in and uh, convince uh, people that what you're doing is verify for their attention, for them, spending on it, for them investing their own resources for the collaboration with you. Um, and so basically you have to be able to, you have to be willing to, and wanting to enable to convince people of that for example. And you also need to be able to build up the organization that is going to deliver on those promises. And you have to be willing to take the uh, responsibility that this is all up on you. And ultimately you have to compensate your team members at the end of the month, every month, and it's up to you to get that money in. Um, and it's not just like HR is going to issue the payroll at the end of every month as in a large organization. It's a very different appetite for risk taking and decision making, um, between these spaces.

Speaker A: And that's what I find fascinating is the very fact that you can find a balance between the two. Because from my experience, a lot of the time you have people that are either suited for one or the other, like natural risk takers that cannot stand still and despise the corporate ladder and everything that comes with it and just want to do something on their own, regardless of what that is, or people that are very natural in that sort of environment where, okay, I have a structure, I have a logo behind me, I can work with a large selection of different people from all over the world. And that's fine. When losing that, being on my own, just with my two hands and the limited resources that I have, that becomes an impossibility. Uh, how do you manage to find that balance? How do you switch from one onto the other? Or are you just drawn naturally to one all the time?

Speaker B: It's not easy to be honest. So in my career, actually I did it a little bit atypically because after some years working as a team member in existing companies, I became a serial entrepreneur. Now I need to point out, with a friend. So it wasn't um, uh, me alone was the two of us who started doing things. So that's already a team and collaboration and so forth. Uh, but one of the reasons I started doing that because I disliked all the companies I worked for.

Speaker A: I

Speaker B: didn't like, uh, the corporate culture, I didn't like the climate, I didn't like the attitude. I didn't like being told off, uh, when I wanted to do things my way. Uh, and then I thought, okay, why not create a company where I would actually want to work for, uh, or work at. And, and that was one of the, one of the, uh, reasons I went in that direction. Uh, because I wanted to, you know, I'm spending eight plus hours per day in my, in my, on weekdays doing work. I want to spend that time in an environment which I enjoy. And so, so that was interesting. Um, one of the, one of my motivations there. And so when you come into an existing organizations, you kind of have to, you know, live with it. You know, that's what you get. And also it's an interesting thing that in uh, when you're a small organization and you work with clients, you, you have that freedom to say to any client that you have the freedom to walk away from any client. So you can say, okay, nice. You know, now I don't want to work with you because I don't like working with you. Obviously other. Sometimes it's you position it even that way. Sometimes you have different messaging, but it's your freedom to do so. Right. And obviously you're not going to get the benefits of that collaboration, uh, but it might give you an easier life or your stress levels are reusing, uh, and things like that. This is not so in a large org, like if you have other teams you have to collaborate with, if you have the organizational structure set up, then you can't just say, oh, I'm going to repurpose my team and do something totally different, uh, because there is other teams I want to disengage from. That's not something you can do. So in that sense you're in a situation where you really, really have to work hard to make all these connections and all these engagements work. And sometimes you can't and then you have to tolerate that or you have to live up with it. So this, this is a very significant difference when you work in a large established, uh, organization as opposed to when you're on your own and you decide who your clientele is.

Speaker A: Okay, that makes sense. That makes sense. Then I guess that desire for freedom has always been inside of you and then the clarity of the vision of what you want to build came with it at some point. Um, but I also understand you do some work for the European Commission as well, right? Can you tell me a little bit about

Speaker B: that? Yeah, sure. So, uh, I'm on the jury of several European level R and D funding frameworks, uh, specifically the European Innovation Council or EIC Accelerator and transition R and D funding frameworks. And I also work for a couple of national R and D funding bodies and similar capacities. So basically there we are um, deciding if um, if companies who are applying uh, for R D funding they get the funding or not. So it's like um, uh, like Shark Tank but for public funding. And it quite literally so because the teams, they come, they pitch in person, you know, we, we ask them questions and basically the jury decides there and then if they get the funding. So, so that's something that we do there. Uh, I want to of course emphasize that I'm one of very, very from a very large pool of judges, so jury members. So I'm not ah, in all of the decisions obviously, but that's something that I do. And it's uh, always a very interesting exercise. The initiatives are for the most part very interesting and exciting and it's very interesting to review them and see what um, you know, people are up to, companies are up to where they want to go.

