Sign up to save your podcasts
Or
Share 66 – The Global Tech Labor Reset
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
66 – The Global Tech Labor Reset
The reckoning is here. Once a safe harbor, Big Tech has finally also gone full out on layoffs. Is this a structural shift to employment in Tech? Will the subsequent talent spillover be great for start-ups and entrepreneurship? Will it positively affect other industries?
In this episode of Tech Deciphered, we will answer these and other questions in a deep discussion on the Global Tech Labor reset.
Navigation:
Our co-hosts:
Our show: Tech DECIPHERED brings you the Entrepreneur and Investor views on Big Tech, VC and Start-up news, opinion pieces and research. We decipher their meaning, and add inside knowledge and context. Being nerds, we also discuss the latest gadgets and pop culture news
Subscribe To Our Podcast
Nuno Goncalves Pedro
Welcome to Episode 66 of Tech DECIPHERED. Today, we’ll talk about the global labour tech reset. Tech and big tech, which seemed immune to any lay-offs, seems now to be under fire. Massive lay-offs over the last 2.5 years, a lot of discussion around the importance of having a computer science, computer engineering background, and so what seemed to be a safe haven for any graduate is now under stress.
Today, we will discuss the structural perspective on what’s happening in the market, if this is a long-term trend or not, what has led us to this, and what is next. We’ll talk about the rise of fractional freelance and solopreneur work, as well as talk about talent spillovers, and some of the usual geographical dynamics around the space. Bertrand, a huge shift in tech.
Bertrand Schmitt
Yes, definitely. It’s pretty big. I think it started probably around 2022, once we got some changing interest rates that have a pretty massive impact in stock prices for a lot of companies. At that time, a lot of companies decided, and usually under some pressure, that it was time to be more efficient, to generate more cash. Yes, you want to grow, but not grow at all cost. You have to go efficiently.
That’s when we started to see some share price going down and step by step, quarter after quarter. Some change in attitude with a lot of big tech and that has created some impact in term of lay-off from different parts of the business, from the sales team to the DNA, to even engineering R&D.
What is also been happening since 2022, 2023 is a change of focus. A lot of focus is being put in AI. A lot of investment in CapEx is going to AI. At some point, if you want to keep doing all this investing, investments, you might have to get some other part of the business in order to create additional savings to do all the spend you can in AI. There has been more recently a switch. It’s not about just efficiency to push all the… But generating the ability to invest in AI.
Nuno Goncalves Pedro
It’s part of a broader movement. Before we step back a little bit and go back in history, even recently, we’ve heard that there’s talks between Meta and a bunch of private equity firms like KKR, Brookfield, Apollo, and others, to actually help in financing data centers. Meta is a gigantic company, so one would assume they have cash to do all these things. Maybe they don’t.
To your point, that level of efficiency that is now needed in the market where you need to throw actual money, CapEx, into the building of infrastructure, the building of the core underpins of what you’re doing is pretty vital. But let’s go back a little bit of a second, and we’ve talked about it maybe in our early episodes.
Companies like Meta, Facebook back in the day, Google, Alphabet now, and others had a tendency, in particular in the Bay Area, to hoard engineers. They would over-hire. There was a lot of discussion that some of these players had as much as 20%-50% the engineering resources they needed, and part of that is that they didn’t want their competitors to have them, and so in some ways, we went from hoarding too much capacity to the days of COVID that in some ways illustrated that there were a lot of people that maybe were not working very hard also on the engineering side, not just on the sales, et cetera side.
That right-sizing, I think, has been a long time coming. We discussed in a past episode that there was a little bit of an adaptation, or an adaptation, rather to the market by a lot of these public companies saying, “Now that I have a mandate where there’s lay-offs happening in other industries, I can do it myself.” But this time is different, and I think the message of probably today’s episode is going to be around that. This time is different.
There are some fundamentals that have changed. Bertrand has already alluded to AI as one fundamental shift in terms of the efficiency of engineers, in particular more junior engineers. There is more behind this. There is the notion that you need less to deliver more. In tech, I would start with a platitude. I think if anything, everyone talks about AI and how AI is going to disrupt financial services, and it’s disrupting functions like marketing, and displacing people around, for example, customer service, and all of this. But a lot of people forget to talk about the most obvious industry that’s being disrupted by AI, and that industry is tech.
Tech is disrupting itself with AI because you have all these tools available, you have all these platforms that you’re deploying, and they have value for your own purposes. They become the underpinning of a more efficient workforce, in particular if you’re a tech company. A lot of people don’t talk much about that, but I think a lot of what we will talk today around lay-offs and efficiencies and all of that actually do relate to the fundamental shifts we’re seeing where the emergence with a lot of these AI tools and platforms that we’re seeing, so tech is disrupting itself at the end of the day.
Bertrand Schmitt
I think AI has two effects. One is it’s so big, it’s so important, you cannot miss it as a business that you are going to reallocate resources. From your traditional product lines, your traditional development, expansion, creation of new features, and you will reallocate instead more money to AI. It can be one side, can be just CapEx, and the other side will be, of course, software that you want to develop connected to AI.
I think that was the first effects that started early on. What it means is that many companies ultimately decide to fire people who are not so AI knowledgeable and instead hire people who have more expertise in AI. I think what you are talking is a second effect of AI. Now we start to reap the rewards of the investment in AI. What we are seeing as a result is that we can do more with less.
Basically, if you’re a software company, or a tech company, you have a lot in software engineering. If you have a lot in software engineering, guess what? One of the best way to leverage AI is actually to use it to develop more efficiently, faster. The tools have gotten better and better, what we got 2 years ago was barely usable, 1 year ago start to be good. Now, I must say, it’s shockingly impressive what you get if you have tool like Visual Studio or if you use some other tools on the market that are even more focused on AI.
It’s quite shocking the quality of the code auto completions that you can get. It’s quite shocking what you can get from other tools directly like ChatGPT. It’s definitely like having a buddy that can support you, check your code, provide you some IDS, run code, help you go faster through documentation, testing. You are going to save time. The first effect might be, you know what? We can do so much more with the same workforce. Maybe we can reduce slightly the workforce, and we’ll still do more thanks to AI.
Of course, there might be a question of you might want to keep the same size of software engineering so that they deliver even way more than before and your pace of development is truly accelerating. Companies are facing multiple options here. It looks like many are choosing to be more efficient.
Nuno Goncalves Pedro
Yeah, there’s just… Let’s talk numbers, right? In terms of lay-offs, 2024 alone, it seems like the number was around 250K-260K, globally, people were laid off.
Bertrand Schmitt
In tech.
Nuno Goncalves Pedro
In tech, yeah. This is global tech lay-offs, right? In tech, 250K, 260K tech lay-offs in 2024. In the US, it varies. We have numbers that go from 95,000-150,000 tech jobs in the US alone that were laid off in 2024. We’ve had numbers that over a period of 2.5 years, 400,000 people in tech were laid off globally. These are massive numbers. This is an industry that by itself was already, in some ways, part of some efficiencies. It’s not the most inefficient industry in the world in some ways, so very significant numbers that we’re talking about.
Bertrand Schmitt
It’s not just lay-off, it’s also who you are not hiring anymore. I think we found an interesting stats and the quantity of entry-level tech jobs have dropped by 50% since 2019. Before the pandemic, new graduates made up 15%, of big tech hires. Today, it’s just 7%. We’re moving from 15% to 7%. It’s pretty shocking.
Where is it coming from? We talk about general cost-cutting measures post-COVID, but for sure, there is a lot around AI and automation, and especially since that AI is replacing a lot of more entry-level jobs. It can be QA, it can be IT support, it can be some software development. It’s definitely a big change. One thing not to forget is that computer science used to be a much smaller field 10, 20, 30 years ago. It has expanded a lot.
To be frank, I don’t think the quality of the graduates has stayed high. The more you expand the pool of people working in industries, the less you have the best of the best. As a result, you could argue there is a lot of people who are not either so good or so really interested in computer science, who still end up in computer science because that was the thing other people were doing. That was a safe way to get a job after your studies. But it was not because they are incredibly excited and knowledgeable about the space. It’s not because they are excellent developers.
My guess is that when everything is great, everything is easy, that’s good. But when there is tension in term of hiring, then it’s going to be trouble for a lot of people who don’t have the interest, appetite, and quite frankly, sometimes the level of what it takes to be a software engineer, for instance, in that new age. If you are not an expert with AI tools, if you don’t know AI deeply, why would a company hire you?
Nuno Goncalves Pedro
I think there are three important things to highlight here. The first one, and we’ve talked about it before, is, and this might be surprising to some people who are not computer engineers or computer scientists in background that might be listening to us, but there used to be the mythical figure of the 10X developer. Being a full-time equivalent can do the job of 10 good engineers.
Those developers do exist. Shockingly enough, they do exist. I’ve met a few in my career. I’ve been lucky to have a couple of them working with us at Chamaeleon. But amazing people that can do really the job of 10 people. I think what’s happening in the industry, to your point, Bertrand, is there’s a lot more variants in the quality. You have a lot of people that are not great. If you’re not a great developer, the problem is you’re actually probably having a negative impact in the projects that are involved in and slowing them down.
