Episode 124
2025 Review
About this episode
In this annual 2025 review special, Alec and Jack take a step back to reflect on one of the most transformative years yet for crypto, AI, LLMs, and digital assets.
From Bitcoin’s institutional turning point to AI becoming a daily tool for millions, this episode captures how Web3 and emerging technologies moved from niche experimentation into the mainstream. It’s a wide-ranging conversation about what actually mattered in 2025, and what it tells us about the road ahead.
Key Highlights
- Crypto and Digital Assets Go Mainstream (But Not Without Volatility). The hosts unpack why 2025 marked a legitimacy breakthrough for Bitcoin and crypto, driven by ETF approvals, institutional adoption, and a more crypto-friendly U.S. political environment. At the same time, Bitcoin behaved more like a risk asset than a safe haven, reacting strongly to macro events like tariffs and global market uncertainty. A major standout was the explosive growth of stablecoins, which quietly found real product–market fit by enabling faster, always-on financial settlement infrastructure .
- AI and LLMs Become Ubiquitous. If crypto found legitimacy, AI found ubiquity. 2025 was the year AI tools—powered by increasingly capable LLMs—became everyday utilities for individuals and enterprises alike. From workplace copilots to creative tools, AI shifted from novelty to necessity. The episode explores how rapid competition between OpenAI, Google, and others drove massive improvements in reasoning, context length, and cost efficiency, while also flooding the internet with AI-generated content and reshaping how people trust information online .
- Wearables, Robotics, and the Next Computing Shift. Beyond software, 2025 hinted at what comes next in hardware. Meta’s smart glasses helped prove the viability of AI-powered wearables, while humanoid robots moved from labs into early industrial testing. From BMW factory trials to Tesla’s Optimus ambitions, the conversation highlights how automation and AI-native machines are beginning to integrate into real-world workflows—setting the stage for much bigger changes in the years ahead .
Looking back, 2025 wasn’t about a single breakthrough, it was about consolidation, legitimacy, and scale. Crypto became harder to ignore, AI and LLMs became unavoidable, and digital assets quietly embedded themselves into global financial plumbing.
As Alec and Jack reflect, the pace of change is accelerating, making predictions harder, but also making the future more fascinating than ever. This episode captures a pivotal moment before the next wave hits, setting the tone for what could be an even more disruptive year ahead.
Transcript
Auto-generated from the episode audio.
Welcome back to untangling Web 3. We've got another great episode for you today, but before that we'd love it if you could like the video and subscribe to our channel. It's the best way that you can support the two of us and help the channel continue to grow. And with that said, on with the show. Welcome back to the untangling Web 3 >> Alec and I just joking about how I sound even more nasal than usual because it's obviously we're in the post-Christmas period. We're actually in the new year as we record this.
And obviously, you know, we've got the perma-flu or whatever the hell's going around, but we're going to power through and we're going to look at today a quick review of our 2025 on the podcast, what happened in the tech world in in Web 3 and crypto and AI of course, and then look ahead to what do we think is going to be, you know, the dominating our our our headspace in 20 26. So yeah, Alec, how was your Christmas first of all? >> Yeah, it was good.
Also had this like perpetual flu. So I think I've been ill for the last 4 weeks. We haven't really seen each other too much lately because we've been so busy with other things. So it's actually nice to get back in and record. But yeah, it's been good. I'm happy. Like right now I'm completely snowed in in Wales. So it's a good time to do nothing but sit here and talk about tech with you. So yeah, excited to be talking about it. >> Yeah, well, it's not at least you're getting, you know, the proper Welsh experience there, Alec.
I'm glad you've got the snow. Um but yeah, so let's let's just kick straight off. I think there's a lot has has happened. We've obviously covered some some of the major stories in the year um in in in in all of Web 3, but you know, crypto in particular has had a bit of a wild year. So I think maybe we should just start with reviewing what happened in crypto because in many ways, I think this is the year that most of us would agree crypto kind of went fully mainstream and not from an adoption perspective but maybe from a legitimacy perspective.
And I'm I'll come back to that point in a second because it wasn't a year of crypto being, you know, fully legitimized for everything. And there were also many, many instances of, you know, problems in the industry as well or, you know, potential misuses of of the technology and the tooling and the tokens as well. Um but from a from a perspective of institutions actually you know, digital assets in crypto, I think this is the year for me where it feels like now the industry is going to struggle to die, okay?
It is as as positive I as as I can spin it. I think this is the year that no longer is it a pipe dream, you know, it's it's not just an industry that's that's begging to be for legitimacy, right? It's not just coming and saying we've got we've got a huge value proposition. Um you know, we need to sell you on this. Actually, it looks like a lot of the public and a lot of the major players in the finance world and governments across the globe now are taking crypto, digital assets very, very seriously.
Obviously, you know, the new US regime with Trump that completely changed the way that the US is playing in the space. And obviously, you know, they tend to lead the market. So that has that's really changed the outlook. Um you know, from January the 5th or whatever it was this year. And it's been pretty much one way traffic from then. But yeah, what what are your thoughts on on the year of >> way traffic, but when you look at the price, the price has definitely not been one way.
I think like the hope that a lot of people felt at the start of the year has kind of subsided a little bit maybe, but we can go into a lot more detail on that in a sec. I think one of the things you said that was quite interesting is it's definitely not going to go to zero. There's that Michael Saylor quote that he always seems pushing out there. It's like not financial advice. Not financial advice. Bitcoin either goes to zero or 1 million. I've seen him saying that more and more lately, which is quite interesting.
So that means it's going to go to 1 million, but not financial advice. I think that it's hard to decouple like talking about institutional adoption without talking about the current administration. Like that was the real turning point in a lot of people's minds was the current administration meant that regulation would be a lot friendlier. That would mean that institutions could move more comfortably, especially in the US and that's where you know, all the money is effectively and that would drive sentiment and expected demand. So some of the first things that happened at the start of this year were the ETF approvals and everyone was like, okay, ETF approval what does that mean?
We've done a lot of episodes on it. We talked to John and Garcia about it. And that is basically making crypto accessible to big, you know, investors, to big hedge funds, having the Black Rocks of the world basically put your crypto assets into an easy investable device so you can basically plug some money into it as a retail user or an institution or whatever and you get exposure and access to crypto. And everyone was like, well, this is going to mean much more demand for crypto assets and the price is going to go through the roof and it did.
Like when we look at the start of the year, the pro-crypto administration, ETF approvals, institutional adoption, all this kind of stuff meant that Bitcoin rallied past 100k on all this kind of optimism and that was extremely exciting and that kind of continued for quite a while, right? Like this was the first part of the year like was a huge was huge for crypto, Bitcoin prices, all this kind of stuff. I think we hit near highs in October of about 120k. Like that's huge. That was absolutely epic.