Speaker A: Yeah, that's a very interesting also additional point of view, at least for you. Right. You've been inside of companies that are large enough, then you've been in charge of companies that would seek that kind of funding. And now you're kind of on the sidelines and you see what kind of companies actually get funding. So maybe some insights you can share over there, like about the whole, um, European Commission selection process is what kind of companies do you feel should usually apply for those kinds of incentives? Is it usually companies that are more mature, that are banded together in conglomerates and stuff like that, or is it mostly for the younger companies to get their first feel for funding and then scale up a little bit from that? Or is it a mix of everything? How does that look like?

Speaker B: So actually, uh, the Commission has specific, uh, funding frameworks for various stages of companies. Uh, so any stage you're in, you're going to find the funding framework for the concept they use. And this is something this company should get familiar with is called the technology readiness level. And this is a concept originating from NASA actually. But the commission has embraced it, uh, quite wholeheartedly. And basically it's kind of a nine stage view of like how do you get from an idea that you just had to world domination? And so basically each of these funding frameworks are tailored to various specific technology readiness levels. Changes with the idea being that you enter at a particular stage and you exit at two stages higher or something like that. So that's a typical scenario. So all stages can enter, but in terms of like who should find, who should look for funding here. Uh, so if you can raise money from uh, the commercial space or FFF or whoever, then do that. Um, um, because the whole concept here, the underlying concept here is that, uh, the Commission is, and most public funding bodies, they intend to fund projects which are not able to raise funding from traditional sources. And this is uh, very central to this concept. This is public funding, this is not charity. It's being spent very strategically in interesting, innovative, risky, visionary, um, proposals which struggle to raise funding, which are worthwhile obviously and which struggle to raise funding out in the market, maybe because they are too risky or maybe they are uh, um, at a particular stage which doesn't fit typical VC narrative or ticket size or a stage. It feels like that. Or also we've seen companies which have been already out in the market with products, or maybe they have already raised funding commercially, but they're uh, you know, coming up with new initiatives that just didn't fit that funding cycle or the funding logic, uh, that they were already in and then it made sense to fund them. Uh, so basically, you know, if you can fund out in the market, do that, uh, it's going to be faster, it's going to be, you know, the process is simpler too. And if you can do that, that's also going to be a validation for yourself that there is people out there who are willing to risk their money for your vision, uh, which in itself is a validation, uh, of what you want to do. And I think that these uh, uh, public funding opportunities are best suited for ambitious, challenging projects, uh, which have a difficult time raising funding out in the open market.

Speaker A: Yeah, that's a good balance. Thank you for that advice. And I feel that's a very nice way of looking at it. I feel that especially in a lot of conversations that I have on the show is that everybody's focusing on your next round, you know, and uh, specifically from like funds or VCs or anywhere else in the private sector. Just basically like it's a closed loop that it's very hard to escape from, that you're just very centralized on how you're going to build your product from the expectations of like VCs and funds and everybody else. But it's nice to hear that there is a space for more risky and more innovative products that don't find that place and that, you know, governmental bodies are willing to put that risk in. Because I have personally seen a lot of interesting pitches, uh, of companies that do not get funding, generally speaking, just because the R and D of it is too complex. Like biotech for example. You need to do a lot of research before you're ready to do anything specific in biotech. And um, for your typical vc, he wants to see, okay, when am I going to get my 10x in a year or two or three or whatever. And it's very, very hard to justify that with biotech. But at the same time the real change comes from those types of risks, right?

Speaker B: Yeah, that's true. At the same time I would argue that there are VCs who focus on biotech. And so basically each of the verticals have specific VCs for them. It's just that you have to address those VCs and if they know that area and they're going to know what the dynamic is and how long it takes to do actual experiments in the lab and get the certifications and especially if it's medical use case like you know, human trials and all that. So, so they would know uh, this area and they wouldn't shy away from it. But at the same time, you know, as we discussed, sometimes we just can't find it, or sometimes it's just too early stage or for a commercial VC to come in as yet and you don't have the set of friends, family and fools that would fund your vision. Um, and so uh, in these cases it makes sense to go down these avenues.

Speaker A: Yeah, completely agree. So then let's talk a little bit about your personal entrepreneurship. So you've mentioned that uh, several times that you have transitioned from engineering into serial entrepreneurship. How did that look like? What was your first company like and what was it about?

