74 – The Prediction Episode

Who dares to make predictions in the current landscape? We do! 

Our Predictions are back. Will our track-record continue on a high or will we be fundamentally wrong? Listen in to our Predictions for 2026

Navigation:

• Intro
• What will 2026 be all about?
• AI, AI and … more AI
• The big Hardware movements
• Of Start-ups and VCs
• Regulatory & Geopolitical Headwinds… and the Wars
• Fintech, Crypto and Frontier Tech
• Conclusion

Our co-hosts:

• Bertrand Schmitt, Entrepreneur in Residence at Red River West, co-founder of App Annie / Data.ai, business angel, advisor to startups and VC funds, @bschmitt
• Nuno Goncalves Pedro, Investor, Managing Partner, Founder at Chamaeleon, @ngpedro

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

Bertrand Schmitt

Introduction

Welcome to Tech Deciphered Episode 74. That would be an episode about some predictions about 2026. What will be 2026 all about? I guess this year is probably starting with a bang. We saw the acquisition of xAI by SpaceX. We saw an acquisition from Grok by NVIDIA. What’s your take about what would be the big themes in 2026? I guess it would be for sure about AI and space.

Nuno Goncalves Pedro

What will 2026 be all about?

Yeah. I predict a year that will be a little bit more of a year of reckoning in some way. There will be a lot of things that I think we’ll start seeing through. The fact that we are in the midst of an amazing transformational era for technology, the use of AI, but at the same time, obviously, a ridiculous bubble that is going alongside it as we’ve discussed in previous episodes. I think that we’ll start seeing some early reckonings of that, companies that might start failing, floundering, maybe a couple of frauds along the way, etc.

I’ll tell you what I will not make many predictions about today, which is geopolitics. Geopolitics, I will not make predictions at all. Who the hell knows what’s going to happen to the world this year in 2026? I don’t dare making any predictions on that.

Back to things where I would make predictions. I think on AI, we’ll have a little bit of reckoning. We’ll talk about it a little bit more in detail during this episode. Interesting elements around the hardware and physical space. Physical space, we just dedicated a full episode to it. We won’t go into a lot of details on that, but definitely on the hardware side, we’ll talk a little bit more about it.

The VC landscape is going through an incredible transformation. We’ll talk about it today as well and some of our predictions for this year. What will happen to the asset class? It seems to be transforming itself dramatically. Obviously, that has a very direct impact on startups, so we’ll talk about that as well.

And then to close a little bit the chapter on this, we will address some regulatory and geopolitical, let’s call it, headwinds without making maybe too many complex predictions. We shall see. Maybe by that time of the episode, we will be making some predictions. You guys should stay and listen to us, and maybe we will actually make some predictions about the geopolitical transformations that we will see this year in the world.

Then last but not the least, we’ll talk about fintech, crypto, frontier tech, and a couple of other areas before concluding the episode. A classic predictions’ episode. We normally have a pretty good track record on some of these, but right now, the world is going a bit interesting, not to say insane.

Bertrand Schmitt

Yes, and going back to some news, Groq technically was not acquired, but, practically, it’s as if it got acquired. I’m talking about Groq, G-R-O-Q. The AI semiconductor company focused on inference AI, and it was late December. It was a way to end the year. This year, we started again with an acquisition of xAI by its sister company, SpaceX.

I guess that’s where we are starting.

AI, AI and … more AI

We are going to start on AI. That’s definitely the big stuff. Everything these days, I guess, is about AI or has to have some connection with AI, or it doesn’t matter. I think every company in the world has seen that. You have to have the absolute minimum on AI strategy. You better execute on this strategy and show results, I would say. For the companies that were not AI native, you truly have to have a way to transform yourself. I guess at some point, the stretch might be too much, and it’s not really reasonable. Then you maybe better stay on what you are doing, especially if you’re in tech, you better be moving faster to AI.