I think the second factor, which is really important to highlight, is some of these roles, and not to basically downgrade what the roles are, they’re very important roles, but some of the roles around DevOps and dev tools, around QA become specialists roles, but they’re roles that are a little bit less engineering-heavy, and they actually match some of the junior developer roles, which are what I’d call breaking stone issues.
You’re trying to do hardcore integrations, convert code, clean up code, et cetera. But at the end of the day, it’s not a fundamental engineering problem that you’re trying to solve for. AI, as we know, is extremely well-equipped to solve issues around that. It’s the issues that are very repeatable, that through testing can have relatively low impact in some ways at the end of the day in terms of the code base that you’re actually maintaining.
Then the third piece, I think, is some people might say, Well, but then is computer science that and computer engineering is that? What matters? No, it’s not. Actually, there are roles in computer science and computer engineering that are becoming even more and more vital. People that have very strong architectural view of code base, of what they’re developing, people that have full on, full stack knowledge of the world, people that have very deep knowledge around legacy systems, for example, embedded system, coding, et cetera. All of those right now don’t seem to be as under threat as other types of roles.
In some ways, we even have a more significant part of the haves and have nots, which you, Bertrand, very alluded to, which is people that have very deep domain knowledge around artificial intelligence. We’re now seeing an arms race for that talent, which the joke is going on that obviously Matt is trying to hire people and paying $100 million in our issues or whatever that is. I’m not sure how much of that is true.
Bertrand Schmitt
Just bonus.
Nuno Goncalves Pedro
Just bonus. I’m not sure how much of that is actually true or not and what the rules are for, but there’s this arms race for that talent. That talent is even more valuable because, again, back to the point, it’s probably seen as more scarce right now. It has a competency that, at least at the current time, AI cannot easily replace. Again, you have to put things, again, horses for courses. That’s why I mentioned these three little pieces of items, which is we have to serve, see it’s nuanced. This is not a one-size-fits-all lay-off, mass, you’re done, thing. It really depends on the role. It depends on the seniority of the person. Depends on the competency in the area that they’re involved in.
Bertrand Schmitt
In terms of AI, there are really two situations. Either your core knowledge and expertise lies in AI per se, you are a data scientist, an AI engineer. The other side is that you are really good at using AI to do your usual work. Suddenly, if you are a software developer and use all the latest AI tools, and as a result, you are 2X, 3X more efficient than before, that’s definitely a level of expertise that companies will also be interested in.
I remember reading an interesting stat that people who best benefit from all the latest AI tools to do software development were actually the most expert developers who made the switch to using these tools some time ago. They were much better leveraging AI than younger developer, for instance, with little expertise. They will benefit from AI tools, but not as much as a pro developer. In a way, the gap is getting wider between entry-level people might have started to use AI and true experts who are also experts at using AI tool. It’s a gap that has been increasing.
I wanted to share two stats I read from the latest COA 2-2025 deck. They shared during their conference. One, they were showing an interesting analysis about Microsoft employee count, where basically between in 2022-2023, the number of employees stabilized. At Microsoft, it went stable at around 230,000 people.
Then we had a peak, a strange peak around July 2024 at around 240K, 250K employees. Since then, so basically since the past 12 months, the number of employees has constantly been going down. One of the questions would be, as Microsoft, which pick employee? Microsoft, by the way, show a stat that 30% of all their code now is generated by AI, 30%. That’s pretty impressive.
Another company, a much smaller one, but quite successful one, AppLovin. Another interesting stat is that in the past 4 years, they have doubled their revenue per employee. Their revenue per employee started from 3.6 million per employee. It was not starting from a low base easy to increase. It was starting from an insane number, very high number. They managed to double that to 7.6 million revenue per employee. During that period, they actually decreased their number of employees. They went up to 1,000 people and went down as of today, to maybe 750 people.
It’s very impressive to see even smaller companies, less than a thousand employee, has managed to increase so much their revenue per employee and took the decision to decrease actively their number of employees because they could have said, “You know what?” We need to invest in more stuff, and we need to do more stuff. Let’s be careful. Let’s keep increase employee size. But no, they decided that they can do more and keep increasing the business while decreasing employees. Even a world-leading revenue per employee.
Nuno Goncalves Pedro
It’s a two-tale side at play. One is that there’s a lot more code being actually generated by AI and tools. Garry Tan, the CEO of YC, has said that at least a quarter of their companies, more than 95% their code was generated by AI, which is silly. More than 94% of the code was generated by AI.
Then the other side is then you have these rapid growth micro companies that have very few employees as startups and are scaling through the roof. We’ll come back to that a little bit later because that changes a lot of the dynamics around employment in engineering, R&D, development, et cetera. But we’re now seeing a lot of these companies that are growing super fast with very few employees. Maybe that’s an interesting segue. Then what happens to salaries? Our salary is going to go through the roof for everyone involved. Not really.
As you said, actually, there’s a lot of functions where salary growth has actually flattened. In many cases, there’s now some salary decreases around certain areas. It’s again going to be a have and have nots. If you have a certain capacity and skill set, you probably can make more and more money as an individual. But actually, across the board, what used to be the core rules, the rules that carry the organization, for example, in terms of development, it might be that there is a salary flattening over time.
Bertrand Schmitt
There is that question in Silicon Valley right now about when will we see the first one employee unicorn? What’s your take on this one? When will we see that?
Nuno Goncalves Pedro
I think sometime soon. I’m looking forward to when we’ll see a zero employee unicorn. An agent will create some…
Bertrand Schmitt
Zero employee, technically.
Nuno Goncalves Pedro
A zero employee. I’m an agent. I created my own self and I created a company. I’m not sharing it with anyone, and I’m a billionaire, and I don’t need VC money. That’s going to be cool. I don’t know. I think we’re probably going to reach it at some point.
There are some of these numbers, again, just to take this with a grain of salt, because a lot of these companies that seem to be growing at the speed of light, we’ve already talked about it in previous episodes, but a lot of these companies that are growing at the speed of light are getting subsidized to grow at the speed of light. They’re getting a lot of money, for example, to bolster their infrastructure. We already discussed it today.
The fact that you might be very lean in terms of your employee count might still mean that you have a lot of costs that you’re incurring in terms of your cloud processing, compute, storage, et cetera. It’s not like you don’t need capital. You might need capital, actually, you might need a lot more capital, but basically just your team is much smaller.
We have to look at some of these numbers with a grain of salt because there’s subsidization going into this. We’ve discussed it also in the past, there’s subsidization going into to go to market. There’s a lot of land grab playbooks out there where companies raise gigantic rounds. Then they might not have a ton of employees, but at a certain point in time, they are spending a ton of money to get to wherever they are in terms of, for example, annualized recurring revenue as a metric.
Again, I would be a little bit paused on… It will happen, I’m sure, the one employee unicorn, or the one-person unicorn, founder unicorn, single founder, only person on the team unicorn. But we’ll need to look at it with a little bit of a fresh lens at that point in time.
Bertrand Schmitt
As you say, if you have big needs in term of CapEx, for instance, because you are building your own training clusters, for instance, for sure, you might spend a lot on that side. As a result, yes, you might have less employees, but does it still mean you are profitable? Maybe not.
You are right that another part could be the go-to-market spend. It could be enormous, especially at times when everybody is going all in in the eye. How do you differentiate yourself? How do you make yourself visible in the middle of all this noise? It might be very hard. It might be very hard. What do you do? You go out and do some marketing spend to correct for this.
It’s not because you have smaller teams that it means that you are profitable. But my bet would be that at least if you are more focused on the software side of things, not investing too much in infrastructure, you probably are in a better spot than before in term of capacity to be profitable. But at the same time, you are right in the middle or in the beginning of a big race, so not investing enough could be a dangerous game.
Nuno Goncalves Pedro
Everything changes. We started talking about salaries. It changes with equity grants, bonuses in general. We had the parentheses around some of these AI roles that seem to be super overpaid at this point in time. Like a superstar rock star signing, a little bit like a football player thing. But in general, there’s a little bit of a stepping back on bonuses, a stepping back on equity grants.
The perks, there was a step back. I’m not sure that’s true anymore with the perks. There’s now a push to get people back into the office in many of the large companies. I think the concern there was a concern around culture, but also a concern around people hiding and not doing anything, of which we knew there were a mass amount of a significant percentage of people hiding and not doing much in terms of work, in particular around some of the engineering disciplines.
I think this perk thing is probably a tale of two cities kind of thing. There’s a reduction in perks in many companies, but at the same time, there’s maybe a comeback of perks in other companies so that people do go to the office. But in general, perks, I don’t know. My perspective is that tech will still be ahead of the game in terms of perks it gives to his employees, in particular if we’re talking about Silicon Valley and some of the big tech companies in the US, Silicon Valley, Seattle, et cetera. But what’s your read, Bertrand?
Bertrand Schmitt
My guess is that I’m not sure perks have been going back in a big way in general on average. But what is true is that the past 2 years, there have been a lot of forcing back on place to come back to the office. In some ways, that has been used as a way to get rid of people who are not willing to relocate back closer to headquarters, not willing to go back working at the office. In a way, it was a soft lay-off. In many situations, there was also no compensation. Basically, if you are not coming back to the office, you are resigning. It’s not that you are getting fired. That was also a way to do it on the cheap, even from bigger companies.