Speaker B: So the first company I would call up in this space was actually a service company, uh, founded with a friend of mine and it was in retrospect very typical East European talent working for providing the talent to the northeastern US market, American market. Um, uh, and that was uh, uh, you know, still today. I mean that's, that's a, that's a business model that works. Um, and, and uh, it was very uh, good because the two of us who founded it, uh, you know, we already had uh, we had the talent but we also had the first client to work with by virtue of my co founder. And so we could bootstrap that without basically any significant investment. Um, and that's also like a very key learning uh, from there is that you know, if you, if you have, if you're in a scenario where you can offer something that somebody's already willing to pay for it, you know, already next month or in two months or something, then, then you can bootstrap your initiative and then you're basically free of any investors or any additional players. And it's very interesting uh, trajectory to take. Uh, but we also, what we did there is that we would uh, do a lot of internal, like basically operate an internal incubation house of sorts and we would look at new technologies that are interesting in themselves or new technologies which can be useful for our clients or uh, you know, that are just uh, exciting not just technology wise, but, you know, have business potential. And this way we spun off a couple of our own initiatives and startups and nurtured them in their own right. And those were then in a way, not no longer service companies, but those were actual startups with their own, um, products and services and so forth. And so that's basically how we kind of leverage the talent that we had in house, leveraged, uh, our business knowledge that was building up during the years. And we could afford to self bootstrap these initiatives as well, which was also very powerful.

Speaker A: But why did you initially even decide to do that? You said yourself the business model works, right? You're bootstrapped, you have the business coming in, theoretically, just find more money, scale it ad infinitum and become super rich and retire to the Bahamas or something. So why that drive to invest into other initiatives that are so distinct from a service business?

Speaker B: Um, so first of all, because, you know, if you're driven by technology, then you're just curious and you want to see what works. Uh, so a lot of these research, you know, a lot of these activities were kind of investments into our curiosity either, uh, from anyone from the team. So here like, you know, the most team members had the option to pitch an initiative or just, you know, not even formally pitch. It's just like, you know, the conversations we have in a small work and you know, one thing can lead to another and just giving these things a try. Um, and, and it was, a lot of it was, okay, let's see, you know, what this new tech does or let's do a pilot project or just, you know, try our fingers at it. Um, so that's one part. The other part is that, uh, service companies don't scale. Ah. So you know, your notion of, uh, scaling it up, it doesn't scale like that. That's not what service company looks like from business model perspective. Uh, the third one was that we were seeing, uh, what our clients struggle with and what their challenges are in terms of their business processes. And of course we are providing tooling for them and creating solutions in adjacent spaces and things like that. But, uh, when you also identify actual pain points that one or a couple of your clients are having, and then you can assume other similar clients are having similar pain points, uh then you can use that insight into creating a solution yourself and then offering that, uh, as a solution in itself as opposed to you developing that for them, uh, on a time and material basis. And so these were kind of the dynamics that were going on there and Also like one of the startups I founded then was uh, just a passion project in that sense. It was like a combination of interests and technologies and timing that just made me want to do it.

Speaker A: That's an often undervalued thing I feel in entrepreneurship. And I really like what you said, that curiosity is inherent to technical people a lot of the time, which is true. And business people tend to lose that curiosity with time just being very profit driven where it does not matter what the business is as long as we're making some money along the way. But yeah, the limitations of scale for a service business would scare anybody off and I understand that. So what was then the winning next, uh, business? When did you decide that, okay, we've done enough with services, now we're going to focus on this next thing? What was that next thing?

Speaker B: Actually it wasn't uh, a shift. We kept the service business going. So this was like trying to sit, you know, trying to ride multiple horses at the same time. Um, which in itself is an interesting challenge. Uh, but, but one of them was a machine learning startup back in 2009. Uh, now we would call it AI obviously. Um, and uh, and that was at the time, you know, just backpedal 17 years ago. Right. Um, so, so that was a time when, when uh, product and content recommendations were not a widespread technology. Like Amazon was recommending new products to buy Netflix. It was already online and was recommending you to look at other movies. That was the time of the first Netflix challenge where Netflix would publish a big anonymized data set that you could then uh, try to predict what people would want to watch and then you could submit your algorithm and then you could win that challenge, for example, things like that. So uh, basically when we were working with our E commerce customers, we also saw how much effort they are, effort and money they are spending on improving customer KPIs. Like um, uh, average cart size, shopping cart size, shopping recurrence, uh, like how frequent do the customers come back? Things like this. And then we kind of put one and one together and then we figured that okay, if we can create a similar recommendation solution that is available for a larger number of organizations as opposed to like, you know, Amazon at that time would have their internal AI research team or machine learning research team develop it for them. Netflix similarly, most companies just don't have the capability to build up such a team, not even fund it. Uh, so if we do that and we offer that as an SAS service, then we would have more companies use this kind of technologies and they would be using our technology for that. So that was for example, one of these initiatives.

Speaker A: Okay, that's very interesting. Um, I guess it would have been very hard to predict where those initiatives would lead some 17 years later, of course, but at the same time you were involved in them, so you probably felt something where the wind was blowing. Uh, and then you mentioned that you left entrepreneurship for a time and went on into big tech. So how did that decision happen?