Nuno Goncalves Pedro

Just to highlight, and I think throughout the episode, you’ll see that there’re obviously a lot of implications that would manifest themselves into capital markets. I mean, we’ll specifically talk about VCs and startups later on. But the fact that everything needs to be AI, the fact that there’s so much innovation happening right now, in my opinion, and this is maybe the first pre-topic to AI, is we’ll see a tremendous increase in M&A activity this year across the board. I mean, we’ve seen already some big acquihires we mentioned in some of our previous episodes, but we’ll see a lot more activity on M&A this year.

Normally, that’s a precursor to the opening of capital markets. I predict also that there will be a reopening of the IPO market that never really reopened last year, to be honest. M&A, a lot more, reopening of the IPO market.

Normally, it happens in the second or third quarter of the year. That’s what my M&A friends tell me. First quarter of year, everyone’s figuring out stuff. Then last quarter of the year, things should be more or less closed. Maybe the third quarter is the big quarter. We shall see. But definitely, as a precursor to our conversation today, I think we’ll see a lot of M&A, and we’ll see reopening of the IPO mark.

Bertrand Schmitt

I guess last year was not as big as you could expect on M&A given the tariff situation announced in April and May. I mean, it became quite tough to do IPO in such market conditions. Definitely, we can hope for something dramatically different in 2026. I guess talking about public markets and IPO, I guess the big one everyone is waiting for is SpaceX. SpaceX getting even more interesting with its xAI acquisition.

Nuno Goncalves Pedro

Do you think that because of the acquisition, it’s more likely that it will happen this year, or because of the acquisition, it’s less likely that it will happen this year?

Bertrand Schmitt

That’s a good question. My guess is the acquisition of xAI is all about xAI needing more financing and cheaper financing. This acquisition is a pathway to that. SpaceX being a much bigger company, a company that is also making much more revenues. I could bet that there is higher probability that, actually, SpaceX will go public in order to finance itself.

At the same time, will it have enough time to prepare itself for the IPO given this acquisition just happened? Can they do that in 6 months? I mean, if anyone can do it, I guess it’s Elon Musk. It’s a strategy to present an even more attractive company with an even more interesting story, a story of vertical integration from AI to space. I guess the story as it’s presented itself right now, it’s one about having your AI data centers in space. Because in space, you have much better solar energy production with solar panels. You have a perfect cooling situation because you are in space. Thanks to Starlink, you have the mean to communicate between the satellites and with Earth itself.

I think if someone can pull up a story like AI data center in space, I guess Elon Musk can. There is, of course, a lot of questions about is it practical? Is it economical? Yes. I certainly agree. I’m not clear on the mass, and can you make it work? Again, I mean, Elon Musk single-handedly, with SpaceX, managed to transform the space market on its head. I mean, they are the biggest satellite launching company in the world. They have the most satellites in the world. I mean, I’m not sure I would bet against him, and I guess I would probably believe that he could pull up something.

Time frames, different story. The 2-3 years data center in space for AI as cheap as on Earth, I have more trouble with that one. I mean, it’s a usual suspect with Elon Musk. You promise something unachievable in a few years, but, ultimately, you still manage to reach it in 5 or 10. Again, I would not bet against the strategy.

Nuno Goncalves Pedro

Yeah. I’ve talked to a couple of space experts, people that have launched rockets, and have worked JPL, NASA, and a couple of other places, etc. For what it’s worth, their feedback is, “No way in hell, and we’re decades away.” We’ll see. I mean, to your point, Elon has pulled very dramatic stuff. Not as fast as he normally says he’s going to pull it, but within a time span that we all see it. Difficult to bet against him.

In terms of actually the prediction, maybe to respond to the prediction as well, will SpaceX IPO? I’m going to make a prediction that has a very high likelihood of missing the mark, but I think Tesla’s going to buy and merge them both into it. It’s going to become a public company through Tesla. That’s my hypothesis.

Bertrand Schmitt

No. That’s supposed to be it. That’s how you solve that.

Nuno Goncalves Pedro

And Elon controls the whole universe. X, xAI, Tesla, SpaceX, all under one umbrella beautifully run. And SolarCity is well in there, of course, so wonderful.