Personally, I understand that. It’s logical to say that if you’re not willing to come back, I mean, you are resigning, basically. But I would say my exception would be depending on what was the company policy, because if you got hired with a specific reason that, “Hey, you can work from remote, and you can work wherever you want.” I mean, it’s a change policy. If there is a policy change, I don’t know if it’s totally super fair to play that game in such a way. I think it really depends on a case-by-case basis.
Nuno Goncalves Pedro
We’ll see how that evolves if big tech will take away a lot of its perks, and they’ll go back to normal. If all of this is happening, what’s the positive end of the spectrum? Is there something emerging that is positive in terms of growth, in terms of different employment, and around big tech in general? We’ve already mentioned AI and some of the AI skills, people that have deep knowledge that have come, for example, from the research side, that have come from the applied maths side of the fence, which is pretty vital for AI at this point in time.
But there are other pieces that are really rising, and there’s a rise of different roles and a different working. Many would call it the fractional roles, freelance roles, solopreneur work. There, we already started seeing it a long time ago. Startups in particular needed fractional talent. Very early on, for example, the classic one is the CFO, a fractional CFO. “I’m a startup. I don’t need a full-time CFO for a long, long time. But I do need someone to keep up my books and my accounting and for me to bounce ideas of and have all the accounts in order, everything like that.”
Now we start seeing fractional CTOs and fractional Chief Marketing Officers and fractional chief revenue officers, et cetera, et cetera. All of this is emerging not only because of lay-offs, but because in particular, if you’re talking about a startup, the half life of some of these roles is pretty limited early on. You don’t need someone, first of all, to be full-time.
Secondly, you need very specific things and very specific help. You could say, “Consulting could be an obvious thing for this.” But consulting is not cheap in general because there’s overhead and there’s a bunch of other things that are being loaded into it. It’s easier to have either a fractional firm that’s a little bit like a body shopping entity or individuals that do this with part of their time.
In some ways, I think what’s cool right now with this emergence of significant lay-offs, which is obviously bad news, is a lot of this talent is in the market. They’re like, “You know what? I could actually make myself available to a variety of organisations at the same time instead of just working for one. Honestly, I could make more money working for five or six organisations.”
I have a very good friend of mine, and he started doing consulting after his last startup. At a certain point in time, he had tons of people working for him, and he built a firm, and he’s been doing it for 5 years. It’s a deep functional expertise that he had and that he built in his firm. Then at the same time, there was a very deep segment and audience that he was catering to in companies. So that can actually work for significant amounts of time. I call that a little bit the podification of work. It might start with an individual doing this as a one-off. But over time, that individual starts building a team and starts building infrastructure and starts building around a couple of things.
Again, similar to what we’ve seen, for example, with fractional CFOs, with outsourcing companies for IT and for development. We’re now seeing it across a variety of functions. I think that’s actually super exciting.
Bertrand Schmitt
Yeah, I think it’s a good thing. I think some individuals will find success in that. But given the increase in this function, of fractional CFO, CTO, CMO, I would say I’m mostly doubtful that it’s more than a temporary situation for a lot of people. I don’t see how the market can adjust to such an increase. I think the trend was up 70X on LinkedIn, so it doesn’t sound something that will still exist forever.
I think some will definitely adjust. As you say, I also know some people, some fractional CMOs who have built their own team, it’s half a dozen people. I think it’s definitely a good trend and something that can be helpful for a lot of startups because when you have to start building a new team, a new marketing team, a new financial team, that can take quite a while to have the right people in the team. In the meantime, outsourcing to people who have real expertise that should also help you build the right team internally step by step. Personally, I think it’s a great idea as a startup to leverage some of these resources.
Nuno Goncalves Pedro
Yeah, maybe new companies will emerge out of this. The question is really, is this a temporary piece or not I’m a little bit more on the other side. That’s why I think there are elements in terms of the stack of functional knowledge that are not temporary. Will most of these people make it? Of course not. I mean, most of the people will fail miserably. Out of this, some companies will emerge, some new companies that will be better at catering to the beginning of the life cycle of a company, so the startup phase. Others will just stay as podification. My friend’s company is a good example of that. They just stayed within a certain size, and they did very well. I’m not sure his objective is to be the next mini Accenture for whatever space he’s in.
I feel there’s a little bit of staying power and people want that, I think, freedom. They want freedom, they want autonomy, they want quality of work. They want to be able to select the projects and the teams that they want to work with and the ones that they don’t.
I think one other effect is people like you and me. We have day jobs, we do stuff. I run a VC firm, you’re in a VC firm. But then we have side things that we do. It’s natural that people have side things. Obviously, our friend, Elon Musk, is probably the one that has more side things than everyone else. He has several companies as side things. But in general, it also means that people like you and I, when we have domain knowledge on the side that we can apply to other domains that are not necessarily the domain we’re spending our time in.
For example, I don’t work extensively with corporations right now. I work with startups, and I have investors. I have limited partners that I need to cater to. But I have a bunch of knowledge around digital transformation, innovation, scale up of companies, et cetera, et cetera. How do I share that? Starting to have platforms in the market, for example, Intro is now making a bit of a splash in the market. Having some of these platforms in the market that allow you to actually be procured by other people that need your knowledge is very powerful. Some of these platforms have existed for a long time, the GLGs of the world and all of that.
But I feel there’s a new momentum here emerging around deep expertise, subject matter expertise, knowledge, and how you scale that knowledge and how you put it into the market. This is still the pre-AI phase, because once AI can create replicas of us and avatars of us, that’s going to be cool, right? I mean, it’s like…
Bertrand Schmitt
I mean, that means we rent our AI by the hour.
Nuno Goncalves Pedro
Let’s get a Nuno avatar to do things, and I can just go and chill on the beach.
Bertrand Schmitt
We can do, yes, 200, 300 hours a week, thanks to our virtual AI. So, that would be the right way to do it. That would be interesting. I start to see this happening, actually. People building chatbots based on what they have written, based on what they have been talking on podcast, building a chatbot based on that and providing access to this chatbot for a fee. It’s actually a very interesting development.
That’s also, for me, an interesting proof that it’s good that you put in writing or in audio what you are doing. Then you can build this type of AI tools. If you are just doing your job, but your sharing is not really public or in a form that is usable for a chatbot, you won’t be able to virtualise this knowledge. I think that’s for me an interesting development. You could argue writing a book is also something good from that perspective. Even if we know that writing a book is not in itself something that will bring you much money, indirectly, in addition to being a way to advertise yourself, it can be a way to start your virtual self from.
Nuno Goncalves Pedro
Maybe moving to other positive effects of this mass transition in terms of workforce that’s happening in technology, is how is this talent spilling over to other areas? I mean, is it that everyone’s getting laid off in tech, and they’re without a job, and they’re unemployed, or is it that they’re finding jobs elsewhere? Our perspective, at least from what it looks, is in the market, there are some interesting migrations happening. Some of them are the adjacencies. I mean, if you look at things like healthcare, biotech, you could say, “Well, that’s tech as well.” Well, probably most people would disagree. It’s a different tech at the very, very least.
There’s obviously a lot of interesting things happening in the deep and frontier tech side of this fence, but that’s still tech at the end of the day. But a lot of the talent is really spilling over to other areas. They’re going now potentially into companies in finance, financial services. There’s been always this tension movement between FSI, financial service industry, and tech, where FSI typically paid well, at least in certain kinds of roles. That transition can happen during these times. Manufacturing, logistics, government, in some cases, we won’t go. I won’t open the Pandora’s box on that one. We’ll just stay at that level.
But obviously, there’s a lot of talent that’s now migrating to industries, defence, for example, where it might have classically not gone to. There’s a lot of tech talent, for example, going into defence. They might end up in defence tech themselves, but defence in and of itself, in terms of the tech talent that was there, was a different animal, so to speak, as an industry. Now we’re seeing a lot of these spillovers, a lot of these movements of talent, which, in my opinion, are actually very positive. Let’s bring in the love, let’s bring in the expertise of technology to other domains that might actually benefit tremendously from it.
Bertrand Schmitt
Yeah, I totally agree with you. In some ways, it’s funny because it’s reminding me what I saw 10 years ago in mobile, where a lot of people were early in mobile apps. At the beginning, early on in the 2010-2015, mobile apps were really mostly about gaming. But step by step, many people were not so keen to stay in the gaming industry. I saw many moving and switching to different industries who were very keen to have people knowledgeable in app business, app development.
I remember seeing many people jumping to the financial industry, for instance, when banks were building their mobile apps and the like. It was for me very interesting to see that switch. These people were highly sought after by these industries because they were known to have dealt with stuff very early on that were not so easy and that were important and could apply to their own business.
Nuno Goncalves Pedro
I do predict there’s even a more interesting spillover effect that’s probably up and coming as well, which is even founders of startups and some of us investors that might say, “You know what? I just want to go and do something else in another industry.” We discussed it in the past, what areas are less impacted by AI. There’s many industries that need so much revamping that even the effects of AI will take years, even decades to be more heartfelt.
The level of core innovation that needs to happen and the underpinning of that innovation, again, some of this talent, founders, investors, they’re more than capable to bring it to other industries. We’re not saying that we’re white knights that are going to save industries. That’s not the point. It’s like the technology toolkit, the understanding that tech skills very, very quickly or quicker than humans is something that’s valuable, we believe, for a lot of industries.