Speaker B: So that was actually a personal decision. So I got uh, m, like me and my family, we frankly got fed up with the social climate in Hungary. Um, and this was 2008. Uh, so surprisingly enough, an election year and uh, there is always the last drop in the bucket, right? Uh, and then, uh, for us that election result was that. And basically then we decided to relocate. And then my kind of personal goals became different. So instead of um, uh, you know, like business, you know, entrepreneurship wise or career wise or whatnot, my personal goal became to relocate to a nice country and just, you know, pull up my sleeves and do whatever it takes to get there. Ah, and do that and establish ourselves and things like that. So, so then the, the simplest way was that, okay, let's get back on the job market, let's find the job. And then just. And this is what happened. And first I joined. So then this is when I moved to Siserman. And um, uh, uh, this was 18 maybe I said 2008 earlier, but this is 2018. Um, and so uh, then first I joined an early stage startup here actually which was working on things that I was uh, somewhat familiar with. Uh, that was AI based flight autonomy. So basically creating an AI based pilot for aircraft. Um, that was what the initiative was about. I spent there one year. Then uh, I looked around in the local job market and then I joined Google, um, as a kind of next step to facilitate us being here. Uh, and that was a very interesting experience obviously also because this was during COVID So my interview was, uh, my interview session was at the first day of the COVID lockdown at the Google office in suitcase. So we couldn't actually have the interviews in person. Um, which was interesting. Um, but yeah, so, so I joined Google. Ah, I worked on YouTube. YouTube search and machine learning. Ah, site reliability of course. So that's a big scale, 2 billion users. Um, my team was responsible for 12 hours per day that uh, search and machine learning services on the YouTube site work, um, five minute on call, also on the weekends, things like that. Um, and then what can go wrong? Um, uh, and then within Google I transitioned to Google Calendar which is half as many users, only 1 billion. Uh, but there I was working on the software engineering side, so feature development side of things. And Calendar is a way more complex product than one would imagine. Especially the free Calendar has way fewer features than the higher tier business versions.

Speaker A: M

Speaker B: um, and then afterwards uh, Meta Tsuri called me up and they were doing the AR VR product line development here in Swedish. Um, so, and that's an area I'm personally very interested in. One of my startups was in the space of ar and so that was interesting and I went over and joined Meta and there I worked on the Quest Pro and Quest 3 and the Orion product line. Um, so AR VR physical products, uh, which is of course engineering wise it's also different but it was quite interesting. And so this was kind of my foray into big tech companies. And then from Meta I went to a scale up company as a Chief Data Officer, um, basically joining the leadership of that company. So that was kind of the employee trajectory here in Swedish, so to speak.

Speaker A: So then uh, first of all you need to give me some tips as to how to appear better in searches. Uh, and second of all, uh, what did Big Tech then teach you? I'm guessing you didn't return to like smaller scale startups and then into your own consultancy just through nothing. Right? You said yourself you have that desire to do something of your own to make an environment that is up to your standards and not somebody else's. But I'm guessing Big Tech did teach you something about how to run a business and how a larger business functions.

Speaker B: Yes, yes it did. I mean it's obviously a humbling experience when you go in and with the heritage of these companies and the size of businesses and the pervasiveness of the solutions that they build and the number of users and the scale. It's very humbling. Um, I mean, and the YouTube SRE, it's also extremely humbling when you have the best resource base at your fingertips. Right? I mean, you know, when there is an, when there is an incident and sometimes the solution is literally just to throw more compute at it and then you have a lot of compute because it's all there on uh, uh, the other end of your console. Um, so that it is very humbling. It's also very, you know, having come from a self built background and self booster background, it's uh, the, the impact you have is also tremendous. Like, like for me it was, it was a very interesting experience to think that, you know, a 1% error would have a 20 million person impact to it, um, with 2 billion users. And when you build your own services and when you build your own products, you fight for every single person you impact or every single customer that you're going to have. And here, by virtue of being a member of this organization, that multiplier is huge. And frankly, I don't think everybody who worked there realized that how interesting the valuable that is. You might be doing the same amount of coding, you might be doing the same quality of coding as uh, someone in a smaller organization, but the impact that that code is going to drive in terms of how many people you're impacting is that multiplier is just tremendously different. And that's a very interesting aspect of it. Um, at the same time, this was already at the time when the culture in these organizations was changing. Um, so I would say that most probably I kind of saw or almost saw the end of the good life at Google, for example, and especially as I joined during COVID where we were not in the office, so it was really difficult to meet people and meet your team members and so forth. So it was also very difficult to get a sense of that end, so to

Speaker A: speak. M

Speaker B: Nowadays these organizations are very different than what they used to be or what kind of the public assumption of how they are is. Uh, so that's also interesting to see how even these orgs with the reputations they have, they change in time and um, and uh, they become different.