Bertrand Schmitt

That’s possible. Certainly, you are not the only one thinking Tesla will acquire or merge with SpaceX. To remind everyone, Tesla is around 1.3, 1.5 trillion market cap. Depending on the day, SpaceX seems to be valued at similar range, 1.2, 1.3 trillion. It looks like it’s the most valued private company at this stage. These are companies of similar size, so that’s one piece of the puzzle. When you think about the combined company, we could be talking about a 3 trillion entity. Playing right here with the biggest companies in the marketplace today.

Nuno Goncalves Pedro

With a couple of tweets from Elon, it will rapidly get to 4 to 5 trillion.

Bertrand Schmitt

That’s so tricky.

Nuno Goncalves Pedro

Yes. On AI and back to AI, one thing I think that we’re about to see is this will probably be the year of agentic AI. Obviously, we predict a lot of growth on that side of the fence, in particular on the enterprise B2B side. We see a lot of opportunities coming through. From our perspective, at least at Chamaeleon, we generally believe that there’s going to be a lot of movements on agentic AI.

It’s also going to be probably the year of the first big fails of agentic AI that will be newsworthy. There will be some elements about that loop and how it gets closed that will happen. I think we might see some scandals already. We’re already seeing the social network of bots talking to bots. We will see other scandals going on this year even in the consumer space and in the bot to bot space, which we now can talk about or in the AI agent to AI agent space.

My prediction is we will see some move forwards. There’ll be some dramatic funding rounds along the way. We’ll see a couple of really cool things out of the gates coming out that are really impressive, but we’ll also see the first big misses of the technology stack. I don’t think we’ll go fully mainstream yet this year, so it’s probably maybe something more for 2027 along the way. That would be my prediction again.

I think enterprise will lead the way. We’ll definitely see a lot of stuff on consumer as well that is cool. Then we’ll all have our own personal assistance in our hands, basically, literally in our phones.

Bertrand Schmitt

Going back to agentic AI, we also started the year with some pretty dramatic move. I mean, the launch of Clawdbot, renamed OpenClaw. I mean, this stuff took fire in like a week or 2. It was coded by just one person who actually didn’t even code the product but used AI to build the product, 100% used AI, proposing some new ways also to leverage AI to do coding. He has a pretty unique approach. It’s not vibe coding. I would say it’s a better way to do that.

Then the surprising evolution with the launch of a social network for AI agents, Moltbook. I mean, this stuff, probably there is some fake in it. But at the same time, I think it’s quite impressive because it’s the first time we see truly 100,000 plus agents communicating directly to each other.

Yeah. I mean, that’s the first time we see surfacing the possibility of some sort of hive mind on the Internet. It’s pretty surprising. Right now, all of this is a hack done in a few days. By end of year, by 2 years, 3 years, we might discover that, actually, the best approach to AI might not be the AI assistant like we are doing today, but a combination of hundreds of thousands of AI working closely together. We might be witnessing the first sign of new intelligence in a way.

Nuno Goncalves Pedro

Things like this social network might either be Skynet, the beginning of Skynet. They might be the beginning of Her, or they might just be a fad and nothing really happens. It’s just interesting to see what these agents are doing.

Bertrand Schmitt

Totally.

Nuno Goncalves Pedro

Obviously, there are real and clear and present dangers of some of the integrations of AI we’re seeing in the market. Interesting enough, and I’ll ask you for your prediction a bit, Bertrand. I think we’ll probably see the first big mishap of AI being used in some infrastructural decision in the age of AI. I mean, we’ve seen AI issues in the past and software issues in the past. We talked in previous episodes about that as well. Mishaps of software that have led to people dying.

But I think probably the first big mishap will happen this year as well. Very public mishap of the use of AI and serve its interactions with infrastructure or something that’s very platform related, etc, that will have big impact that everyone will notice. That’s my prediction for the year as well. We’ll have the first big oops moment, as I would call it, for AI in this new age of full on AI.

Bertrand Schmitt

I would say first some perspective. I think today, people are not using AI directly for life and death decision, at least not that I’m aware. We’re not going to let AI fly a plane, for instance, tomorrow so you can be, reassured. At the same time, given there is such a race to AI, there definitely might be some mistakes. We were talking about the social network for AI agents, Moltbook.