Like everyone in the space was very happy. And then some of the the negatives came in with the current administration around tariffs. And I think this really emphasized the fact Bitcoin isn't an insular asset. Like it isn't kind of self-contained, right? It is very exposed to macro trends around the world and like I think that was one of the most interesting things is that the tariffs that kind of happened and affected like global stock markets, you know, global everything basically, also had a huge impact on Bitcoin and crypto.
and we haven't really fully recovered from that position. >> Yeah, I think um it's it's a good point, and I'm glad you you call me out on on, you know, the the one-way traffic point. I guess I should have been clearer. Uh and when I say one-way traffic, I'm not really talking about the price. I'm talking about in terms of the the industry moving from strength to strength, right? And and and it being accepted by more and more more organizations, people uh becoming more comfortable with it.
But yeah, in terms of the the speculative asset, the speculative side of Bitcoin and other other crypto, I think you're absolutely right. People, you know, there was there was to some degree priced in, but to some degree maybe not. There was a lot of anticipation around, you know, the regime change in America when it at the start of the year. You obviously had a bunch of big changes that came with that. You had the executive orders being signed immediately for Bitcoin strategic reserve, right? That was a big moment.
That had been anticipated for a long time. Um but still significant nonetheless. And you know, this is all following, as you said, the ETF approvals from the previous year. Like this is building on uh further levels of layers of legitimacy and and kind of putting this in the hands of your everyday investor as well as your institutional investors. Um so you had the, you know, you had big events like the executive order. You have crypto task forces springing up. You have people at the highest levels of government now being employed to look after the the strategy around uh Bitcoin and other crypto, right?
So I think that that's kind of where I'm coming from. >> I get you. Yeah, I think one thing I've noticed more I think more than any other year we've been doing the show because, you know, we started this in what, 2023? uh I found it interesting. We actually had um you know, I remember one of our previous guests mentioned how uncorrelated Bitcoin is from other stocks. It was something like 1% correlation or something something strange like this. Um but I think this year it feels like again not not not not I haven't done the technical analysis on this.
But [snorts] it does feel to me like more than since maybe 2020 when we saw that huge you know, market crash when COVID-19 hit uh in Bitcoin and and other cryptos. This year it's felt like it's more behaving like a tech stock than anything else. It's more reactive to the typical market sentiment new cycles that we've been having, you know, we we we're talking about the AI bubble very soon, you know, it's a spoiler alert for next episode but lots of the market narratives that affect the kind of magnificent seven and the and the big techs tech stocks I think have been influencing the the price of these things in a in a similar way that I don't think we've seen for a while which I think again just reflects that that point that Bitcoin and and and and Ethereum and other cryptos were to some sense more decoupled in the past and now it's that you know, that thing we we always talk about with the the maxis.
The more this gets ingrained in the traditional financial world, the more and more this will look like a traditional financial instrument, right? Even though under the hood you can still access it and use it in a more kind of unilateral self-custod- custodial way, self-sovereign way. The actual price dynamics are being more and more influenced by these bigger players. So yeah, it's um it it that was one of my big observations for the year I think for crypto. >> Get you and it's not to say that I know we started this stuff being like you know, we're a tech podcast.
We don't want to talk about price too much and I've kind of like come at it from the price angle and I think the point is that what I've noticed from the space is that there is an obvious coupling there like the price tends to mean activity when there's more money, there's more bags, there's more you know, funds to throw at developers to build projects, there's more excitement, more people move into the space, all this kind of stuff. It's impossible right now to decouple price and activity and institutional adoption and all this kind of stuff in my opinion.
So yeah, that's why I talk about price. And then you are right though institutional adoption is coming. I think one of the big things that sorry, institutional adoption has come. It is like will continue on the upside. I think that one of the things that maybe was like locked in a little bit is like they people as they often are were very impatient. Like the administration coming in, they expected everything to kind of happen instantly. When we talked about the strategic reserve being announced people were like, "Oh, you know, Trump's going to start buying Bitcoin for America." No, no, the strategic reserve was like we're not going to sell it and we're going to look at what happens next.
Like these things even with the current administration which is kind of like quite, you know, slapdash and move quite quickly. Um, they take years to do. So it's like what you're saying the kind of the groups have been started, the research is being done, the steps are being taken in the positive direction. So it is likely that these things are only going to become more mature. And what does that mean for the space? Like we're already starting to see like what I think one of the big success stories of this year has been less around like Bitcoin, Ethereum, all these kind of assets has been about stable coins.
Like stable coins have absolutely taken off. And like we talk a lot about like are stable coins just an interim solution? Like do they have longevity? Is there like this kind of like utopian end goal we think of where you just have crypto assets you don't need stable coins? Potentially, but like if that future is the inevitability you still need like the interim of stable coins to bring traditional fiat currency into the crypto space and you know, stable coins have like I said absolutely taken off.
Like they're starting to compete like USDT and USDC are starting to compete with the number of transactions like Visa and MasterCard are doing which is just bonkers. Like I think the market for them is hundreds of billions right now. Which again like just looking at that like two to three years ago is is incredible. And like Tether itself always comes up at the top of like the number of employees to how much revenue is generated because it's such a small team generating so much revenue. I think the only company that does a bit better is only fans, which is really weird.
But yeah, it's we've come a long way. >> Yeah, I think that is that is the other key theme that I'm glad you've highlighted for the year in crypto really is, you know, alongside the the hype and the mania and the craziness frankly of what's happened with the US and what that's done for, you know, Bitcoin primarily. I think that the the fact that stable coins have found this product market fit. Is something that surprised me a little bit and maybe I don't know if it surprised you Alec, but it's like it's like NFTs a little bit.
Not that I think NFTs have ever found product market fit, but they they look like they had for a short while in the sense that this is technology that's been around for a long time. It's nothing new to us. It's kind of was, you know, if you like if you ask me at the beginning of 2025, I would not have expected it to be as big a theme as it was last year because, you know, it's kind of it's it's simple. It's it's non-volatile crypto assets that are very very good as a settlement layer, right?
It doesn't have all the benefits and properties of kind of native crypto necessarily, but it serves a big purpose. But what we did see this year is all these companies, institutions, financial institutions starting to recognize its value for settlements, you know, out of hours for example. I know that you know, you you look at something like PayPal now, it was big news that they they were getting involved. You know, they've launched a stable coin PYUSD this year. And now by all accounts they are just settling, you know, Visa transactions using [snorts] using the stable coins, right?