Speaker A: So what do you think is then the main change that the public isn't aware of?

Speaker B: I mean, you know, as a service company they are all there, so we know how they operate. I mean, I guess that's for the general public. I guess, you know how the Org is set up and what the team member experiences. That's for people who aspire to join. Um, I think it's like most people in the technology know that these organizations have changed, uh, especially even since then with the layoffs and everything. So that's an even bigger cultural change.

Speaker A: Yeah, I think that can't, uh, speak for everybody, of course, but generally speaking, the public perception for Meta, for example, is okay. Mark Zuckerberg has been building something in the metaverse for a long time and then that didn't work and now he's trying to do something with AI and for Google, I think, okay, people underestimate how much Google actually owns in terms of infrastructure that everybody uses day to day. But at the same time I feel that everybody should remember when they removed, don't be Evil from their modus operandi. And that kind of the general shift that people might be aware of.

Speaker B: Um, yeah, I mean, the signs are all over, right. Of what the direction is that, uh, people are trading. But maybe we can touch a little bit on the layoff cycle. Right. So I think that pretty much kind of underlined or made the culture shift clear. Um, because with the layoffs, people who were performing well were fired from the company.

Speaker A: Mhm.

Speaker B: Uh, and that didn't happen before. So imagine a company which is almost two decades old, which always, uh, was proud of the engineering workforce that they have. And they used to list the number of highly skilled engineers as part of their assets in the quarterly stock exchange uh, publication filing. And so this company, which prides itself on the engineering workforce that they have, is one of their main assets, one of their value drivers, decides to get rid of people who are performing well, who are not failing on their performance review or executing what they should have been, who landed impactful projects, and they still decide to get rid of them. And this in a situation where they were still tremendously profitable, actually, the profit margin was not shrinking for any of them. Uh, so you could not really highlight a strong financial incentive to do that. Um, so that was like, you know, it's an end of an era in a lot of ways. Uh, whereas beforehand, if you were, uh, an engineer who performed to expectations or better, then you could just assume that you're just going to stay on because it's a company that basically transforms engineering output into business value. That's what it does. You're the one who is generating that engineering output. And then the company as an org is going to transform that into business value. And this contract was broken with the layers. And then by making clear that there's always going to be layoffs like this, um, afterwards, I mean, then the writing is really on the wall and culture changes as well with that. Um, you know, people are not as open to sharing things because it's going to be like a battle royale at the end of the day. Um, you know, who is staying, who is not staying. Um, it becomes competitive, but not in a good way because of that. So it goes a very long way. And obviously from the way I'm narrating this, it's obvious that I don't think it's going in the right direction.

Speaker A: M I definitely agree with you on that. I feel that you've said it very, very well when you said that the contract is broken between what a tech company used to be and what it is now. And uh, I think it has a ripple effect into the whole industry as a whole. So if the top of the industry has changed its mindset, then most of the industry follows suit without really necessarily making the same changes internally. Right. Some are not as profitable and they do need to make those layoffs. Others are even more profitable maybe, but still do those kinds of layoffs. And I know it's a long philosophical question, but since we did start talking about it, why do you think that change happens in the end?

Speaker B: Frankly, I don't know. Like uh, um, I really don't. Uh, there's a lot of speculation on how and why. I mean obviously one of the pro forma reasons given is that they over hired during COVID and uh, but it still doesn't change the basic contract that you have a lot of skillful engineers and basically as an organization your job is to turn that into business value. And I didn't see them running out of problems to solve. Um, so why not solve them and then, um, you know, why not do what you're supposed to be doing? So, so that, so, so that, that, um, so I don't buy the argument, to be honest. Um, you know, obviously this is a change in culture. So that is what this is indicating that there has been a change in culture. And this is a, uh, great signal that first one can investigate and probably would need to dive deeper into why and how the culture change happened. I really want to speculate, but this is a signal.

Speaker A: Yeah, that's true, that's true. But the very nature that you don't have a simple answer is telling because again from the public's perspective, I think the narrative is pretty clear. We don't need as many engineers because now AI does everything and engineering itself is no longer valuable. Uh,

Speaker B: layoffs happened before AI coding came about.