Apparently, all the keys used to secure the AI were shared by mistake because it was not properly locked down. We can see that indirectly, mistakes will be made for sure. Two, it’s highly probable that some people will trust AI too much to do some stuff, and this stuff might not work and might have some grave consequence. Hopefully, there is not so much of this. Hopefully, it’s mostly AI used for the good. But you’re right. I mean, at some point, the more we use the technology, the more there would be issue. I mean, it’s highly probable.

Nuno Goncalves Pedro

That will lead me to another prediction, which is, and we’ll talk about more of it later, but it probably will lead to the first significant movement in terms of regulatory environment certainly in the US at some point if it happens in the US in particular, where there will be some movement that will be like, “Hey, you guys can’t do this anymore.” Because this will probably emerge from mismanaged interfaces. From systems having access to stuff that they shouldn’t have access to in the first place.

Talking a little bit more about what’s happening in AI. You’ve already mentioned some of the issues that relate actually to security and cybersecurity. We keep talking about AI. We keep talking about all these infrastructure pieces and platforms that are being built. I think we’ll have a lot more incidents like the one you just mentioned where things will be shared that shouldn’t have been shared, where people will break systems and get into it, etc.

Let’s see where that takes us, which is a little bit ironic because, obviously, with AI, the promise is that cybersecurity becomes more robust as well because there’re agents working on our behalf on the cybersecurity side. There’s also agents working on the other side.

Bertrand Schmitt

It’s a constant race. It’s the attackers, defenders. Each time you have new technology, you have a new race to who is going to attack or defend the best. Each new wave of technology, it’s an opportunity to challenge the status quo.

Nuno Goncalves Pedro

The attackers have been winning, and I feel they’ll continue winning in 2026. I think it’s going to still be a year of attack. We’ll see more and more breaches, more and more stuff that will happen.

Bertrand Schmitt

I don’t know if they will win. I mean, it’s normal that they win once in a while. For sure, some infrastructure is not updated as it should. Some stuff are not managed as it should, so there will always be breaches. I don’t know if things are dramatically going to change because, again, everyone who cares who is going to update his infrastructure with AI for defense. There is no question that you have no choice. We will see. That I don’t know.

For sure, AI will be used to attack directly with AI. Maybe you’re able to do bigger, larger scale attack. Or thanks to AI, you are simply able to create new type of attacks more easily. AI can be used behind the scene as a way to prepare and organise new type of attacks, even if it’s not used directly live in the battle.

Nuno Goncalves Pedro

One topic that we’ll come back to later is the geopolitics of everything, but maybe more broadly. On the geopolitics of AI, it’s very clear that we have an arms race going on. Obviously, the US on the one hand, China on the other hand is the two extremes, putting tremendous amount of capital into data centers just at the base of that infrastructure. Chipset development, chipset access, a huge theme in terms of the export restrictions, etc, that are being forced by the US.

I think it will continue. From a European standpoint, obviously, they’re stuck between a rock and a hard place, to be very honest. Let’s see what happens on that side of the fence. My view of the world is that certainly from a US and China perspective, we’re going to see a lot more movements in 2026, like big movements. The Chinese movements we always see in delay.  It takes us a couple of months, sometimes even more than that to understand exactly what’s going on. I think we’re going to see some huge moves this year in terms of the States, the United States of America, and China really pouring capital into the creation of the next big winners around AI.

I think the US is obviously more visible. We see a lot of these companies. We’ve just discussed xAI and its acquisition by SpaceX or merger. I don’t know what they’re calling it exactly. Effectively, on the China side, the movements I think are already very big. As I said, it will take a while to figure out exactly what those moves are. One thing that I propose is that at some point, China will have very little dependency on chipsets from the US. I’m not sure it’s going to happen this year, but I think the writing is on the wall. Irrespective of any other geopolitical issues that is coming to the fore at this moment in time. That’s one of the key areas or in arenas of fight.

Bertrand Schmitt

It makes sense. If you are China, you will look at what happened. You would think that you cannot just depend on the largest of one country. It makes rational sense, the same way it makes rational sense for the US to limit exports to China because there is value to delay some peer pressure that could use these technologies for good but also for bad. If you were an ally of the US, that would be one thing. But when you are not an ally of the US, that certainly should be a different perspective.