Bertrand Schmitt
One thing, though, to keep in mind because sometimes I see some companies with things they can do as well as Meta or Google or others in terms of developing their own tools or product based on AI. Sometimes I’m just shocked. I’m like, “Guys, come on, let’s be realistic 2 minutes. You don’t have the capacity to pay some of the top talents like Silicon Valley can pay.” I’m talking about cash, signing bonus, salaries, stock options that we increase in value.
I would say I would just highlight that piece of the puzzle because I’ve seen again and again some industries that believe that, “Hey, I’ve hired a few guys relatively cheap from the tech industry, and now I don’t need to buy the products from the Google of the world. I can do it myself.” I mean, good luck with that. Good luck with that. You are not going to be able to compete. Find a way to use this talent in a way that is efficient and smart, but don’t think that give you the ability to compete with the big boys, because you won’t be able to compete.
Nuno Goncalves Pedro
The operating models, to your point, are fundamentally different. If you just hire a few people and say, “Well, now go and do it.” It’s like, well, you’re fitting into an organisation that has specific APIs in mind, a specific value system, how things are done, how the cycles are done, you need to cater not just to individuals, but really create organisations that really are able to scale technology and actually shift how the organisation would work.
If you want a more tech-enabled organisation, you need to create an operating model that is tech-enabled, not just come here and fit in. It’s like you can’t make much work. In some ways, fully agreed, just hiring people won’t solve the problem. Making big bets into pools of talent, into new operating models, into new processes, even into new spin-out organisations is going to make a much bigger impact.
Bertrand Schmitt
Yes. Again, at some point, it’s the very best. We stay in tech where they can get pretty insane salary compensation that no normal company will be able to afford. You won’t be able to fight meta who wants to hire the top AI engineer or top AI scientist and spending dozens of millions on some of the top talents. I mean, there’s no way you can compete there. You have to be realistic about the level of talent you can bring and, of course, making your organisation compatible with this talent. If you don’t change your ways, there is not much anyone can do. But even if you change your way, even if you do a lot of things positively, if you are not able to compete on the comp side, you have to have reasonable expectations as well.
Nuno Goncalves Pedro
Let’s talk maybe about some of the cultural and geographical shifts that we see happening. We’ve talked about many of them because of COVID and during COVID and post-COVID, the logic and what was happening.
A very obvious one is the remote versus return to work dynamic as two extremes: fully remote or full return to work. In some ways, big tech seems to be going back to the return to work, so the whole, you guys can be wherever you are, and you get paid. That has come crashing and burning, so to speak, very visibly and very publicly. A lot of the big tech companies saying, “You need to return to office, and you need to return to an office that you belong to. Otherwise, either we change your salary because you’re somewhere else. If you have an office for you to go into in the area that you’re in now, or you should leave.”
In some ways, there’s a momentum back. I think at least I have been a proponent that the world, maybe with everything that happened with COVID, would go more towards a hybrid model where you have hubs, where people work from the office, again, a lot more, but they get to work a bit from home.
In some cases, we’re seeing a really like, no, you return to the office, maybe you can work from a one day a week, and that’s it. There’s a little bit of a rebalancing, I think, of workforces. Obviously, smaller companies, startups still very much have very distributed teams, in some cases, almost fully liquid still and all remote. But if we’re talking about mid-size, high-growth companies, large big tech companies, it seems like we’re going full on back to the good old days. We’ll see how that pans out.
Bertrand Schmitt
Yeah, I will say on this topic, one thing I have seen and has not changed is that if you are some of the top talents around there on your topic, companies will let you do whatever you want. I know some people who are AI scientists, they are still working from home, and trust me, they are not going to get fired because the companies know that they have a good deal. The guy is there, and he can pick any job he wants if he’s out. I think it’s true of your average worker, but top talent can still impose their rules to most systems.
Nuno Goncalves Pedro
Let me just comment on that. I think it is true, but just for those listening, we’re talking about really top-tier talent. I was just privy to a transaction involving two parties, a larger party and a smaller startup, where the transaction didn’t go through because they were doing an acquisition of a company, the larger party, just for AI talent, and they recognised the smaller company had amazing AI talent, and still they were like, “No, you guys are going to have to come to these two hubs, and there’s no flexibility. We don’t give a shit.”
Again, we’re talking about really high top-tier talent, rock star talent. We’re not talking about the best performers. We’re talking about the the the best performers is what I would say. Then you’re allowed to do whatever you want. Then you’re prima donna, that’s fine.
Bertrand Schmitt
Yes, and I’m sure different companies they have different approach, but some people I have in mind are working for companies that are very well known to not have remote work policies, but they still have for some people. The other piece of the puzzle that some things I’ve heard is some companies that are mandating a return to the office, but they don’t have enough seats.
Nuno Goncalves Pedro
Yes, I’ve heard that.
Bertrand Schmitt
For me, it’s just so unbelievable that you’re mandating employees to come back to the office, but you don’t have the number of seats. People basically don’t have a desk when they arrive, don’t have a parking spot, practically cannot work from the office. You have 20, 30, 40% not enough capacity in the office. For me, just this stuff blows my mind because I have no way. I simply cannot understand how an HR team, how some execs have no clue to such a level that they mandate one thing on one side, but on the other hand, they don’t have the capacity. They do both at the same time. We mandate, and then we don’t have the capacity.
I’m not sure what to take about that. It really made so little sense to me that it’s shocking. I’m talking about really big companies that have the ability to have the visibility and to rent whatever needs to be rented. It’s shocking. For me, it’s interesting. You have what he said, what he talk about. When you see the practical reality, it’s quite shocking to see the, I would say quite frankly, incompetence of some in terms of how the organisers return to the office.
Nuno Goncalves Pedro
Definitely a lot of incompetence. Fundamental shift, maybe switching gears a little bit is, how is work being done? Obviously, we’ve talked a lot about demographics, and now Gen Zers are fundamentally different and looking for authenticity, mission-driven work, having side gigs. They have their work-life balance as a given, and all these elements that are being pushed onto. We’ve also discussed the social issues. Are they ready socially to manage and scale to be managers, et cetera. But I think there’s a more fundamental shift happening right now, in particular, if I look at tech, which is this notion of individual contributor versus manager.
I think we went through a time where there was a lot of more and more management layers. There’s now this discussion where the structure should be flatter. There should be still managers because there needs to be orchestration, but the structures need to be flatter because there’s a lot of individual contributors at the table.
Bertrand Schmitt
It’s more efficient. Too many layers destroy your capacity.
Nuno Goncalves Pedro
It’s always been more in computer engineering. This notion of small teams, highly empowered, highly capable to deliver with a thin layer of management is always, I think, what people have talked about in computer engineering as the best case scenario. I feel in some ways we’re going back to that a little bit. We are going back to the age of individual contribution, and particularly with this AI augmentation, the ability for me to do stuff myself.
I was talking to a founder the other day, and he’s an older founder, a little bit more experienced in the market. He was telling me, “I’m not sure I need a ton of junior developers. I can just go and look for stuff.” He’s the CTO, in this case, of the company. “There’s a lot of stuff I can do myself. I could just look for code, clean it up, debug it, compile it, run it. That’s fine. For basic stuff, I don’t need people.” In some ways, we’re like, this is the max manager, the CTO. In some cases, the CTO is saying, “I can be a contributor myself, and then I don’t need to be managing three or four people. I can do basic jobs myself.”
I feel there’s a little bit of flattening of engineering orgs going on. It might at some point go back to a lot of management layers, but that it feels, at least to me right now, that’s where we are in terms of the cycle.
Bertrand Schmitt
Yes. An interesting thing is that it goes in parallel with that new trend of AI as a manager. Not only you might have less manager for efficiency’s sake, a lot of companies realise they cannot have too many layers of management without losing their capacity to do stuff quickly, to turn around quickly. But on top of it, there is that new trend of management through AI. A lot of people say AI might be able to replace middle management. It’s also an interesting development. It might not just be about AI on the front lines of replacing the most junior jobs. There is a lot of talks about AI replacing managers.
Nuno Goncalves Pedro
Do you want to close down?
Bertrand Schmitt
Yeah, sure. Some takeaways on the topic. Definitely, there is a big shift in tech in hiring. If you want to be focused as a founder, you want to, of course, be cost-effective in hiring top talent, you should probably think about leveraging more fractional roles, talents that are not easy to hire, you cannot afford to hire full-time. You want to stay in. AI keeps improving, if not by the day, by the week, so be careful.
Stuff that you don’t believe could be done in 6 months might be doable with AI. If you are an operator, focus on a retention of your team, focus on best practices, remote, hybrid. I think that what some of the big companies are doing could be your benefit as a smaller company. Don’t think about applying the policies of the big guys because applying a different policies could be a way to get a specific advantage.
If you’re investors, I think that you probably want to think carefully about how this company is using AI, how this company is optimising resources. Are they still living like we used to do years ago? Are they full on in that brave new world, and I think it’s important to push for that.
Of course, some of this is painful when people are fired from their job, are laid off. But on the positive side, that’s usually a way to start again, to start new, to think about how you could do things differently, potentially start a new type of business as we discuss. I think everyone should think very carefully to upgrade their skills in order to better use AI, because that’s definitely the way things are going. Thank you, Nuno.
Nuno Goncalves Pedro
Thank you, Bertrand.