Just because the infrastructure is always on, always available, low friction, right? And that is that is a real win that is completely separate from all this other price related hype or to do with any, you know, key individuals pumping the price and driving the the kind of the the mania adoption of crypto. It's actually just recognizing that the the is has value that these companies can can can you know extract them and um and utilize. So yeah, I think that was that was a really again it surprised me cuz it's kind of boring technology.
>> But you know, often boring technologies are the ones that are the most benefit most benefit. >> That's where you make the money. Boring technologies are where you make the money. Like I think that it's like the promise of crypto and like Bitcoin specifically is so abstract to so many people and like it requires so much trust in something that is very new and novel in the grand scheme of like fiat and assets and commodities and all these types of things. But when you talk about stablecoins, again, it sounds simple, it sounds really boring, it sounds really safe.
But when you're talking about like a trillion-dollar industry and you say, "Okay, I can make it slightly faster or I can give you 24/7 uptime." Like that is huge, right? And that's the kind of like the incremental steps that big institutions do tend to want. Like they don't want to radicalize absolutely everything and overhaul their entire system. They want to make micro improvements and squeeze efficiency out like as easily as possible. So that is a kind of when you It's one of these things like in hindsight to say like that is obvious that it's always going to be a massive gain for them.
And yeah, it has kind of come to fruition. I think the one big thing I want to say on this stuff as well is I think one of the things that kind of is or some of the maybe the more crypto maxi people again stressed about is that like you say Bitcoin exists in like the kind of in the scheme of everything else. It has the macro trends affecting it constantly. That's something that 2025 has really emphasized, right? Like key events really affecting the price of Bitcoin and the adoption of Bitcoin and all this kind of stuff like tariffs for example, like the AI sell-off when the you know, deep seek was announced.
All these kind of things directly affecting Bitcoin. I think one of the things that has shocked a lot of people, surprised a lot of people is that Bitcoin is being treated like a risk asset. And I think the long-term proposition of Bitcoin for so many people, especially the people from the early days, is Bitcoin acting as a safe haven. Like we always describe it as a digital gold. Like you look at gold right now, gold is at all-time highs. Like gold is absolutely performing because there's so much uncertainty in the markets, in the world, fear of AI bubbles, wars kicking off, all this kind of stuff.
Gold and silver are at absolute all-time highs because they are, you know, safe Bitcoin is not acting like a safe haven as many people thought. Like that's the end state of Bitcoin, right? It's acting as a risk asset. So when people get a bit afraid markets might tumble and there's volatility, actually Bitcoin is not taking advantage of that and tends to be one of the first things that is sold off. Obviously, it's actually, you know, all the altcoins first, then Bitcoin, but yeah, it'd be interesting to see you know, what 2026 looks like and if Bitcoin can move from a risk asset to a safe haven as many people think that the end goal should be.
>> Yeah, I mean, if you just look at, you know, as we record this, the past 1 year performance of Bitcoin doesn't look so great. It's down something like 12% in US dollar terms, right? Which is here as you see, it's >> zoom out. That's what they always say, zoom out. >> Yeah, zoom out, but if you zoom out on the gold chart, it just probably looks even better, right? So Yeah, it's um again, it's it's something that I I just I do wonder about market saturation for Bitcoin.
There are two kind of competing things. It's like, yeah, is the demand to hold Bitcoin, you know, cuz at this point as we say, like using Bitcoin is is not exactly what people are there, you know, buying it for. They're not acquiring Bitcoin to use it in any meaningful sense apart from to hold it and save it as a as a potential inflation hedge, right? As a store of value that will outpace inflation ultimately. >> But isn't that what you do with gold? Isn't that the same reason people buy gold and >> I just I just mean in terms of the So the the reason you could have bought 10 years ago, right?
The promise of this decentralized financial future where you can pay outside of the, you separately to the existing systems. Then you say well, how much value can a system like that capture? Well, now the question is more okay, how much can value can system capture if it's purely this kind of asset and store value? that that that's the question is, you know, what what's the ceiling on that? Like what point do we hit saturation? We've got all these you know, big companies offering like even in the UK now we've got exchange exchange-traded notes you can buy exposure to Bitcoin in a tax tax-efficient account like an ISA, right?
So it's not hard to to to get it to to get hold of Bitcoin these days. It's not hard to get hold of exposure to Bitcoin. So, you know, what what point do we hit that ceiling? Does it just keep increasing? I don't know. But obviously you do have that scarcity baked in. So yeah, I'm interested to see like will if if there's a second year of this kind of of this kind of performance then will people still be quite as bullish? Will Michael Saylor will his plan continue?
With strategy? I don't know. But yeah, I mean I don't have it I don't think I have any I've given up predicting things in crypto. Because I just think it you know, you you never know. >> Come on, you got to predict something. You got to give us some predictions. >> Uh oh. I mean, I don't know what is there to predict, right? Because that's the other thing I think Okay. My prediction is a boring year. There we are, that's a different type of cop-out. But I think a lot of the a lot of the the bounce to crypto and Bitcoin that were being talked about for the last three or four years, right?
With a second Trump presidency, with a change in SEC, with new regulation. Like we've had all those things happen now. What with with adoption of the big TradFi, you you institutions. What what's left? What are the big events coming up that can that can boost the industry? I don't necessarily see too many right now. So, I feel like this will be an interesting just kind of consolidation year, in my opinion. Um and I think this will be one of those years where, you know, who knows what's going to happen in the world next year.
This will be a one where it's really interesting to test the hypothesis of store value. Like, does it play out as one? Or does it continue getting spooked by what happens in the market? Cuz if if it if it doesn't perform well in times of crisis, then then, you know, maybe we do go full circle and and and we're back in the danger zone with it. But, yeah, I don't think I have any more meaningful prediction than that for for crypto. >> I think that's fair enough.
I think the big thing will be is like if traditional fiat systems really start to capitulate, and I think that Bitcoin's success outside of just No, within the spectrum of Bitcoin's success is about how much people want it, right? And how much people trust in it. And it's as simple as that. I think that what will accelerate Bitcoin's success is capitulation and collapse of fiat systems. If people start to trust less and less in fiat systems, if say, you know, the US or Britain or Europe or whatever it is starts to over inflate, print more money, there's less trust in reliable institutions, that kind of thing, I can see that Bitcoin becoming more and more adopted.