Speaker A: True,

Speaker B: they

Speaker A: started before, at least they continued throughout. But that is, I think I've mentioned this before in the show. I think that's very convenient excuse a lot of the time like AI has its uses, you can agree or disagree, but like it's not replacing a good engineer in a team one to one, almost never sure. The process of writing code has been commoditized. Everybody can write code, but again to a layman, maybe that's what most engineers do, but that's not what most engineers do. Writing code is the first and the most simplest thing that you learn as an engineer, like solving problems, having the vision of uh, a large scale system and how to control it, how to scale it, how to maintain it. That's what an engineer's value is. Well, I, uh, want to put words in your mouth. I don't know whether or not you agree, but how do you see this whole AI situation influencing those kinds of cultures?

Speaker B: Uh, I mostly agree with you. At the same time, I think that there is that level of writing code which is extremely highly skilled,

Speaker A: of course.

Speaker B: So there are those highly skilled experts who are like, you could, you would call them coding machine inside these big tech companies, but basically who are extremely efficient in writing very high quality code. And, and uh, and, and those people, the impact of those people on, on their larger org is, is, is significant. And, and it's uh. Um. So, so I wouldn't, I wouldn't just say that, you know, writing code is the first thing you learn. I mean there is the levels doing that. Uh, but it's interesting because a good friend of mine who is such a cool person that I just described actually told me in the summer that the way he, he narrated it was beautiful. He said that his profession is gone. Is. It's not his job. His job is not gone, his profession is gone. Um, what he is very extremely efficient at and what he loves doing is no longer there. Um, basically even him at his level, he is working with AI agents. And um, obviously that's how you multiply your output and so forth. Um, uh, but, um, it's a significant change. Um, but I guess your question was mostly about how do we relate to that. So this is how we relate to it. So basically instead of you writing code firsthand, you become a manager of agents.

Speaker A: Uh,

Speaker B: um, and this brings with it all the peculiarities of such a change. Even if when you were elevated from an individual contributor to a team lead or a tech lead, that basically instead of handwriting the code, what you do is that you coordinate with other people or other agents in this case, so that that code is written. And then you do that coordination, you supervise the results and you make sure that that code is there and it's being written. Um, and that change is not so easy for everyone as even in the old days it was never easy for a person to grow from a person who individually creates good architecture and writes good code to a person who is working with other peers to do the same. That's a different, uh, skill set. It's something different that you have to do. And in this case the nature of the change is different though. So this is why a lot of people, I think at first they enjoy the efficiency of, oh, I can just tell it to do this, tell it to do that. But at the same time you find yourself basically pestering agents left and right to get the job done instead of you diving deep into it and you getting it done. And that's a very different, uh, kind of activity. Um, but this is how we scale now. And also I see in a lot of ways that the people who expect that, okay, I can get the agents, write some of the code, but I still want to review it, all of it. I think they kind of missed the point because the point is that they're going to generate so much code that you won't be able to review it. Right, that's the point that you get so much code generated that even reviewing by yourself is out of the question because of just the amount. Um, and that's what we have to get used to now, basically. It still means that we need the expertise, obviously, because you need to ask the right questions, you need to be able to review at scale so that you focus on the important part of it. And then of course, when it kind of starts to fall apart, then you need the expert insight to immediately see what's going on or what the issue could be at any level of the code base. Uh, so a lot of that expertise is needed, but you have to utilize it differently.

Speaker A: But that then breaks most conventions of how you're supposed to work though, right? Uh, like you said with your friend, the profession is gone. Then what's left? You get, uh, much less focus on how you're building because specifications and analytics and business decisions are done in the same environment. Right. You're not like, only isolated to coding. Your head of business development or whatever is the same with their agents, where they're looking at the market, trying to predict where to go. And marketing analyzes huge amounts of data through AIs and AIs down and down and down and down and goes where everybody's just integrated already so much where the whole process of delivering quality products has changed. Right, so how does that then look in the future? Do we need to create more products or is it just going to be kind of like natural selection where everybody builds products the same way? Most of them will not make it because most of them are not that good. But some of them will probably be good just by that monkey and typewriter thing where you give them enough and at some point it will work.