View all episodes
By Bertrand Schmitt & Nuno G. Pedro
5
2828 ratings
The reckoning is here. Once a safe harbor, Big Tech has finally also gone full out on layoffs. Is this a structural shift to employment in Tech? Will the subsequent talent spillover be great for start-ups and entrepreneurship? Will it positively affect other industries?
In this episode of Tech Deciphered, we will answer these and other questions in a deep discussion on the Global Tech Labor reset.
Navigation:
Our co-hosts:
Our show: Tech DECIPHERED brings you the Entrepreneur and Investor views on Big Tech, VC and Start-up news, opinion pieces and research. We decipher their meaning, and add inside knowledge and context. Being nerds, we also discuss the latest gadgets and pop culture news
Subscribe To Our Podcast
Nuno Goncalves Pedro
Welcome to Episode 66 of Tech DECIPHERED. Today, we’ll talk about the global labour tech reset. Tech and big tech, which seemed immune to any lay-offs, seems now to be under fire. Massive lay-offs over the last 2.5 years, a lot of discussion around the importance of having a computer science, computer engineering background, and so what seemed to be a safe haven for any graduate is now under stress.
Today, we will discuss the structural perspective on what’s happening in the market, if this is a long-term trend or not, what has led us to this, and what is next. We’ll talk about the rise of fractional freelance and solopreneur work, as well as talk about talent spillovers, and some of the usual geographical dynamics around the space. Bertrand, a huge shift in tech.
Bertrand Schmitt
Yes, definitely. It’s pretty big. I think it started probably around 2022, once we got some changing interest rates that have a pretty massive impact in stock prices for a lot of companies. At that time, a lot of companies decided, and usually under some pressure, that it was time to be more efficient, to generate more cash. Yes, you want to grow, but not grow at all cost. You have to go efficiently.
That’s when we started to see some share price going down and step by step, quarter after quarter. Some change in attitude with a lot of big tech and that has created some impact in term of lay-off from different parts of the business, from the sales team to the DNA, to even engineering R&D.
What is also been happening since 2022, 2023 is a change of focus. A lot of focus is being put in AI. A lot of investment in CapEx is going to AI. At some point, if you want to keep doing all this investing, investments, you might have to get some other part of the business in order to create additional savings to do all the spend you can in AI. There has been more recently a switch. It’s not about just efficiency to push all the… But generating the ability to invest in AI.
Nuno Goncalves Pedro
It’s part of a broader movement. Before we step back a little bit and go back in history, even recently, we’ve heard that there’s talks between Meta and a bunch of private equity firms like KKR, Brookfield, Apollo, and others, to actually help in financing data centers. Meta is a gigantic company, so one would assume they have cash to do all these things. Maybe they don’t.
To your point, that level of efficiency that is now needed in the market where you need to throw actual money, CapEx, into the building of infrastructure, the building of the core underpins of what you’re doing is pretty vital. But let’s go back a little bit of a second, and we’ve talked about it maybe in our early episodes.
Companies like Meta, Facebook back in the day, Google, Alphabet now, and others had a tendency, in particular in the Bay Area, to hoard engineers. They would over-hire. There was a lot of discussion that some of these players had as much as 20%-50% the engineering resources they needed, and part of that is that they didn’t want their competitors to have them, and so in some ways, we went from hoarding too much capacity to the days of COVID that in some ways illustrated that there were a lot of people that maybe were not working very hard also on the engineering side, not just on the sales, et cetera side.
That right-sizing, I think, has been a long time coming. We discussed in a past episode that there was a little bit of an adaptation, or an adaptation, rather to the market by a lot of these public companies saying, “Now that I have a mandate where there’s lay-offs happening in other industries, I can do it myself.” But this time is different, and I think the message of probably today’s episode is going to be around that. This time is different.
There are some fundamentals that have changed. Bertrand has already alluded to AI as one fundamental shift in terms of the efficiency of engineers, in particular more junior engineers. There is more behind this. There is the notion that you need less to deliver more. In tech, I would start with a platitude. I think if anything, everyone talks about AI and how AI is going to disrupt financial services, and it’s disrupting functions like marketing, and displacing people around, for example, customer service, and all of this. But a lot of people forget to talk about the most obvious industry that’s being disrupted by AI, and that industry is tech.
Tech is disrupting itself with AI because you have all these tools available, you have all these platforms that you’re deploying, and they have value for your own purposes. They become the underpinning of a more efficient workforce, in particular if you’re a tech company. A lot of people don’t talk much about that, but I think a lot of what we will talk today around lay-offs and efficiencies and all of that actually do relate to the fundamental shifts we’re seeing where the emergence with a lot of these AI tools and platforms that we’re seeing, so tech is disrupting itself at the end of the day.
Bertrand Schmitt
I think AI has two effects. One is it’s so big, it’s so important, you cannot miss it as a business that you are going to reallocate resources. From your traditional product lines, your traditional development, expansion, creation of new features, and you will reallocate instead more money to AI. It can be one side, can be just CapEx, and the other side will be, of course, software that you want to develop connected to AI.
I think that was the first effects that started early on. What it means is that many companies ultimately decide to fire people who are not so AI knowledgeable and instead hire people who have more expertise in AI. I think what you are talking is a second effect of AI. Now we start to reap the rewards of the investment in AI. What we are seeing as a result is that we can do more with less.
Basically, if you’re a software company, or a tech company, you have a lot in software engineering. If you have a lot in software engineering, guess what? One of the best way to leverage AI is actually to use it to develop more efficiently, faster. The tools have gotten better and better, what we got 2 years ago was barely usable, 1 year ago start to be good. Now, I must say, it’s shockingly impressive what you get if you have tool like Visual Studio or if you use some other tools on the market that are even more focused on AI.
It’s quite shocking the quality of the code auto completions that you can get. It’s quite shocking what you can get from other tools directly like ChatGPT. It’s definitely like having a buddy that can support you, check your code, provide you some IDS, run code, help you go faster through documentation, testing. You are going to save time. The first effect might be, you know what? We can do so much more with the same workforce. Maybe we can reduce slightly the workforce, and we’ll still do more thanks to AI.
Of course, there might be a question of you might want to keep the same size of software engineering so that they deliver even way more than before and your pace of development is truly accelerating. Companies are facing multiple options here. It looks like many are choosing to be more efficient.
Nuno Goncalves Pedro
Yeah, there’s just… Let’s talk numbers, right? In terms of lay-offs, 2024 alone, it seems like the number was around 250K-260K, globally, people were laid off.
Bertrand Schmitt
In tech.
Nuno Goncalves Pedro
In tech, yeah. This is global tech lay-offs, right? In tech, 250K, 260K tech lay-offs in 2024. In the US, it varies. We have numbers that go from 95,000-150,000 tech jobs in the US alone that were laid off in 2024. We’ve had numbers that over a period of 2.5 years, 400,000 people in tech were laid off globally. These are massive numbers. This is an industry that by itself was already, in some ways, part of some efficiencies. It’s not the most inefficient industry in the world in some ways, so very significant numbers that we’re talking about.
Bertrand Schmitt
It’s not just lay-off, it’s also who you are not hiring anymore. I think we found an interesting stats and the quantity of entry-level tech jobs have dropped by 50% since 2019. Before the pandemic, new graduates made up 15%, of big tech hires. Today, it’s just 7%. We’re moving from 15% to 7%. It’s pretty shocking.
Where is it coming from? We talk about general cost-cutting measures post-COVID, but for sure, there is a lot around AI and automation, and especially since that AI is replacing a lot of more entry-level jobs. It can be QA, it can be IT support, it can be some software development. It’s definitely a big change. One thing not to forget is that computer science used to be a much smaller field 10, 20, 30 years ago. It has expanded a lot.
To be frank, I don’t think the quality of the graduates has stayed high. The more you expand the pool of people working in industries, the less you have the best of the best. As a result, you could argue there is a lot of people who are not either so good or so really interested in computer science, who still end up in computer science because that was the thing other people were doing. That was a safe way to get a job after your studies. But it was not because they are incredibly excited and knowledgeable about the space. It’s not because they are excellent developers.
My guess is that when everything is great, everything is easy, that’s good. But when there is tension in term of hiring, then it’s going to be trouble for a lot of people who don’t have the interest, appetite, and quite frankly, sometimes the level of what it takes to be a software engineer, for instance, in that new age. If you are not an expert with AI tools, if you don’t know AI deeply, why would a company hire you?
Nuno Goncalves Pedro
I think there are three important things to highlight here. The first one, and we’ve talked about it before, is, and this might be surprising to some people who are not computer engineers or computer scientists in background that might be listening to us, but there used to be the mythical figure of the 10X developer. Being a full-time equivalent can do the job of 10 good engineers.
Those developers do exist. Shockingly enough, they do exist. I’ve met a few in my career. I’ve been lucky to have a couple of them working with us at Chamaeleon. But amazing people that can do really the job of 10 people. I think what’s happening in the industry, to your point, Bertrand, is there’s a lot more variants in the quality. You have a lot of people that are not great. If you’re not a great developer, the problem is you’re actually probably having a negative impact in the projects that are involved in and slowing them down.
I think the second factor, which is really important to highlight, is some of these roles, and not to basically downgrade what the roles are, they’re very important roles, but some of the roles around DevOps and dev tools, around QA become specialists roles, but they’re roles that are a little bit less engineering-heavy, and they actually match some of the junior developer roles, which are what I’d call breaking stone issues.