Less outside of the regulation, this important person says this, I'm going to speculate to this, more about everyday people being like, oh, actually, I don't trust that my money is worth what it will was 10 years ago, 5 years ago, a year ago. I am starting to see that things like Bitcoin could be a means to hedge against inflation. Uh Is that going to happen in a year? No, but I think that a lot of the big players like the Coinbase of the world are starting to get smarter in their pitch and starting more and more to be like, you know, pitch it towards your pound was worth this and could buy this 10 years ago.
Now it can't. Bitcoin can. And yeah, there is still something around the volatility and the stability. Like you look at the actual price of Bitcoin, how is that a hedge against inflation when it, you know, halves over the course of 4 months? But I think yeah, it'd be interesting to see if like the the definite prediction is that fiat will continue to depreciate and like your pound is not worth your pound. Um so yeah, let's see. >> I think um yeah, that that's that's maybe a good place to end our our crypto segment on cuz that is the the first That's the that's the first big, I guess, tranche of what we've seen in the year, right?
That's dominated the Um and yeah, it will be interesting to see how it continues to perform and survive in in times of potentially increased crisis. Not that I'm predicting it, but it seems like everybody else is. So, you know, or like maybe even if if if there is a a global recession something like that that as as many people are predicting. Um it'll be it'll be at least interesting as data to see how how Bitcoin and crypto, >> you know, perform in that in that environment.
Uh so that's one good thing that could come of it um even if that might be the only thing. Um anyway, let's let's let's >> Recession comes, money printer goes brr, and then Bitcoin go up. That's what a lot of people say. Not that we're saying >> of people say that. Will it be true? Um but anyway, the other thing a lot of people is talking about this year, obviously, is AI. Everything AI. Obviously, we covered robotics and the nuances of how that relates to AI.
I know you're going to want to talk about this as kind of the other big theme of the year. So, yeah, why don't why don't you kick us off with what What did you What did you make of the year in AI and and and what was interesting for you? >> Um I think AI is kind of become quite ubiquitous, which is not You know, surprising. Yeah, I don't know. But like I think more like I think what did blow my mind is that it was more and more like every single person you spoke to, not just people in tech, not just people that we were friends with, but moms and dads.
That was one of the things like we often talked about, you know, Bitcoin and crypto to come back to that like a year or two ago when we started the podcast is when is that going to become mainstream where they are using it. And the answer is they're still not, right? They don't understand why they would use it outside of like money go up, but AI is different. Like their killer application like do you remember when we first started this podcast like early on we had the chat GPT corner.
Like how new and novel that seemed like two and a half years ago. I mean we quickly dropped that because >> that was a great idea. >> Yeah, yeah. Well, we realized that, you know, you and the robot were quite similar in terms of tone and delivery. So we thought we'd only have one and you got a bit jealous. But I think the obvious thing is I say 2025 like before 2025 usage was an even shallow. People didn't really know what it was. 2025 was the year that it became an absolute household name and absolutely everyone I know uses AI on a weekly basis like which I think could we have predicted that at the start of 2025?
No, I actually don't think we could have. I didn't think it would be as commonly used at the start of the year as it is now. Like it that has really surprised me. But I think yeah, it's been an interesting year where it's gone from maybe 20% of people I knew using models on a regular basis to now basically everyone. 80 to 90% of the people I know are using AI models on a daily basis. I think maybe we are slightly ahead of the curve in the UK and the West more generally and that probably isn't like an accurate depiction of like the what the global setup is, but there are some positives and also a lot of like disadvantages to that as well.
>> Yeah, I think you're you're you're bang on that. Yeah, this was the a year where it felt like everyone and their mom was was using AI. Um like just just some kind of anecdotes very early in the year I used I think it would have been it just Dolly at the time cuz I would you know it was good enough anyway. I made a birthday card for a friend based on a photo of them and everyone was like, "Oh my god, that's hilarious. That's amazing.
It looks just like him." How do you do that? It was a it was it was it was very wholesome actually. It was just showing them in a by the hills. >> But so that was like, you know, I I was quite impressed with myself with how well I was able to use it then. And then yeah, throughout the year it's been like, "Oh, this person's generated this. There's you know, everyone I know has posted something to a group chat that they've generated now in some form or another, right?
Whether it's a picture or text or something." >> You've lost your edge. >> Yeah, yeah, exactly. We're falling behind. And obviously I think I've also noticed you know, we both work in the in the industry. So we've been used to the tools for a lot longer and the vernacular of it. And now in this last year, I've noticed like all my friends, anyone I know, you know, past colleagues etc. They are now they're now working with these tools because their companies are you know, you you you can feel that pressure from the companies to we need to be using AI, right?
We need to be >> And you remember like before this year like before 2025, there were like a lot of internal bans, right? Lots of legal IP concerns, companies prohibiting the use of AI cuz they didn't know what it meant from like a regulatory or implication standpoint as well. That completely flipped in 2025. >> Yeah, I agree. It feels like the the kind of the floodgates have fully opened like people are those conversations around I think there is a lot still to be said and there will be a lot to be said on ethical AI, but for a lot of people it felt like in this year the horse has bolted.
There's no time to sit around and argue about the the merits of should we get involved with this? It's like we have to get involved with this because everyone else is, right? And we'll we'll you know, there'll be recriminations later. We'll we'll deal with the fallout in 5, 10, 15 years time when we realize it was all a miss big mistake. But yeah, you can you can see how everyone's companies are saying, you know, use you need to use Copilot, right? Or just in the small tasks and now everyone has this kind of low-level in the same way that I imagine, you know, when we were like I guess two or three years old, people were being introduced to like word processors, right?
And like everyone was getting this you know, how to use the Microsoft Office suite, right? Back in the day. Now it's just another workplace tool, which is quite powerful way of adopting a technology when it's when it's mandated through workforces. Like obviously that's going to that's going to get everyone, you know, involved in using it. But it's obviously increased the education around it as well, which is interesting. Like yeah, you can talk to anyone about it. Um the other thing I've noticed is just how how prevalent AI content is.
Like >> AI slop. >> AI slop. Yeah, yeah. Everything is AI slop and it's you know, it's not just Twitter anymore. It's like every whenever I watch the TV now I like two thirds of the ads I'll be like, was that AI? About one third of them I'll be like, that was definitely AI. Uh you can't really [clears throat] tell. Like >> that game show where you have like is is it cake or something like that where it's like you don't trust anything anymore. Anything could be cake or AI generated at any time.
Do not trust anything. I think yeah, I agree with that. Uh the the social media landscape is awful right now. It's like dystopian where everyone has the exact same format. Everyone's saying things in the exact same way. It's the exact same opinions and insights being copy-pasted across the board and it's really depressing. I think like we've lost individualism. Like I have stopped engaging with social media as much because I one, I'm not getting interested insights from there cuz everything is just AI generated. I can get that from my AI.