Speaker B: I would turn this around a little bit and look at it from, ah, from the customer's perspective. Um, so now because these tools are very powerful and they give you agency, they Give you a lot of agency to yourself as well. And then when it's an agentic AI, then you gave the agency to the uh, to the AI, so it can not just recommend stuff but actually execute, reach decisions and execute on it. That's what an agent is, that it has agency. That's why it's an agent. Uh, and if I turn it around, what I see is that uh, a lot of people who are working in a specific area, like you know, you're an architecture studio or you're a bakery or a cafe or whatever it might be, uh, now you can roll your own mini tools. You don't need to find a solution to your bespoke problem on the market, or you don't need to hire an external engineering team to develop a custom solution for you. Uh, you roll a lot of what you want on your own. And because you know your problem space, you know your domain, you're the expert, uh, and you know your tooling. And that's a, nowadays in every profession, people know their tooling, know their digital tooling as well. Um, then having these facilities at your disposal, you can just create your own custom bespoke, uh, connectors, integrators, mini tools to do your bidding. Um, and that gives you a lot of leverage, a lot of power, uh, without having to employ an external engineering company or even a single engineering consultant or anything like that. And I think that that's very powerful. Also at the same way, if you look at how to get menial things done, uh, at a kind of mid level quality, which is mostly what you need for anything, that's not your differentiator, that's not your core service. Like you run an organization, there's administration to be had, there's finances to be had, there's things like that.

Speaker A: Sure.

Speaker B: Uh, you can go very far now without actually employing a specialist. Um, and so basically, uh, a lot of the kind of mid range needs that you have can be or will be covered by this kind of tooling. And I think that's very powerful and it's great. It also has an impact on the service providers who traditionally would have provided those services. Obviously. Right. Because

Speaker A: then

Speaker B: you're not gonna engage with them for menial things that are not unique or not special or anything. It's just needs to be done right. And then there's always the specialist, that is always the difficult, there's always the high end, um, version of everything where you need the specialists and obviously that space is there. Um, so this is how I look at it mostly Is that uh, it's a great enabler and then coming back to like okay, what kind of product services and what survive. I think it also brings in the age of the throwaway mini solutions,

Speaker A: uh,

Speaker B: um, as in you roll your own and you're never going to maintain it. You just write a new one in six months. Um, and because the landscape has changed and the integration points have evolved and, and now you're the architecture studio now working on a different building product building project and um, you need a different kind of tool for that. And so in that sense also the maintenance question is very different now. And basically you need very strong reliable foundations that you can uh, build your bespoke tooling on and your solutions on which are going to be probably temporary and not maintained in long term. As long as you have these foundations and integration points, um, I think then you're quite well set. And so this kind of translates into what is important and foundational services are important that you can basically provide uh, the important facilities of your profession or your vertical. It's important that these have integration points, machine capable integration points or some kind of API or maybe even an MCP endpoint, something um, uh, and there is a clear bias towards that now. So up until recently, yeah, machine ah, integration, I mean it became more and more important but it wasn't like uh, a really must have and now I think that's kind of a no brainer now. Like let's not select tools and services which you can't integrate it because then you can't actually have the uh, you know, your own tooling created easily. Side note, you can always kind of mimic the UI experience with whatever. But, but, but, but uh, but so, so I think that that's the direction it's going to. And of course then this rolls into an impact into companies who were, you know, what kind of solutions you want to create or what kind of space is there to create customized solutions or integration houses. You know that uh, changes the landscape quite a

Speaker A: bit. Mhm. Yeah, uh, I can definitely see that happening. Let's call it that. I definitely have much less expertise than you do to make predictions on this market, so I'm not going to try to. But I do appreciate your view on it because it does seem like a very balanced one. You know, no doom and gloom and no transformational change of humanity and singularity or something like that. Very down to earth and I really appreciate it. So at the end of the day would you call yourself like an optimist or a pessimist when it comes to the evolution of the current AI tooling, will it make things worse or better net for us in the industry?

Speaker B: I'm generally attack optimist. So, so I'm, I'm usually uh, I think scaremongering is easy and people are, I mean we know that people are receptive to, to the concept of fear. But no, I'm, I'm usually a, uh, a tech optimist. Ultimately the, the decision is always a societal one. It's not, doesn't lie with technology. So obviously that question is really up in the air now, like what are we gonna do with these? You know, if I look at it from that perspective, improvement in efficiency, uh, so where does it end up, uh, where does it, where does it end up in terms of the societal impact? Is it going to be that people are out of the job and therefore they can't um, sustain themselves? Uh, but then if machines do all the work, why do we need jobs? Uh, so I think that this is really the big societal question of this yet again disruption and yet again another significant improvement in efficiency of how is the value that is generated here, how is that distributed between society or within society, uh, and how we as uh, societies and like the whole collection of societies, humanity at large, you know, how are we gonna uh, make that work for us or for whom? And I think that's the question that needs to be answered.

Speaker A: Not an easy one to answer, unfortunately. I would have asked you this like I am interviewing you. I can ask that question. But I'm afraid that it is a little bit harder than that.

Speaker B: But it's an open question and genuinely interesting. But here it can be genuinely frightening as well. So really I'm really curious on which way it's going to go.