You’re trying to do hardcore integrations, convert code, clean up code, et cetera. But at the end of the day, it’s not a fundamental engineering problem that you’re trying to solve for. AI, as we know, is extremely well-equipped to solve issues around that. It’s the issues that are very repeatable, that through testing can have relatively low impact in some ways at the end of the day in terms of the code base that you’re actually maintaining.
Then the third piece, I think, is some people might say, Well, but then is computer science that and computer engineering is that? What matters? No, it’s not. Actually, there are roles in computer science and computer engineering that are becoming even more and more vital. People that have very strong architectural view of code base, of what they’re developing, people that have full on, full stack knowledge of the world, people that have very deep knowledge around legacy systems, for example, embedded system, coding, et cetera. All of those right now don’t seem to be as under threat as other types of roles.
In some ways, we even have a more significant part of the haves and have nots, which you, Bertrand, very alluded to, which is people that have very deep domain knowledge around artificial intelligence. We’re now seeing an arms race for that talent, which the joke is going on that obviously Matt is trying to hire people and paying $100 million in our issues or whatever that is. I’m not sure how much of that is true.
Bertrand Schmitt
Just bonus.
Nuno Goncalves Pedro
Just bonus. I’m not sure how much of that is actually true or not and what the rules are for, but there’s this arms race for that talent. That talent is even more valuable because, again, back to the point, it’s probably seen as more scarce right now. It has a competency that, at least at the current time, AI cannot easily replace. Again, you have to put things, again, horses for courses. That’s why I mentioned these three little pieces of items, which is we have to serve, see it’s nuanced. This is not a one-size-fits-all lay-off, mass, you’re done, thing. It really depends on the role. It depends on the seniority of the person. Depends on the competency in the area that they’re involved in.
Bertrand Schmitt
In terms of AI, there are really two situations. Either your core knowledge and expertise lies in AI per se, you are a data scientist, an AI engineer. The other side is that you are really good at using AI to do your usual work. Suddenly, if you are a software developer and use all the latest AI tools, and as a result, you are 2X, 3X more efficient than before, that’s definitely a level of expertise that companies will also be interested in.
I remember reading an interesting stat that people who best benefit from all the latest AI tools to do software development were actually the most expert developers who made the switch to using these tools some time ago. They were much better leveraging AI than younger developer, for instance, with little expertise. They will benefit from AI tools, but not as much as a pro developer. In a way, the gap is getting wider between entry-level people might have started to use AI and true experts who are also experts at using AI tool. It’s a gap that has been increasing.
I wanted to share two stats I read from the latest COA 2-2025 deck. They shared during their conference. One, they were showing an interesting analysis about Microsoft employee count, where basically between in 2022-2023, the number of employees stabilized. At Microsoft, it went stable at around 230,000 people.
Then we had a peak, a strange peak around July 2024 at around 240K, 250K employees. Since then, so basically since the past 12 months, the number of employees has constantly been going down. One of the questions would be, as Microsoft, which pick employee? Microsoft, by the way, show a stat that 30% of all their code now is generated by AI, 30%. That’s pretty impressive.
Another company, a much smaller one, but quite successful one, AppLovin. Another interesting stat is that in the past 4 years, they have doubled their revenue per employee. Their revenue per employee started from 3.6 million per employee. It was not starting from a low base easy to increase. It was starting from an insane number, very high number. They managed to double that to 7.6 million revenue per employee. During that period, they actually decreased their number of employees. They went up to 1,000 people and went down as of today, to maybe 750 people.
It’s very impressive to see even smaller companies, less than a thousand employee, has managed to increase so much their revenue per employee and took the decision to decrease actively their number of employees because they could have said, “You know what?” We need to invest in more stuff, and we need to do more stuff. Let’s be careful. Let’s keep increase employee size. But no, they decided that they can do more and keep increasing the business while decreasing employees. Even a world-leading revenue per employee.
Nuno Goncalves Pedro
It’s a two-tale side at play. One is that there’s a lot more code being actually generated by AI and tools. Garry Tan, the CEO of YC, has said that at least a quarter of their companies, more than 95% their code was generated by AI, which is silly. More than 94% of the code was generated by AI.
Then the other side is then you have these rapid growth micro companies that have very few employees as startups and are scaling through the roof. We’ll come back to that a little bit later because that changes a lot of the dynamics around employment in engineering, R&D, development, et cetera. But we’re now seeing a lot of these companies that are growing super fast with very few employees. Maybe that’s an interesting segue. Then what happens to salaries? Our salary is going to go through the roof for everyone involved. Not really.
As you said, actually, there’s a lot of functions where salary growth has actually flattened. In many cases, there’s now some salary decreases around certain areas. It’s again going to be a have and have nots. If you have a certain capacity and skill set, you probably can make more and more money as an individual. But actually, across the board, what used to be the core rules, the rules that carry the organization, for example, in terms of development, it might be that there is a salary flattening over time.
Bertrand Schmitt
There is that question in Silicon Valley right now about when will we see the first one employee unicorn? What’s your take on this one? When will we see that?
Nuno Goncalves Pedro
I think sometime soon. I’m looking forward to when we’ll see a zero employee unicorn. An agent will create some…
Bertrand Schmitt
Zero employee, technically.
Nuno Goncalves Pedro
A zero employee. I’m an agent. I created my own self and I created a company. I’m not sharing it with anyone, and I’m a billionaire, and I don’t need VC money. That’s going to be cool. I don’t know. I think we’re probably going to reach it at some point.
There are some of these numbers, again, just to take this with a grain of salt, because a lot of these companies that seem to be growing at the speed of light, we’ve already talked about it in previous episodes, but a lot of these companies that are growing at the speed of light are getting subsidized to grow at the speed of light. They’re getting a lot of money, for example, to bolster their infrastructure. We already discussed it today.
The fact that you might be very lean in terms of your employee count might still mean that you have a lot of costs that you’re incurring in terms of your cloud processing, compute, storage, et cetera. It’s not like you don’t need capital. You might need capital, actually, you might need a lot more capital, but basically just your team is much smaller.
We have to look at some of these numbers with a grain of salt because there’s subsidization going into this. We’ve discussed it also in the past, there’s subsidization going into to go to market. There’s a lot of land grab playbooks out there where companies raise gigantic rounds. Then they might not have a ton of employees, but at a certain point in time, they are spending a ton of money to get to wherever they are in terms of, for example, annualized recurring revenue as a metric.
Again, I would be a little bit paused on… It will happen, I’m sure, the one employee unicorn, or the one-person unicorn, founder unicorn, single founder, only person on the team unicorn. But we’ll need to look at it with a little bit of a fresh lens at that point in time.
Bertrand Schmitt
As you say, if you have big needs in term of CapEx, for instance, because you are building your own training clusters, for instance, for sure, you might spend a lot on that side. As a result, yes, you might have less employees, but does it still mean you are profitable? Maybe not.
You are right that another part could be the go-to-market spend. It could be enormous, especially at times when everybody is going all in in the eye. How do you differentiate yourself? How do you make yourself visible in the middle of all this noise? It might be very hard. It might be very hard. What do you do? You go out and do some marketing spend to correct for this.
It’s not because you have smaller teams that it means that you are profitable. But my bet would be that at least if you are more focused on the software side of things, not investing too much in infrastructure, you probably are in a better spot than before in term of capacity to be profitable. But at the same time, you are right in the middle or in the beginning of a big race, so not investing enough could be a dangerous game.
Nuno Goncalves Pedro
Everything changes. We started talking about salaries. It changes with equity grants, bonuses in general. We had the parentheses around some of these AI roles that seem to be super overpaid at this point in time. Like a superstar rock star signing, a little bit like a football player thing. But in general, there’s a little bit of a stepping back on bonuses, a stepping back on equity grants.
The perks, there was a step back. I’m not sure that’s true anymore with the perks. There’s now a push to get people back into the office in many of the large companies. I think the concern there was a concern around culture, but also a concern around people hiding and not doing anything, of which we knew there were a mass amount of a significant percentage of people hiding and not doing much in terms of work, in particular around some of the engineering disciplines.
I think this perk thing is probably a tale of two cities kind of thing. There’s a reduction in perks in many companies, but at the same time, there’s maybe a comeback of perks in other companies so that people do go to the office. But in general, perks, I don’t know. My perspective is that tech will still be ahead of the game in terms of perks it gives to his employees, in particular if we’re talking about Silicon Valley and some of the big tech companies in the US, Silicon Valley, Seattle, et cetera. But what’s your read, Bertrand?
Bertrand Schmitt
My guess is that I’m not sure perks have been going back in a big way in general on average. But what is true is that the past 2 years, there have been a lot of forcing back on place to come back to the office. In some ways, that has been used as a way to get rid of people who are not willing to relocate back closer to headquarters, not willing to go back working at the office. In a way, it was a soft lay-off. In many situations, there was also no compensation. Basically, if you are not coming back to the office, you are resigning. It’s not that you are getting fired. That was also a way to do it on the cheap, even from bigger companies.
Personally, I understand that. It’s logical to say that if you’re not willing to come back, I mean, you are resigning, basically. But I would say my exception would be depending on what was the company policy, because if you got hired with a specific reason that, “Hey, you can work from remote, and you can work wherever you want.” I mean, it’s a change policy. If there is a policy change, I don’t know if it’s totally super fair to play that game in such a way. I think it really depends on a case-by-case basis.