I don't need to go on you know, LinkedIn. But also I feel less ready to post on there. Like I feel like you've either got to go down the conform and it's AI so no one has a voice anymore or you go to the opposite route and try and go human and have all these errors and all of a sudden like you you're losing the ability to be creative because you haven't written anything like 6 months. So, yeah, I think that the social media is really dystopian right now.
I think the the workforce stuff like I think it's worth bearing like another minute on that just cuz I think it's so interesting that every like you're saying, the floodgates completely opened, every company rushed because, you know, rather than just the benefits that could be had from it, more and more they saw the risks of not having it. And imagine a lot of that pressure came from the big AI companies, came from the consulting firms that are in the pockets of the big AI companies saying if you don't adopt this, you're all going to go out of date in a year's time.
But these changes take a long time to really become core workflow stuff. I think like to adopt AI into a company of like 20,000 people or some of these huge companies, it takes months if not years for those workflows to become common practices. I think like you see read a lot in the papers and some who reads papers anymore? You read a lot in about like a lot of companies that have tried to adopt AI strategies haven't seen the types of returns and productivity and output they expected.
But like I said, these take months and years to kind of really come through. And also you would then see like, okay, they were using Copilot like the first version or something like that and like who is using Copilot? Like it's just that is not like cutting-edge frontier stuff. So, I think like as companies become more in tune with with the solutions as they can take more risks and move to more frontier models as well, they're really going to start to see like the productivity gains that they're expecting.
But individuals are loving this stuff like developers, analysts, all these people that had to pull from many resources, you know, condense things, turn it into an output, like they are seeing huge benefits from this stuff on an individual local level. >> Yeah, I think that that's a really important point is about the frontier models as well that I think was a huge feature of 2025 is that okay, the last couple of years you've had the big names at OpenAI, Anthropic, Google duking it out trying be, you know, the best.
But, I feel like this year that it the fight got much more intense with how, you know, the the the release trying to undercut on price, and like the actual leaps and bounds each model was making, you know, I think the most recent one really that impacted me was with like Gemini 3. And when that came out, that kind of turned everyone's heads and was like, "Oh, this can kind of do actually a lot of things better than all the other models." And I think the key thing when it as it relates to the work the workforce kind of adoption that we're talking about is that things like the Gemini 3 is a step change from what, you know, it was before that unlocks like a different problem space.
Like, it's like, "Oh, I was trying say I am, you know, February of this year, I was trying to solve a problem using LLMs. And based on the capability of LLMs back then, you'd end up hitting a brick wall quite quickly, potentially. If it's a complex task, it might get you some of the way, but then if it's like a very very, you know, complex coding task or reasoning task you might struggle, and you're going to be like, "Well, actually, I'm not sure I need to use LLMs that much in this workflow." Now, as of, you know, the January 2026 it is completely different.
Like, they could that wall that you hit is much much further down the line. I think that is letting a lot of companies and individuals, you know, Vibe Coders, but, you know, companies in particular say, "Actually, it is worth investing in these tools because they can get you so much further. They can help you solve problems that you actually couldn't solve 6 months ago or even like 2 or 3 months ago, right?" Um when we I think that's kind of what I'm trying to articulate is that we went from having models that there was kind of like you could dip your toe into and [clears throat] still be unconvinced about should I should I invest in in using this to solve X problem?
And then every day, every month, with the release of new frontier models that that objection just starts getting decreased, right? And you just say yes, actually it is definitely worth it. And I know I'm going to get improvements like this going forward as well, hopefully. So. >> And there's like that quote from Sam Altman, I'm going to butcher it, but it was like there were two types of AI startups that he's seen right now, all of which are gone now with OpenAI and their constant updates.
But one was the the companies that solve a problem and assume that AI will get much smarter, and the one that think AI is stagnating right now and just solve the problem as is right now. And I like the former one is going to be where they make money. And I I totally agree in that it's such an investment now to have workflows and maybe replace or enhance roles with AI because those models will only get smarter, cheaper, faster. Whereas like when you invest in an individual at the end of the day, like it's fine this dynamic, but they might leave in a year's time and you have to start again and you take a step back and like you might not always have that that kind of intelligence being built upon, whereas models will get smarter, those knowledge bases will still be there, the flows will still be there, they will just get better.
And I think the hyper competition has been extremely exciting. Like it's one of those bonkers things where a new version is released every couple of weeks, so it's impossible to keep up. And then you look at each version, they do a benchmark of a hundred different tests, and they always come top on all of them, and then OpenAI do something have a different benchmark and they're top of all of them. So it's impossible to know which is best. But I think most people don't care about what is best.
I think that they are converging. I think the difference between Claude, GPT, and OpenAI and Gemini is close enough that it's quite hard to tell the difference. I think some of them have edge cases where like Gemini is really good at doing like big context, lots of research, like maybe more scientific problems. But I think most people care about style, and like there's there's a stickiness as well. Like the thing I like about OpenAI is that I have like um message histories and threads of like months long where I've given it all the context, all the instructions, and you're just like it knows what you want from that and the personalization is there.
There's a lot it would take for a lot for me to move away from that to then push some of those threads into say a Gemini or Claude. So yeah, I think they are getting stickiness. My fear with these things is right now it's great. The hyper competition between these big three means that none of them can inflate prices. They always have to stay hyper competitive pushing creatures constantly which is nice for a user. I don't want one of them to win and then to become like an Uber of the world and hike prices and dominate become monopoly.
I think the situation room right now is kind of like a golden era where there's lots of money coming in. None of them really make much profit from these things but they all have to be hyper competitive. Will that last forever? I don't know. But some of the effects we've seen is that the inference costs for similar types of questions have gone down by a factor of one 10 times over the last year and which is >> Yeah, have you are you familiar with the term enshittification, Alec?
>> Say that again? >> with the term? >> Enshittification. No, tell me what it >> You should yeah, Google it after but it I mean it's I won't but I won't butcher the whole concept. Maybe we'll get him on in the new year to explain. But there's this guy who's basically written a book recently and is doing the with a very compelling argument that it's you know when we talk about web 2 and all the problems of web 2 and web 1 was better, web 3 is kind of was trying to improve on that.
And it's a way of articulating actually all of these kind of industries, especially in technology, over time they get enshittified. I they get worse for the end user because of the inherent incentive dynamics in Um so like initially like you're saying with Uber, initially services prioritizing users, then it prioritizes corporate customers and then eventually it prioritizes the company selling it ultimately, right? And everyone else loses. And I think we're in that as you said, the golden age of AI where because of the land grab they need because of that they're in the blitz scale phase.