Maybe one last point concerning agents, I think there will be a lot that will revolve around coding. We can see OpenAI with Codex. We can see Cloud with code. There was, of course, [inaudible 00:18:28] that was trying to be big on agentic coding. I think agentic coding was one of the big transformation in 2025 and is going to get bigger in 2026. I think for a lot of people who do coding, there was a radical transformation in terms of what you can achieve, what you can do, how much you can trust AI to help you code. I start to think we might see this year, the replacement of not just one AI replace one coder, but one AI replace a full team because of the new ability to manage that at scale.

Coding might be a common activity where you are going to think about outcomes, think about objective, think about how you organise, but not really coding by itself anymore. A big change, like you used to code, directly your hand on the stuff, but step by step, everyone is going to become a manager of agent. I think in one year, we saw enough transformation to think that in the coming year, the transformation can be even more dramatic.

Nuno Goncalves Pedro

The big Hardware movements

Now switching gears to hardware. Obviously, a lot of movements in 2025 and over the last few years. One piece of thesis that we’ve had long-standing at Chamaeleon is that we will see the emergence of AI devices. Some of them have been tremendous failures as we discussed in the past. I predict that we’ll have a couple of really interesting full stack AI devices in the market this year. Why does that matter? Because, as many of you know, obviously, there’s compute that can happen in data centers and cloud infrastructure all over the world, but also there’s compute that can happen at the edges. The more you can move to the edges and the more you can create devices that actually allow you to have user experiences that are very distinctive at the edge, the more powerful some of these devices might become.

I predict Apple will not be the first to launch anything on this. I predict probably OpenAI, after the acquisition of IO, will maybe not launch something this year, but will announce something this year. I’ll step back on that prediction. They’ll announce something this year, but maybe not launch. But we’ll start seeing some devices that have some interesting value in the market, probably devices that are AI devices, but they are very focused on very specific user flows, and so very much adequate to specific activities.

I won’t make a prediction on that, but I think areas that would make sense for that to happen would be obviously around fitness, health, et cetera, et cetera, where we already have the ascendancy of products like Oura Ring and others out there. Definitely, that’s one area that might have quite a lot of developments.

I think AI-first devices, devices that are very focused on compute at the edges, providing user flows that are AI-enabled to end users, we’ll see a lot more of that and a lot more activity this year. Again, I don’t think Apple will be necessarily ahead of the game. Again, maybe OpenAI will give us something to at least think about and look forward to.

Bertrand Schmitt

First, I’m not sure it will be that transformational because if it’s not in your phone, in your pocket, there is only so much you can do with it, and there is only so much computing power you will have. I’m doubtful it would be really impactful this year.

Nuno Goncalves Pedro

I feel we’ve been discussing this shift of paradigm in input and output. For me, some of these devices could lead to that shift. Because, again, a mobile phone is not a great long-term paradigm for the usage that we have because it’s really constrained by the screen. The screen is really what takes most of the battery life away. If we didn’t have that screen, what could we do? If we have the block that is as big as a mobile phone, and it didn’t have a screen, it was just compute, that’s a mini computer, a microcomputer.

Bertrand Schmitt

That’s a fair point, but I don’t see that transformation this year. That’s really more my point. I can see that you can have AI-enabled smart glasses, and it’s clear there is a race to AI-enabled smart glasses. My point is more to go beyond the gadget, it would take quite a while. It would need to have cameras. It would need to analyse what you see. It would need to hear what you hear. Again, it might come, but then at some point, it would be okay, what do you do with it?

We have the example of the movie Her. That’s showing Her what it could be. There are definitely possibilities. It’s clear that if you take the big VR headset like the Apple Vision Pro, there is a failure from that perspective in the sense that I think it’s a great, amazing device. The big problem is that it’s doing way more that makes sense.

I think there will be a clearer separation between your smart AR glasses that has to be light, that has to be always unconnected, and that’s primarily there to help you make sense of the world around you. The true VR headset that doesn’t really require much in terms of AI, and it’s just there to immerse you in a different world. For this, we know, unfortunately, in some ways, that there is not a lot of demand for it.