Nuno Goncalves Pedro
We’ll see how that evolves if big tech will take away a lot of its perks, and they’ll go back to normal. If all of this is happening, what’s the positive end of the spectrum? Is there something emerging that is positive in terms of growth, in terms of different employment, and around big tech in general? We’ve already mentioned AI and some of the AI skills, people that have deep knowledge that have come, for example, from the research side, that have come from the applied maths side of the fence, which is pretty vital for AI at this point in time.
But there are other pieces that are really rising, and there’s a rise of different roles and a different working. Many would call it the fractional roles, freelance roles, solopreneur work. There, we already started seeing it a long time ago. Startups in particular needed fractional talent. Very early on, for example, the classic one is the CFO, a fractional CFO. “I’m a startup. I don’t need a full-time CFO for a long, long time. But I do need someone to keep up my books and my accounting and for me to bounce ideas of and have all the accounts in order, everything like that.”
Now we start seeing fractional CTOs and fractional Chief Marketing Officers and fractional chief revenue officers, et cetera, et cetera. All of this is emerging not only because of lay-offs, but because in particular, if you’re talking about a startup, the half life of some of these roles is pretty limited early on. You don’t need someone, first of all, to be full-time.
Secondly, you need very specific things and very specific help. You could say, “Consulting could be an obvious thing for this.” But consulting is not cheap in general because there’s overhead and there’s a bunch of other things that are being loaded into it. It’s easier to have either a fractional firm that’s a little bit like a body shopping entity or individuals that do this with part of their time.
In some ways, I think what’s cool right now with this emergence of significant lay-offs, which is obviously bad news, is a lot of this talent is in the market. They’re like, “You know what? I could actually make myself available to a variety of organisations at the same time instead of just working for one. Honestly, I could make more money working for five or six organisations.”
I have a very good friend of mine, and he started doing consulting after his last startup. At a certain point in time, he had tons of people working for him, and he built a firm, and he’s been doing it for 5 years. It’s a deep functional expertise that he had and that he built in his firm. Then at the same time, there was a very deep segment and audience that he was catering to in companies. So that can actually work for significant amounts of time. I call that a little bit the podification of work. It might start with an individual doing this as a one-off. But over time, that individual starts building a team and starts building infrastructure and starts building around a couple of things.
Again, similar to what we’ve seen, for example, with fractional CFOs, with outsourcing companies for IT and for development. We’re now seeing it across a variety of functions. I think that’s actually super exciting.
Bertrand Schmitt
Yeah, I think it’s a good thing. I think some individuals will find success in that. But given the increase in this function, of fractional CFO, CTO, CMO, I would say I’m mostly doubtful that it’s more than a temporary situation for a lot of people. I don’t see how the market can adjust to such an increase. I think the trend was up 70X on LinkedIn, so it doesn’t sound something that will still exist forever.
I think some will definitely adjust. As you say, I also know some people, some fractional CMOs who have built their own team, it’s half a dozen people. I think it’s definitely a good trend and something that can be helpful for a lot of startups because when you have to start building a new team, a new marketing team, a new financial team, that can take quite a while to have the right people in the team. In the meantime, outsourcing to people who have real expertise that should also help you build the right team internally step by step. Personally, I think it’s a great idea as a startup to leverage some of these resources.
Nuno Goncalves Pedro
Yeah, maybe new companies will emerge out of this. The question is really, is this a temporary piece or not I’m a little bit more on the other side. That’s why I think there are elements in terms of the stack of functional knowledge that are not temporary. Will most of these people make it? Of course not. I mean, most of the people will fail miserably. Out of this, some companies will emerge, some new companies that will be better at catering to the beginning of the life cycle of a company, so the startup phase. Others will just stay as podification. My friend’s company is a good example of that. They just stayed within a certain size, and they did very well. I’m not sure his objective is to be the next mini Accenture for whatever space he’s in.
I feel there’s a little bit of staying power and people want that, I think, freedom. They want freedom, they want autonomy, they want quality of work. They want to be able to select the projects and the teams that they want to work with and the ones that they don’t.
I think one other effect is people like you and me. We have day jobs, we do stuff. I run a VC firm, you’re in a VC firm. But then we have side things that we do. It’s natural that people have side things. Obviously, our friend, Elon Musk, is probably the one that has more side things than everyone else. He has several companies as side things. But in general, it also means that people like you and I, when we have domain knowledge on the side that we can apply to other domains that are not necessarily the domain we’re spending our time in.
For example, I don’t work extensively with corporations right now. I work with startups, and I have investors. I have limited partners that I need to cater to. But I have a bunch of knowledge around digital transformation, innovation, scale up of companies, et cetera, et cetera. How do I share that? Starting to have platforms in the market, for example, Intro is now making a bit of a splash in the market. Having some of these platforms in the market that allow you to actually be procured by other people that need your knowledge is very powerful. Some of these platforms have existed for a long time, the GLGs of the world and all of that.
But I feel there’s a new momentum here emerging around deep expertise, subject matter expertise, knowledge, and how you scale that knowledge and how you put it into the market. This is still the pre-AI phase, because once AI can create replicas of us and avatars of us, that’s going to be cool, right? I mean, it’s like…
Bertrand Schmitt
I mean, that means we rent our AI by the hour.
Nuno Goncalves Pedro
Let’s get a Nuno avatar to do things, and I can just go and chill on the beach.
Bertrand Schmitt
We can do, yes, 200, 300 hours a week, thanks to our virtual AI. So, that would be the right way to do it. That would be interesting. I start to see this happening, actually. People building chatbots based on what they have written, based on what they have been talking on podcast, building a chatbot based on that and providing access to this chatbot for a fee. It’s actually a very interesting development.
That’s also, for me, an interesting proof that it’s good that you put in writing or in audio what you are doing. Then you can build this type of AI tools. If you are just doing your job, but your sharing is not really public or in a form that is usable for a chatbot, you won’t be able to virtualise this knowledge. I think that’s for me an interesting development. You could argue writing a book is also something good from that perspective. Even if we know that writing a book is not in itself something that will bring you much money, indirectly, in addition to being a way to advertise yourself, it can be a way to start your virtual self from.
Nuno Goncalves Pedro
Maybe moving to other positive effects of this mass transition in terms of workforce that’s happening in technology, is how is this talent spilling over to other areas? I mean, is it that everyone’s getting laid off in tech, and they’re without a job, and they’re unemployed, or is it that they’re finding jobs elsewhere? Our perspective, at least from what it looks, is in the market, there are some interesting migrations happening. Some of them are the adjacencies. I mean, if you look at things like healthcare, biotech, you could say, “Well, that’s tech as well.” Well, probably most people would disagree. It’s a different tech at the very, very least.
There’s obviously a lot of interesting things happening in the deep and frontier tech side of this fence, but that’s still tech at the end of the day. But a lot of the talent is really spilling over to other areas. They’re going now potentially into companies in finance, financial services. There’s been always this tension movement between FSI, financial service industry, and tech, where FSI typically paid well, at least in certain kinds of roles. That transition can happen during these times. Manufacturing, logistics, government, in some cases, we won’t go. I won’t open the Pandora’s box on that one. We’ll just stay at that level.
But obviously, there’s a lot of talent that’s now migrating to industries, defence, for example, where it might have classically not gone to. There’s a lot of tech talent, for example, going into defence. They might end up in defence tech themselves, but defence in and of itself, in terms of the tech talent that was there, was a different animal, so to speak, as an industry. Now we’re seeing a lot of these spillovers, a lot of these movements of talent, which, in my opinion, are actually very positive. Let’s bring in the love, let’s bring in the expertise of technology to other domains that might actually benefit tremendously from it.
Bertrand Schmitt
Yeah, I totally agree with you. In some ways, it’s funny because it’s reminding me what I saw 10 years ago in mobile, where a lot of people were early in mobile apps. At the beginning, early on in the 2010-2015, mobile apps were really mostly about gaming. But step by step, many people were not so keen to stay in the gaming industry. I saw many moving and switching to different industries who were very keen to have people knowledgeable in app business, app development.
I remember seeing many people jumping to the financial industry, for instance, when banks were building their mobile apps and the like. It was for me very interesting to see that switch. These people were highly sought after by these industries because they were known to have dealt with stuff very early on that were not so easy and that were important and could apply to their own business.
Nuno Goncalves Pedro
I do predict there’s even a more interesting spillover effect that’s probably up and coming as well, which is even founders of startups and some of us investors that might say, “You know what? I just want to go and do something else in another industry.” We discussed it in the past, what areas are less impacted by AI. There’s many industries that need so much revamping that even the effects of AI will take years, even decades to be more heartfelt.
The level of core innovation that needs to happen and the underpinning of that innovation, again, some of this talent, founders, investors, they’re more than capable to bring it to other industries. We’re not saying that we’re white knights that are going to save industries. That’s not the point. It’s like the technology toolkit, the understanding that tech skills very, very quickly or quicker than humans is something that’s valuable, we believe, for a lot of industries.
Bertrand Schmitt
One thing, though, to keep in mind because sometimes I see some companies with things they can do as well as Meta or Google or others in terms of developing their own tools or product based on AI. Sometimes I’m just shocked. I’m like, “Guys, come on, let’s be realistic 2 minutes. You don’t have the capacity to pay some of the top talents like Silicon Valley can pay.” I’m talking about cash, signing bonus, salaries, stock options that we increase in value.