>> Where they try to become the category leader, right? So, we're we are in that that honeymoon period for AI. Um when it goes down the incentivization pathway, and I encourage you to look this up uh if you're listening to this, but yeah, I think I think that that's that's something we will see. Also, >> I'll jump on that really quickly, Jack. I just want to say that you're already kind of starting to see where like right now, I think a lot of the models, they kind of they're unbiased in the sense that they will extract things that kind of weight it based on the training, but like they've you know, they're already talking and opening I'm already talking about prioritizing ad-based sources and like moving down the Google search route where I ask it a question and it gives me like a top five sources for that, and maybe three of them will be paid, and all of a sudden like one of the things that you I like about you know, OpenAI is
that when I look at something or ask a question, it gives me many sources, it kind of treats them without an obvious bias anyway, but if all of a sudden it prioritizes paid advertis- advertisements over, you know, non-paid, that neutrality comes out of it, and that again, that's really going to screw up the experience as a user. I I don't like that. But, that's the problem with, you know, growth. You have to grow, you have to make money. At the end of the day, all they care about is their >> And advertising seems to come for everything eventually, right?
You know, there's already yeah, as you said, like all these indications that they're going to go the same way, and and advertising is just going to move from What do they What do they call it? Um like LLM optimization instead of SEO. >> It's like LLMO or something now. >> Where they're like, how do you optimize your content to be indexed and surfaced by LLMs? Because people aren't Googling anymore, right? Well, actually, I mean, one of my behavioral changes is that I I Google stuff still, but I tend to just use the AI summary, right?
Because for most things I've ever Googled, it's just it gives you what the what I was looking for and a bit more, right? And I don't tend to need that much more than that. Um so, yeah, that that that's one of them. And I guess we will you'll we'll talk more about it in the next episode, the kind of implications of the business models around AI. That was another big theme. There's a lot of talk about, well, when are they going to make money? And hence, you know, potential advertising in the future.
Like, are they when are they going to make profit, I should say. Because they are they're making revenue, a lot of them, like OpenAI is making revenue, but they're not making profit yet. They're burning through trillions of dollars um in like capex and opex. And yeah, it's like an unsolved open question is to, you know, can they become profitable in the future? Um one thing maybe I just wanted to touch on to end my thoughts on the year is on on AI is just noticing it's the first year I've noticed actual behavioral change, like beyond it myself, beyond just, okay, I'm using AI tools, right?
That's that's a behavior change, but it's you know, it's it's it's not it's not it's not one that is it's an explicit one, right? It's like I'm making a choice to do this thing. But I've noticed more and more, like, it's affecting my behavior in other things, right? So, as I said, the way I use Google now is different. I will use it and just I won't ever And again, I'm using Google the same, but I'm I'm doing something different after. I'm just using the AI summary.
There's other stuff where, as you mentioned earlier with like, you can tell when something's written with AI. Like, you know, people sometimes try and mask it, but usually they don't. Um and you know, there are some great videos highlighting the the telltale signs, like that it's not X, it's Y. >> Oh, that's so much. >> This is not a something, it's a >> I hate that so much. The long hyphens and all this stuff. >> Yeah. Yeah. And I've just noticed myself, like, completely switching off from anything that has that.
And like, even if the argument's correct, I I almost discount it and I get angry when I read someone posting, especially if it's like from an an individual's account, if I see those patterns in like an individual's, you know, personal post on Twitter or LinkedIn or something, it actually makes me angry and it makes me discount that and I wonder if you know, we're changing our we're changing our perception filters, right? For like what we're picking up from social media, which is interesting. >> I I definitely agree.
I think like one of the interesting things because we obviously play around with a lot of models in my like day-to-day, um is that that so much money has been spent on these models and the people training these large language models to optimize like for readability and to optimize for human digestibility. So all this it's not X, it's Y. That is well, was anyway 2 to 3 years ago, the best way to get someone's attention and to clarify a point. Like so much research and money went into that.
But like you're saying, like we're starting to get to a point now where like the models are starting to change behavior and trying to starting to change how we perceive things and change preferences because we know that's AI generated. I think if it was just coming from the model itself and you're asking a question it did that, you wouldn't care and that would be fine. But it's the fact that people pretend that it's their own personal opinion and they've written something fresh and novel and and insightful and it being AI that really pisses people off, not the actual form >> Well, I I partly agree with that.
I think that is that's a good observation and I do agree with it. I But I think what's triggering me more is that it's so ubiquitous. It is like it's like >> It's the same. >> I You should watch um Pluribus, the new >> uh Vince Gilligan show on Apple Apple TV is really, really good. But it reminds me of that show and I won't give anything away there really, but everyone is parroting the same style and it's like >> we're losing we're losing the art of communication already, very quickly.
on in a digital realm, definitely. But then because we've become for the last 15 years so dependent on digital communication and we do less and less of it in person. Then, when I start hearing these things in person, like I haven't really heard it yet, but when people come up to me in the street and they start saying, "It's not X, it's Y." I think then I'm going to actually lose my brain. I'm I'm going to just I'm going to start crying. >> That's funny. That's the prediction for 2026.
Jack starts crying when someone says it's not X, it's Y. >> when someone says, "It's not X, it's Y." >> No, but I think like that that's a good recap of where we're at. I think for the year ahead, like my prediction, I think maybe it's an obvious one, is the way that most people use the AI right now is still quite shallow. It's very much like question response, Google something, get a quick TLDR version of many sources in one go, you know, write this in a different way, change the tone, check all this kind of stuff, fairly simple.
I think that we're going to move from AI answers questions more to more AI handles processes. And right now, we don't really use AI to make decisions. I think most people don't anyway. I think we're going to move more and more to AI making decisions and just giving you the final output. So, rather than you reviewing each step, letting AI handle entire workflow, it gives you the result and you just sanity check it. I think like some of the tools that I'm typically using, really recommend that a lot of people use, are agent frameworks like NAN, LangGraph, AutoGen, like these are really epic to kind of have AI working within itself, running for minutes if not hours to handle a request and to produce some dialed output at the end.
Like you're seeing huge growth in these systems. Like it can filter, it can prioritize, it can follow up, it can work out what you need to see, what you don't need to see. I think that the reason this stuff is really going to boom in the next year is, like we said, like AI is starting to become cheap enough and ubiquitous enough that you can run it continuously, not just like question and answer step-by-step, you know, episodically. Um the context is getting much bigger, so it can really handle like hours of processing, hours of responses, and won't go many, many emails, many, many documents.