Maybe there is little demand because you are too hidden in your own world. The technology is not working well enough yet. There are a lot of reasons. But I think Apple trying to do both at the same time, AR and VR, with the Vision Pro, was a pretty grave structural mistake. I think we would see a clearer line of separation between the two. There is bigger market opportunity for AR glasses. That, I certainly agree. There is opportunity to connect that to a computing device. As you talk about, your glasses are your screen, your phone becomes something in your pocket connected to your glasses.

Nuno Goncalves Pedro

For me, Apple has their way of doing things. From the perspective of what you said, they normally really plan their devices. Even if it’s a big shift in terms of a new area, like they tried with the Vision Pro, and we criticised them for launching it as a device that should have been more of a dev device that they really launched as a full-on device, but that’s their playbook, classically.

I think Apple needs to change how they put products out and how they experiment with those products, et cetera. I think they have enough money to be doing everything all the time and figuring it out. If they don’t want to put it out, then they need to do a lot more hell of testing internally with their silos, but they should be playing across all these arenas, VR, AR, everything. They just should put devices out that are either ready for prime time, or they should call it something else. They should call it like this is a dev device or whatever it is.

Bertrand Schmitt

I agree with you. My complaint is more that it was marketed as a consumer device when it was not. It was a true developer device. Two, they tried to mix the two at once, and it made no sense. No one is going to walk in their home or in the street with their Vision Pro on their head. You have to be deranged, quite frankly, to have use cases like this.

I think that for me is a crazy mistake from a company like Apple that prides itself in pure UI, pure user interface, very well-designed device for one specific use case, not mixing the two use cases. We still don’t have Macs with a touchscreen, you know?  We still don’t have an iPad with a good OS that makes use of this great hardware. For some strange reason, they decided to mix everything in the Vision Pro with a device that weighs a ton on your head and is so uncomfortable. That’s why, for me, I’m like, “Guys, what is wrong? Why did you let this team run crazy?”

I hope at some point, Apple will go back to the drawing board. My understanding is that that’s what they are doing. They are going to have two devices, one smart glasses, an evolution of the Vision Pro, just focus on VR. They might actually abandon the concept of the pure VR-oriented headset. Because, from a market size perspective, it might not be big enough for Apple, quite frankly.

Nuno Goncalves Pedro

I read on all of the above, and people at this point was like, “Why are then players like Samsung and others not doing it. LG, et cetera?” Because those players historically have not invented new categories. They’re amazing at catching up once the category is invented, and then they scale the hell out of it, and that’s what these companies have been exceptional at. I wouldn’t see a dramatic innovation, I think, in terms of devices coming from any of the big ones on that side of the fence. Not to disrespect them in any way, but I think that’s not been their playbook ever.

Again, if the origination doesn’t come from a start-up or from an Apple, I don’t see those guys going after it. My bet is that we’ll see some start-up activity and, again, hopefully, some announcement from IO now within the OpenAI world.

Bertrand Schmitt

I would slightly disagree with you. I see where you are coming from. But take the Samsung Galaxy Note, that sudden much bigger headphone that no one was doing that was launched by Samsung, at some point, it forced Apple to launch an iPhone Max. Let’s look at the Z Fold that Samsung launched 7 years ago, copied by everyone. Now Samsung launching a trifold. Apple has still not launched their foldable phone. I think there is a mix, actually, of sometimes-

Nuno Goncalves Pedro

For me, that’s not a proper new category. It’s still a mobile phone. It just happens to have a screen that folds in half.

Bertrand Schmitt

The iPhone was still a mobile phone, you could argue. 

Nuno Goncalves Pedro

No. I think the iPhone was…  I could actually agree with you on that point. Maybe Apple is not as innovative in that case. I think what Steve Jobs was exceptionally good at in terms of his ability as this master product manager was to be an exceptional curator of user flows and user experiences, and creating incredible experiences from devices based on that. That was his secret sauce.

Could you say, “Wasn’t all of this stuff already around?” It was. You just put it all together very neatly and very nicely. But if you’re talking about significant shifts in how a category is done, the iPhone was a significant shift in how the category was done. The Fold is still an interesting device. I actually have a Fold right now in front of me. The 7 that you highly recommended to me that we both got, the Z Fold 7.