Speaker A: Yeah, we will see. I guess that's the only thing that's left to do for now. We still survive. There's still code being written, there's still jobs to go around, and we have some time to find the answer to that question. So to then kind of wrap up our discussion and to finish, what would your advice be for people that are coming into this industry or that haven't spent as much time in this industry, what they need to focus on maybe what kind of skills or the attitude they need to cultivate in order to succeed in the current climate?

Speaker B: You mean as individuals and trying to become employees or

Speaker A: I guess people from all sorts of walks of life that you have seen, like people that are looking to get into big tech and entrepreneurs that are looking to launch their products and People that are basically on the fence of whether or not should they build products, should they go for the big job or something in between.

Speaker B: Yeah. So from a technical knowledge perspective, one still needs to learn the tech. I mean that's table stakes. You need to know, you know, what, what a computer is, how it's running, um, you know what, what the bits in the in edit are, you know what, how it turns into bytes and how many bits in a byte. And that's not even an easy question itself. Um, so uh, you need to know these things and you need to um, educate yourself, uh, on the, on, on the, on the ground level. Um, so that's not something you can just skip and hope not to have to learn. Uh, um, on one side, on the other side you can, you know, you have very efficient, like, like you can go very far. Um, even as a single person or you know, with a very, very small team. Um, and that gives you a lot of leverage if you have like a vision of what you want to do. Um, and here I'm still talking about coding tooling, so not utilizing some kind of AI as part of your, whatever you're doing. Uh, it's just that the tooling super powerful now. Like uh, uh, you know, if you, even if in an area that you kind of marginally are aware about, uh, but you know the basics and you know the book group foundations like now seeing a scientific publication and replicating it as work in code, that's a couple of days now, uh, for the most part. Uh, and that level of uh, speed, that's extremely powerful I think. So if you really look around and you have a vision of okay, this is what I want to get and these are kind of the papers that came about on that subject and now there are these algorithms that can facilitate what I want to do and you can get to a working version on your own machine. Uh, even if the researchers didn't publish any code in a couple of days, I think that's a very powerful environment. And then you can build on that result and maybe you can create your own result that you can publish later and so forth. So, so in that sense moving forward is really fast. Um, so that's a very good environment. But this is kind of also then the bar that you need to have. So that means that yeah, you have to have that mindset and you have to have that expectations, uh, that this is the speed you can move at too. And it's not like a one off and it's not like okay, I replicated same Thing, this is something that's going to be a couple of days. You can replicate two papers per week, um, and then work with those results and integrate that into whatever you're planning to achieve. That's the space, uh, that you're going to be working in. So very fast, very moving forward, um, but also enabling you to do so. I think that's super interesting.

Speaker A: Yeah, I think that's a very important part of it. Right. That it is there. It is enabling you to move that fast, but so does it. It does the same to everybody else. So now you have to have a better idea where you're moving. And I feel that that movement is life in the end of the day. Not moving is the worst of the two, I feel.

Speaker B: And ultimately it gives you the freedom to focus on the things you want to focus on because you can touch these areas. And then we know freedom is associated with fear as well and anxiety. So it's not always easy, uh, at the same time. Uh, but in terms of needing problems to be solved, I don't think we run out of problems in general. Uh, so I don't think that we are. The situation where you would like, lay back and say that, oh, there's nothing more to do. Uh, there's, I mean, just look, let's look around. There's a lot of. Yeah, so, so in that sense, I don't think that, you know, that that kind of nihilism is here. It's, it's more that. Okay, then, then let's focus on what we want to focus on personally and contextually. And of course, if you want to make money, like, what can you make money out of? And then, then you go down on the path with this very enhanced toolings that you, that you're. Or at your fingertips.

Speaker A: Yeah, no, well said, well said. I agree with you on that. Thank you, Akash. Um, thank you for being on the show as well. So where do people find you? Where can people read more of you or hear more of you or see more of you?

Speaker B: Yeah, so, um, people can find me on LinkedIn, I guess. That's the easiest. And there I also post sometimes of, uh, what I'm thinking about and, and uh, that's the best way to follow what I'm doing.

Speaker A: Thank you. Thank you very much. And thank you as well for taking the time to listen to our episode here today. What have you found the most interesting out of everything? I believe that the balance between the speed of creation and the versatility of the solutions we can now create, coupled with a very optimistic view of the future is at the core of our discussion here today. You have the capabilities right at your fingertips. Now, it is easier than ever to build your own startup, open your own company, go find that job, and learn something around the profession that you're willing to invest into. But at the end of the day, you need to do it. So do things like liking this video, subscribing to the channel, and telling your friends about it because the algorithm likes it. Akush built it that way. Thank you all, and bye. Uh, bye.