And the final thing is like integration. Like tools like NAN, they can integrate with almost any software. Your emails, your Gmail accounts, your Google Drives, your social media, all these things. But yeah, I think there's going to be a massive move from more chatbot style, which you've seen in 2025, search engines, questions, answers, to handling entire workflows, handling entire processes, using its own judgment, and you're just seeing the final results. >> Yeah, I think that's a that's a solid prediction. I think also one that is based on a little bit of what we've seen in 25.
You know, I something that comes to mind is Claude Code here. When you if you use Claude Code, you know, to develop, you know, software, you have like options. You can be like plan mode, where you're just editing any code. Then you can have, you know, ask me to review mode, and then you can just have accept edits, where you just you just don't really review, and you let it do, you know, what it does. And I think people nine out of 10 times are just like, yeah, accept edits.
I don't I don't [laughter] care. You know more than me at this point. Um and I think, convenience is always king, and people just getting to that the answer, the the result as fast as possible is definitely the way things are going to go. I don't think that's necessarily a good thing whatsoever, but I think it's a fair prediction. Um and yeah, like what systems get built around validating outputs, uh you know, the eval systems you have to see like, well, okay, AI did all this, but how do I know it did it right, or, you know, do I even know how do I even know the criteria?
Because if you're going like uh you know, out on a limb building something new, you don't even know how to evaluate something necessarily. So, I think, yeah, that that's an interesting one. Um for me, I don't I think that then main my name prediction is I I do think we are going to hit a point in the next year where distinguishability is a problem. I think up until now uh you know, GenAI has been distinguishable enough. As I said, like when I'm watching TV, I can nine out of 10 times I think confidently say what is AI generated and what isn't.
I think that's that's higher than most people because cuz you and I are so used to it. I did get fooled by one today. So, I was like, "Oh, here we go. It's getting better. It's >> Well, no. What was the What was the >> It was actually one It was uh posted by one of the big Bitcoin Twitter accounts. >> It was with like It was like a girl on a saying she had 90,000 Bitcoin and it was meant to be like an old home movie thing.
It was very convincing. And then someone in the And then someone in the comments pointed out that she's not actually holding a mouse. Her hand's just like So, I was like, "Ah, probably should have seen that." But, the rest of it was so so good that I think yeah, I think people are going to get tricked much more easily, I think, in the next year. Um which has a whole has a a bunch a bunch of bad implications that we've covered, right? We don't think we need flog that that dead horse at the minute, >> The spam is coming.
>> Yeah, and again, if we I always use the Will Smith eating his pasta. I think the version of that next year, like by the end of next year, I think if you did uh scientific tests with big sample sizes, I think people are going to approach that 50% accuracy. Like, can they predict it one in one in two? Probably, you know, that's about where we're going to land, I think, at this >> Scary because who knows what's authentic, what's really more, but also exciting because like the barrier to entry to say make movies, make content is so low.
For like $20, you can then become a movie maker. Like it's bonkers. It's really going to just completely change the world. >> What do they say, Alec? If everyone's a movie maker, no one is. >> Just classic line. >> Um okay, next one. This is a fairly short one cuz I think we've talked about this a fair bit is wearables, smart glasses. Like smart glasses with we've done a few episodes on this. You know I'm a massive proponent of these. >> I don't have them on right now, but I do have them over here just in case.
>> They've been around for, you know, many, many years. 10 years. I think like like there were still smart glasses being manufactured. Like like less than a million units a year mostly for like testing prototypes, like just research and stuff like that, but no real big consumer adoption. I think 2025 was the year that that changed. Like it's not that they went mainstream, but it kind of proved the viability and like it was Meta that did that. Like Meta started shipping or continued shipping consumer smart glasses.
They really started to focus on wearability, audio, camera, AI integration. Like whereas like previous versions had focused much more on the tech, what's possible, research, this focused on okay, making an epic pair of glasses that looks like, you know, stylish glasses, will fit on people's faces, fit into their lives, fit into their style, their fashion choices, whatever, but has smart features. And like I say, they weren't full AI headsets. They weren't absolutely revolutionizing the world, but they were lightweight, they were always-on devices, and they were things that people can actually wear day-to-day and get a lot of cool features out of.
>> Yeah, I think as you know, I'm I'm I'm not sold yet on what's there. >> I would like to clip of you saying you were sold when we did the episode and you tried them on. >> I think I said I was getting close to sold. I'm sold that there is potential there. Um I think we're still a a little way off, but I think yeah, for me like this was the year where I could I could see where it's going. And I can see some really good use cases for it.
Like there are definitely benefits of this new wearable category. And I'm excited just from a you know, a possibility standpoint of what can they what can they engineer from it? But yeah, I think yeah, will they will they do it in 2026? Is another question. I think we'll see some of the interesting improvement by the end of next year. >> Huge prediction. I think we'll see some interesting improvements. >> My prediction is that you will buy brings out next year and I won't. I'll probably be more like 27, 28.
>> I think that so like in 2025 Meta sold, I think they reckon about 3 million headset 3 million devices, 3 million pairs of smart glasses. I think the predictions are upwards of 5 to 10 million in 2026. So they're starting to really ramp up production. Like I said, I think the real game-changer is the features right now are simple. Like they do what your you know, your phone does but worse. It like can you know, do basic voice assistant, can play music. It merges some technologies into one.
Obviously it's a bit more seamless. You can take a photo. It's hands-free like but that it's not so much of a game-changer that you're going not going to buy a phone. I think that there's there's a few interesting points here. One is proving the feasibility, proving they can make these lightweight enough, long enough battery that is actually usable. Like that was one thing that Meta did. I think the other thing is related to our previous conversation is AI is the unlock in 2026. Like AI inference is becoming cheaper, faster, increasingly being pushed to the device which wasn't really possible.
And when latency is dropping from like seconds to to milliseconds, you start to have this game-changer in smart glasses that is not possible on mobile devices. Like the idea of context and persistence improving because you have visual, audio, and memory on a device. Like this makes smart glasses genuinely more useful than a phone cuz they're on my face anyway. At any moment I can say, "Do something. Give me advice. Look at this engine. Tell me what's wrong. Look at these ingredients. Tell me what I can make.
Schedule something based on what you can see." And that is a real switch from what we have right now, which is a phone. I take it out. There's more steps involved. Like everything in tech and UX about minimizing the number of steps. And smart glasses right now don't do that. But once we really have the AI unlock and having models that are super smart, super contextually aware, I really think it's going to revolutionize the game. I think we're going to start to see that in 2026, where you can do things on smart glasses with these new AI models coming out you can't do on a smartphone.