I think they do amazing devices. I don’t think they normally are the most innovative players. Then, when they come to innovation, it comes from technology edges. Obviously, they have Samsung Display, there’s a bunch of other things. They had the ability to do foldable screens in-house themselves.

Bertrand Schmitt

I don’t disagree with you. I think there is an interesting situation where some companies have some strengths, another one has some strengths. My worry with Apple is that this was not demonstrated with the Vision Pro. The Vision Pro was a hot pot of technologies barely integrated together, with use cases absolutely not well-defined and certainly not something that makes sense for most of us.

There is a question of has Apple lost it? While Samsung actually keeps doing their own stuff, that, yes, might be more minor improvements, but at least they are doing it. Because it looks like Apple is missing the train on even the minor improvements. By the way, you might not be aware, but Samsung launched its Vision Pro competitor. Interestingly enough, it might be a better product in some ways, being much lighter and much more comfortable.

Nuno Goncalves Pedro

We should play around with that and report back to our listeners.

Of Start-ups and VCs

Moving to venture capital and the startup ecosystem and what’s happening there, I think it is very much a bifurcated environment, and it’s bifurcated for both VCs and for startups. If you’re a startup in the AI space, and you have the hottest team since sliced bread, and you can create FOMO at the speed of light, you can raise ridiculous rounds. Five hundred million at the $3 billion, or $4 billion, or $5 billion valuation, and you still haven’t really even started.

First round, you can raise 500 million. That’s back to the whole discussion on Bubble and where are we, et cetera. Some of these companies might actually become huge, some of them might not. But definitely, we are seeing really the haves and have-nots on the startup ecosystem with incredible teams raising a lot of money very, very early on or mid-stage if they’ve already existed for a while, and then the rest not being able to raise.

We see a lot of non-necessarily AI sectors, some of the areas of SaaS that don’t necessarily have AI in it, or fintech, or the consumer space that are really, really struggling. If you don’t have an AI story for your startup right now, it’s extremely difficult to raise money unless your numbers are just the best numbers ever.

That’s, I think, the first part of the element of bifurcation that we’re seeing today. The second element of bifurcation that we’re seeing today in terms of fundraising is for VCs themselves, and really propelled by the large VC firms raising more and more capital in recent orbits, announcing 15 billion across funds raised. Lightspeed, I think, had made an announcement a couple of weeks ago as well. They’ve raised a bunch of money as well. The big guys are all raising a lot of money.

At some point in time, the question some of you might ask is, “These VCs are redeploying more and more money if they have a couple of billion for a VC fund. How does that look like? Is that still VC?” My perspective, I’ve shared before in some of our previous episodes, is that that’s no longer venture capital. At that point in time, we’re talking about something else. Private equity hedge funds, if you want to call them, maybe funds that are really driven by growth investment or late-stage investment.

If you have a couple of billion under management, you’re not going to make your returns by writing a $3 million check in a series seed and leading that round.  That has implications for everyone in the ecosystem. It has implications for smaller funds that obviously have a lot more difficulty in raising capital. It’s difficult to differentiate.

Last but not least, also for startups that really continue searching for that capital that is out there. Andreessen Horowitz, for example, runs Speedrun, which is a great program for companies around consumer in particular. Initially, it was a lot for gaming. But at some point in time, Andreessen Horowitz could decide that they don’t want to invest more in you. They just put money from Speedrun, which is obviously a very small check compared to the very large checks they could write mid to late stage and that will have an effect on you as a startup.

What happens at that point in time if Andreessen Horowitz is not backing you up in later stages? More than that, what happens if I can’t get these big funds interested in me? Are the small funds still valuable to me? Punchline, my view is yes. Obviously, we’re a smaller fund, so there’s parochial interest in what I’m saying. Small funds can still create a ton of value for you, also in terms of credibility, ability to accompany you in those first stages of investment, and the ability to bring other larger investors later down the road as well.

