The media is awash with speculation that we’re in an AI bubble – but are we?
Global Futurist and AI Expert Andrew Grill was back on Radio Bristol with John Darvall to talk about the state of artificial intelligence at the end of 2025. We explored the fascinating parallels between today’s AI excitement and past tech booms, like the dot-com era.
AI is on everyone’s mind, with enormous investments and big promises, but also questions about sustainable value. As I shared, there are still real challenges to broad adoption—training, budget constraints, data quality, and the need for organisations to rethink their processes, not just layer new tech on old systems.
It’s clear that the real breakthrough will come not just from the technology itself, but from changing how we work and think. 2026 will be an exciting year for those ready to embrace these changes.
Interview Transcript
John Darvall
Let’s talk AI, remember the space race America and the Soviet Union battling it out. What’s going to happen now with the likes of AI and Google and Microsoft and meta and apple and many others who are all having a crack at this, and then, of course, we’re talking about the billions and trillions of dollars that it might create, or not, as the case may be. Is it a bubble? Could it go pop? Let’s try and understand this with Andrew grill, who’s an AI expert and author of digitally curious Andrew, morning again, good to have you back with us. John, good to talk to you again. So here we are at the end of 2025, I would think most people at the beginning of this year would have probably thought AI was a movie, not a not a thing. It’s a thing. Is it a bubble?
Andrew Grill
It’s a really good question. Well, I’ve been around for a while. I was around during the.com bubble and the subsequent crash. And the signs are there. The first sign, as you mentioned, these massive valuation open AI, the makers of chat, GPT, is valued at $500 million there are enormous amounts of money being spent and and pledged from all those companies you mentioned. There’s a strange things happening with circular financing. One company’s investing in another, and then they’re buying products from them. And the major problem, I think, is that supply massively outstrips demand. We don’t need all these AI products just yet.
John Darvall
There are lots of them. And you talk about the fact that open AI is worth this. I’ve used it a couple of times. How does it make money? Because I understand how, if you Google something and you click on the link, the click is the generative for the advertising. But you ask chat GPT a question, you get an answer. You don’t get links. You don’t get clicks. It tells you. So how is it making money?
Andrew Grill
Well, a couple of ways. Yes, you can use the free version. I use the paid version. It gives you a bit more. They also licence their technology to other companies. So a lot of the generative AI tools you’re using are actually powered by open AI products. They’re making money there, and they’re making billions of dollars, but they’re not making the sort of money they’re saying they’re going to spend $1 trillion in the next five years, they won’t make that back yet. So that’s, that’s the issue. So they’re making good money, but not the sort of money they need to to support all these, these investments. So that gets to the
John Darvall
heart of something quite important here, Andrew, doesn’t it is that it is making money, but they have this technology. But once you’ve sold it to all of these companies, you’ve sold it to all of these companies, you can’t sell it again. I mean, you can upgrade it. So is that where the bubble lies, that once it’s there, it’s there, and if people don’t use it properly, or companies have it, that’s it.
Andrew Grill
It’s grown. Well, here’s the rub at the moment, in that people are using it, but they’re becoming this interested because it’s not doing exactly what they want. I heard earlier, one of your colleagues talking about it, giving silly answers. I think there are actually four things that are stopping adoption. One is training. Not everyone knows what it can do. Probably about 80 different events this year. I’ve spoken at where people have gone. I didn’t know I could do that. The next is budgets. People not putting aside money to spend proper money on this into next year. Data is a problem because it’s not the right format. And here’s the kicker, process, people are not prepared to change the way they get work done in an AI world just yet, and that’s why the demand is limited at the moment.
John Darvall
And is that actually where we’re probably going to see the change next year is that we have to change how we do something if we’re going to embrace AI, rather than just sticking AI on top of it exactly.
Andrew Grill
If you stick AI on top of a broken process, you get a very broken process. I’m actually looking at how I change what I’m saying to companies next year, and it’ll be less about the technology, because it’s there and it works. Be more about how do you actually use it in your day to day work? I imagine most of your listeners like you have played with it a couple of times, but you don’t use it every day, and unless you use it every day, like you do your mobile phone, the money just isn’t there.
John Darvall
Andrew, great to have you back with us. I have a funny feeling we’ll talk again next year on this because it’s not going to go away anytime soon, but I wish you a Merry Christmas. Thank you for your time this morning. Andrew Grill, AI expert, author of digitally curious.