>> Yeah, I think that's a that's a really good point. I think you're you're you're right. There will be definitely things we can't we can do now with these wearables that we couldn't do, you know, the previous categories of wearables or iPhones or phones generally. Um you know, I think that that's the stuff that was selling me I think when when when you showed me them in person I was like, "Oh yeah, I can see it now." Like I don't know exactly what all those problems will be, but even things like, you know, the the note taking, for example, or like being more present in conversations.
I think that kind of stuff is interesting. Recording things as well. Like I don't I don't think there's a huge amount I would use it for in my day-to-day, but you know, there I can also see there being times in your life when that is actually useful and you don't you want your hands-free. Uh maybe you're going on a trip somewhere. Like that that is quite cool. It's why people have bought bought GoPros for years, except this is going to be, you know, as you said, with AI much more much more enabled to do more powerful things, right?
That your your your kind of simple cameras couldn't used to do. So, yeah, I think I think that's a a really good prediction. Um, and then should we wrap things up the one that you love the most, even more than than the than [snorts] the AI glasses, is the the humanoid robots. So, I say humanoid because in 2025, a lot of these humanoid robots were were people remote operating them or in suits as as far as I gather, but >> You say a lot. Like, I think 99% of the content that people saw online related to humanoid robots was AI generated.
The other 1% was people in suits. And then 0% that people didn't see was actually real. Like, again, humanoid robots have been around for 10, 15, 20 plus years. Actually longer. I think it's like 1995. And the concept of them, we're not going to go into that cuz I'll bore you. >> when you were born, Adam? >> That's what Yeah, exactly. Me, I was the first. And the concept of them has been around for hundreds if not thousands of years. Not going to go into that.
We did a whole three-parter on it. What happened in 2025 was the a lot of these moved from lab-based environments, super controlled, to actual industrial application initial testing. Like, some of the things I really want to talk about is BMW started to run one of the first live BMW factory test where they put a actual humanoid robots on the product line, on the assembly line, and tested them over like 10-hour shifts. They recorded like thousands of hours of run time because they're really at the level now where they feel confident that these can actually start to replace people.
Scary, optimistic, exciting, whatever, it doesn't matter. The point is they feel comfortable comfortable enough now to start doing large-scale testing to say, "Okay, can we replace people? What are the results we're getting?" Like, another big stat that I want to pull out here is that, um, Tesla Optimus, they reckon next year they are going to manufacture 100,000 humanoid robots. So, 100,000 of that. It's not to say that this is going to be in everyone's home next year. That is definitely not the case. Like I think a lot of these a lot of them like cost around 40 to 80,000 dollars.
Like a lot of money. It's obviously not going to be something you put into, you know, every home in America. But when you start to compare that with your typical factory worker, but this thing goes 24/7, costs a lot less, um can be deployed for 10 years. The cost really makes sense on assembly lines in industrial units. That is what we're going to start to see in 2026. We're going to start to see these deployed in warehouses, in assembly lines, on manufacturing lines, and then hopefully we'll start to see costs go down.
We start to see applications go up. We start to see them being used in volatile, crazy, hazardous, dynamic environments like houses, like on the streets. And I'm hoping we get that in 2027, maybe. But let's see. >> Yeah, I think I think I'm with you there. Obviously, you know more about this than me, but I think the the domestic humanoid robots, right? The ones like we had a few announcements about this year. There was that, you know, company in San Francisco saying we've got these humanoid robots that actually turn out to be remote operated for now by someone who I don't know, with a remote control.
Um I think that kind of stuff is clearly further off and um you know, likely should be. And I I think yeah, the BMW testing is really interesting. I think that's definitely what we're going to see more of next year. It's It's It's natural right? Like major mechanized industries typically are the the pioneers in new technologies, right? And I think it it makes sense that you're replacing human parts of the supply chain, especially if they're dangerous or difficult or you know, they're risk to human life and things.
I think that's that's that's a really nice, clear value proposition for humanoid robots that doesn't make you feel queasy and think about, you know, having one of these in your home, right? I think that will at least kick get people comfortable with the idea. Not that they have to, right? This stuff can just happen in the black box as far as these manufacturing facilities are concerned. But it's a great way of like again getting data, the testing phase, getting data on how humanoid robots can work and how to improve them generally, even if it even if it only ever gets as far as industrial applications.
You know, I do think eventually we will have them in a more domestic setting, but I think this will be like a year of consolidation of all that that testing phase too. >> Exactly. And to bring this full circle, like it's the reason that stable coins succeeded so much more than crypto did this year. It's the micro enhancements and like improving to things to say make them 24/7. Don't overhaul an entire system. So much of the world is made for humanoids, so much of assembly lines, industrial kind of setups is made for humans right now.
You don't want to change all that to move it to like a conveyor belt style thing or to make, you know, you kind of edit it to make it usable by an industrial robotic arm or something like this. Humanoid robots fit into existing workflows. If you can just have them slightly cheaper, slightly faster, 24/7 uptime, they're going to be bought in huge volumes. I think that's what we're going to start to see in the upcoming year. Exciting. Hopefully we'll get less people in suits and more actual robots being deployed out there.
>> Yeah, well, I think that's a it's a good it's a good way to open the I think we've got actually quite a few fairly positive predictions, you know, coming up. I think we think broadly that, you know, crypto might have a a slow, quiet, consolidatory year, but it's going to be an interesting data point for us. AI, [snorts] well, I I was a bit more pessimistic about AI. I think we're going to see this the slop become less sloppy and it's going to be as ubiquitous, but uh but harder to to distinguish.
Um and yeah, I think AI and robotics is going to be an interesting one to watch next year. But yeah, I'm excited. I'm excited for another year in in technology. I think it's it's becoming For me, it feels like it's becoming harder and harder to predict how we'll feel at the end of the next year. So, I just hope we're happy. >> The rate of change of technology keeps accelerating, so it's getting harder and harder to predict anything. But, this is our second, you know, full year, I guess.
The second year start of the year we've done 2026 >> and then yeah. Yeah. >> There we go. Like, very excited for 2026. Let's see if any of our predictions hold up. And yeah, thank you for listening wherever you may be, and join us next time to untangle a little more of Web3. And I hope you had a good Christmas and New Years. >> Thanks for tuning in to another episode of Untangling Web3. >> Don't forget to like and subscribe and to follow us on social media.
It really does help us out. >> The views we express here are our own and do not reflect the views of our employers. And with that, we'll see you next time.
Topics: Blockchain & Crypto · Business & Society