There’s definitely a big movement happening in terms of the fundraising for VC funds, which we shouldn’t neglect, which is the big guys are raising a lot more capital and are therefore emptying the market to smaller funds that are having more and more difficult raising at this point in time. We had discussed that there would be a need for concentration in the industry, that micro funds would need to concentrate, and we didn’t have the space for so many micro funds as we had around. But the way it’s happening is extremely dramatic at this moment in time. I think it will continue through 2026.

Bertrand Schmitt

Remember a few years ago, with the rise of AI, there was more and more of the question about, “What’s the point of SaaS at this stage?” Because SaaS was around for 15 years. Basically, how do you come up with something new that was not already tested, validated by the market? How do you bring something new? We say this was reinforced to the power of 10.

If your product is not clearly built from the ground up for a new use case enabled by AI, anyone could then might have built your product 5, 10 years ago, and therefore, why now has no clear answer, and it’s a big problem. I’m still surprised myself to still see some entrepreneurs where you talk to them about AI because you don’t see them in the deck, and they explain to you, “It’s not yet there,” and you’re like, “What’s wrong with you guys?”

Fine. Do whatever you want. Do a small business and whatever, but don’t think you can come up pitch and raise without an AI story. The second category is people who come with an AI story, but you can feel very quickly, I guess you saw that many times, Nuno, where just a story layered on top with little credibility. It’s not better. It’s not enough to just have a story. Your business needs to be radically built differently or radically proposing some brand-new use cases that were impossible to solve 5 years ago.

Nuno Goncalves Pedro

To stack up on that, absolutely in agreement. If you’re just adding to the story, and it’s an afterthought, and you’re just trying to make the story somehow gel, once you go into one or two layers of due diligence, your investors will very quickly realise that you’re not really AI-first or dramatically AI-enabled or whatever. It’s just you’re sort of stacking something on top of another thesis. It needs to make sense from the product onwards. It’s not just, let’s just put it together with chewing gum, and magically, people will give you money.

It was true also if we remember the good old crypto blockchain days, where everyone’s investing in crypto. A lot of stories that didn’t make much sense. In that sense, it’s not very different. I would go one step further. I think in the world of the VC winter that we’re a little bit in, where it’s more and more difficult if you’re a smaller fund to raise your fund at this moment in time, there’s a lot of sources of distinctiveness still talked about, like proprietary networks, access to deal flow, fast track record, all that stuff that really, really matters.

But our bet continues at Chamaeleon continues being that you need to be AI-first as a VC fund yourself. You need to have core advantages in using not only readily-available AI tools or third-party available AI tools, data sources, technology stacks, but actually building your own stack over time, which is what we did with Mantis at Chamaeleon.

Again, just to reinforce that, I think we’re at the beginning of that stage. We, Chamaeleon, are ahead of the game, but we think that the rest of the market will have to move towards that as well. Still, to be honest, very surprising to me to see that many significant large players are doing very little still around some of these spaces. They have data scientists. They’re running some tools. They’re running some analysis and all that stuff, but it’s still, again, back to the point I was making for startups, all glued up with chewing gum. It doesn’t all come together nicely, which it does need to from a platform standpoint.

Bertrand Schmitt

It’s quite surprising. I agree with you that some VC funds might think that they can do business as usual in that brand-new world. It’s difficult to believe.

Nuno Goncalves Pedro

Maybe moving a little bit toward the capital formation piece. We already discussed the M&A space really accelerating. We’ve also discussed the IPO market and some predictions on that. Secondaries, there’s obviously a lot of liquidity coming from secondaries from mid to late stage. I think it will continue throughout the rest of 2026.

A lot of activity in buying, selling in secondaries as some asset managers are becoming more distressed, as some very high net worth individuals and family offices are becoming more distressed as well, at the same time, where there’s a lot of opportunities to potentially arbitrage around some investments. I believe a lot of money will be made and lost this year by decisions made this year, just to be very, very clear in terms of equity, purchases, et cetera.

Exciting year ahead of us. Definitely a very, very interesting market ahead of us. Secondaries, M&A, growth, and late-stage investing, also, early-stage investing will continue just for those that were wondering. Last but not least, the public markets, the IPO market as well.