I would say I would just highlight that piece of the puzzle because I’ve seen again and again some industries that believe that, “Hey, I’ve hired a few guys relatively cheap from the tech industry, and now I don’t need to buy the products from the Google of the world. I can do it myself.” I mean, good luck with that. Good luck with that. You are not going to be able to compete. Find a way to use this talent in a way that is efficient and smart, but don’t think that give you the ability to compete with the big boys, because you won’t be able to compete.
Nuno Goncalves Pedro
The operating models, to your point, are fundamentally different. If you just hire a few people and say, “Well, now go and do it.” It’s like, well, you’re fitting into an organisation that has specific APIs in mind, a specific value system, how things are done, how the cycles are done, you need to cater not just to individuals, but really create organisations that really are able to scale technology and actually shift how the organisation would work.
If you want a more tech-enabled organisation, you need to create an operating model that is tech-enabled, not just come here and fit in. It’s like you can’t make much work. In some ways, fully agreed, just hiring people won’t solve the problem. Making big bets into pools of talent, into new operating models, into new processes, even into new spin-out organisations is going to make a much bigger impact.
Bertrand Schmitt
Yes. Again, at some point, it’s the very best. We stay in tech where they can get pretty insane salary compensation that no normal company will be able to afford. You won’t be able to fight meta who wants to hire the top AI engineer or top AI scientist and spending dozens of millions on some of the top talents. I mean, there’s no way you can compete there. You have to be realistic about the level of talent you can bring and, of course, making your organisation compatible with this talent. If you don’t change your ways, there is not much anyone can do. But even if you change your way, even if you do a lot of things positively, if you are not able to compete on the comp side, you have to have reasonable expectations as well.
Nuno Goncalves Pedro
Let’s talk maybe about some of the cultural and geographical shifts that we see happening. We’ve talked about many of them because of COVID and during COVID and post-COVID, the logic and what was happening.
A very obvious one is the remote versus return to work dynamic as two extremes: fully remote or full return to work. In some ways, big tech seems to be going back to the return to work, so the whole, you guys can be wherever you are, and you get paid. That has come crashing and burning, so to speak, very visibly and very publicly. A lot of the big tech companies saying, “You need to return to office, and you need to return to an office that you belong to. Otherwise, either we change your salary because you’re somewhere else. If you have an office for you to go into in the area that you’re in now, or you should leave.”
In some ways, there’s a momentum back. I think at least I have been a proponent that the world, maybe with everything that happened with COVID, would go more towards a hybrid model where you have hubs, where people work from the office, again, a lot more, but they get to work a bit from home.
In some cases, we’re seeing a really like, no, you return to the office, maybe you can work from a one day a week, and that’s it. There’s a little bit of a rebalancing, I think, of workforces. Obviously, smaller companies, startups still very much have very distributed teams, in some cases, almost fully liquid still and all remote. But if we’re talking about mid-size, high-growth companies, large big tech companies, it seems like we’re going full on back to the good old days. We’ll see how that pans out.
Bertrand Schmitt
Yeah, I will say on this topic, one thing I have seen and has not changed is that if you are some of the top talents around there on your topic, companies will let you do whatever you want. I know some people who are AI scientists, they are still working from home, and trust me, they are not going to get fired because the companies know that they have a good deal. The guy is there, and he can pick any job he wants if he’s out. I think it’s true of your average worker, but top talent can still impose their rules to most systems.
Nuno Goncalves Pedro
Let me just comment on that. I think it is true, but just for those listening, we’re talking about really top-tier talent. I was just privy to a transaction involving two parties, a larger party and a smaller startup, where the transaction didn’t go through because they were doing an acquisition of a company, the larger party, just for AI talent, and they recognised the smaller company had amazing AI talent, and still they were like, “No, you guys are going to have to come to these two hubs, and there’s no flexibility. We don’t give a shit.”
Again, we’re talking about really high top-tier talent, rock star talent. We’re not talking about the best performers. We’re talking about the the the best performers is what I would say. Then you’re allowed to do whatever you want. Then you’re prima donna, that’s fine.
Bertrand Schmitt
Yes, and I’m sure different companies they have different approach, but some people I have in mind are working for companies that are very well known to not have remote work policies, but they still have for some people. The other piece of the puzzle that some things I’ve heard is some companies that are mandating a return to the office, but they don’t have enough seats.
Nuno Goncalves Pedro
Yes, I’ve heard that.
Bertrand Schmitt
For me, it’s just so unbelievable that you’re mandating employees to come back to the office, but you don’t have the number of seats. People basically don’t have a desk when they arrive, don’t have a parking spot, practically cannot work from the office. You have 20, 30, 40% not enough capacity in the office. For me, just this stuff blows my mind because I have no way. I simply cannot understand how an HR team, how some execs have no clue to such a level that they mandate one thing on one side, but on the other hand, they don’t have the capacity. They do both at the same time. We mandate, and then we don’t have the capacity.
I’m not sure what to take about that. It really made so little sense to me that it’s shocking. I’m talking about really big companies that have the ability to have the visibility and to rent whatever needs to be rented. It’s shocking. For me, it’s interesting. You have what he said, what he talk about. When you see the practical reality, it’s quite shocking to see the, I would say quite frankly, incompetence of some in terms of how the organisers return to the office.
Nuno Goncalves Pedro
Definitely a lot of incompetence. Fundamental shift, maybe switching gears a little bit is, how is work being done? Obviously, we’ve talked a lot about demographics, and now Gen Zers are fundamentally different and looking for authenticity, mission-driven work, having side gigs. They have their work-life balance as a given, and all these elements that are being pushed onto. We’ve also discussed the social issues. Are they ready socially to manage and scale to be managers, et cetera. But I think there’s a more fundamental shift happening right now, in particular, if I look at tech, which is this notion of individual contributor versus manager.
I think we went through a time where there was a lot of more and more management layers. There’s now this discussion where the structure should be flatter. There should be still managers because there needs to be orchestration, but the structures need to be flatter because there’s a lot of individual contributors at the table.
Bertrand Schmitt
It’s more efficient. Too many layers destroy your capacity.
Nuno Goncalves Pedro
It’s always been more in computer engineering. This notion of small teams, highly empowered, highly capable to deliver with a thin layer of management is always, I think, what people have talked about in computer engineering as the best case scenario. I feel in some ways we’re going back to that a little bit. We are going back to the age of individual contribution, and particularly with this AI augmentation, the ability for me to do stuff myself.
I was talking to a founder the other day, and he’s an older founder, a little bit more experienced in the market. He was telling me, “I’m not sure I need a ton of junior developers. I can just go and look for stuff.” He’s the CTO, in this case, of the company. “There’s a lot of stuff I can do myself. I could just look for code, clean it up, debug it, compile it, run it. That’s fine. For basic stuff, I don’t need people.” In some ways, we’re like, this is the max manager, the CTO. In some cases, the CTO is saying, “I can be a contributor myself, and then I don’t need to be managing three or four people. I can do basic jobs myself.”
I feel there’s a little bit of flattening of engineering orgs going on. It might at some point go back to a lot of management layers, but that it feels, at least to me right now, that’s where we are in terms of the cycle.
Bertrand Schmitt
Yes. An interesting thing is that it goes in parallel with that new trend of AI as a manager. Not only you might have less manager for efficiency’s sake, a lot of companies realise they cannot have too many layers of management without losing their capacity to do stuff quickly, to turn around quickly. But on top of it, there is that new trend of management through AI. A lot of people say AI might be able to replace middle management. It’s also an interesting development. It might not just be about AI on the front lines of replacing the most junior jobs. There is a lot of talks about AI replacing managers.
Nuno Goncalves Pedro
Do you want to close down?
Bertrand Schmitt
Yeah, sure. Some takeaways on the topic. Definitely, there is a big shift in tech in hiring. If you want to be focused as a founder, you want to, of course, be cost-effective in hiring top talent, you should probably think about leveraging more fractional roles, talents that are not easy to hire, you cannot afford to hire full-time. You want to stay in. AI keeps improving, if not by the day, by the week, so be careful.
Stuff that you don’t believe could be done in 6 months might be doable with AI. If you are an operator, focus on a retention of your team, focus on best practices, remote, hybrid. I think that what some of the big companies are doing could be your benefit as a smaller company. Don’t think about applying the policies of the big guys because applying a different policies could be a way to get a specific advantage.
If you’re investors, I think that you probably want to think carefully about how this company is using AI, how this company is optimising resources. Are they still living like we used to do years ago? Are they full on in that brave new world, and I think it’s important to push for that.
Of course, some of this is painful when people are fired from their job, are laid off. But on the positive side, that’s usually a way to start again, to start new, to think about how you could do things differently, potentially start a new type of business as we discuss. I think everyone should think very carefully to upgrade their skills in order to better use AI, because that’s definitely the way things are going. Thank you, Nuno.
Nuno Goncalves Pedro
Thank you, Bertrand.
More shows like Tech Deciphered
View all
The Daily
112,880 Listeners
Eixo do Mal
16 Listeners
The Prof G Pod with Scott Galloway
5,559 Listeners
All-In with Chamath, Jason, Sacks & Friedberg
9,957 Listeners