AI in Offshore Wind Forecasting: The Human Edge

Intro (00:00):

Welcome to PowerPod, the renewable energy podcast from TGS 4C, a leading provider of offshore wind intelligence. In each episode, we explore the key developments shaping the offshore wind market and bring clarity to the biggest trends shaping the industry and the wider energy transition.

Dan Wiseman (00:23):

Hello and welcome to another episode of PowerPod with me, your host, Dan Wiseman, market analyst here at TGS 4C. Today I'm joined by our manager here on the market research team, Jamie Bernthal Hooker, and returning guest Chris Ayers. Today we're going to be talking about something that is talked about a lot, I guess, in various different contexts, and this is the matter of AI and its place in forecasting and analysis when it comes to offshore wind. So, let's start by introducing ourselves. Jamie, if you'd like to just introduce yourself and let us know what you do here at TGS 4C?

Jamie Bernthal Hooker (00:58):

Hello, I'm Jamie Bernthal Hooker. I've been on a couple of episodes of PowerPod before, back in the olden days, and I manage the market research team at TGS 4C.

Chris Ayers (01:10):

Hello, thanks for having me back, Dan, for my second podcast in three. My name's Chris Ayers, and I'm the European Union analyst for offshore wind, grid reform, and interconnection, and I'm also the analyst for Nordic countries and Ireland as well.

Dan Wiseman (01:31):

Excellent. Now, obviously a major part of our job is market analysis. It's kind of our bread and butter on the market research team. So, as some of our listeners may know, we release a quarterly market overview report, which is coming out next month for our latest report, where we delve into our individual markets and we talk about forecasts and what's happening in individual nations and their offshore wind goals. Now, Jamie, I think we've entered kind of a day and age where it's very easy if you have a question or if you want to know something, just to jump online and pop it into ChatGPT or Grok or Claude or whatever AI software you choose to use. What do you feel we do as a team that maybe AI can't do in terms of our analysis and our research?

Jamie Bernthal Hooker (02:35):

Oh, that is a very wide question. There's a lot I could talk about there. What I think we have that I'm so proud of is real interest in the market. That sounds weird, but we're experts in our markets because we're devoted to looking at all the holistic aspects of these parts of the world and the industry. And so, our reports are assembled by people whose jobs are really to become fully immersed in offshore wind in a part of the world or in a specific technology. And we bring together that kind of holistic expertise, that knowledge and interest in the market that brings in not just measurable data, and we have the best data in the world, but also that context and that knowledge of what makes it important, what is important, how it works together. And so when it comes to forecasting, we're building on that very solid and firm and important data alongside a lot of opinions that are going on in the world, in the space, in the internet. We're able to use our insights and our expertise and our genuine passion, which I think is so important and so hard to imitate passion, to build credible forecasts.

Dan Wiseman (04:36):

Yeah, I mean I completely agree. Like you said, having that passion from our analysts is super important and I think every one of us has that. Myself, I deal with North America, so I deal with the US and Canada quite a lot, and they're just fascinating and it's very easy to get drawn in to this microcosm of offshore wind in each of these markets. And I think it's two different questions when it comes to either asking AI or asking a real person. And it's not so much "what happened?", more so "so what?". You know, and the thing that we do here that I feel is we bring a lot more context to information than maybe you can get from AI. For example, if you ask AI a question of like "how much is this wind farm looking to install by X date?", it's like "oh, 1 gigawatt by 2030". We'll have a different idea because although that's printed on paper somewhere, that may not be the backstory. And I guess, Chris, in our markets as analysts, we see it every day where although there is a target in place and that's down on paper, it's not always the case that what that says is actually what's going to happen, is it?

Chris Ayers (05:30):

Yeah, no, exactly. So what a kind of living, breathing, genuine, metabolizing analyst can do that I don't think AI can substitute, or even come close to substituting at this moment, is really delving into those long-term structural patterns and changes for their particular market. You know, I have my five and you have three or four, isn't that right? Four markets, and you have now three. We all have our own markets that we devote hours and days and weeks and blood, sweat, and tears to. And we gain that understanding of what are the wider geopolitical and political contexts, what is the legislative agenda that's underpinned not just offshore wind, but all the other relevant areas of policy, whether that be energy, of course, infrastructure, economic growth.

There's obviously the economic context, there's the regulatory environment—how feasible is it in the regulatory environment in our markets to actually come close to reaching the targets that countries, the governments set themselves. And all, and have, then the synergies between them, of course. And so that really, really well-rounded understanding, that intimate understanding of our own individual markets is so important. So, I don't think, I think AI is very, you know, it has its obviously its use, it can retrieve information and then afterwards ask you if there's anything else they can help with. But at the end of the day, it's that intimate understanding that I really think separates us, not just from AI, but from our, from other data intelligence providers.

Dan Wiseman (07:27):

Yeah, and I think I mean, firstly, I think it's another thing that I've noticed with that we are able to do that AI possibly can't is that we can detect the subtleties. Like if we notice a developer comes out and says something, you're like, "hold on, they're not usually like this, or they wouldn't disclose this unless something else is happening." So then we end up kind of going down another rabbit hole of information and trying to find out. So, I think AI finds it very difficult to pick up on contexts of things. But like you were saying, I mean AI does have a place. It's very good for analyzing massive amounts of data. It's something that we're using here at TGS 4C, especially with an upcoming product that we have, where we use AI to try to get a massive amount of information in one go, because some things are just too big for us analysts who have our markets to be able to concentrate on. So, another thing, I think, Jamie, we've seen a lot of when we've used AI in maybe not work perspectives, but kind of in other ways is the fact that it doesn't always give you the answer that you want. It kind of just gives you the answer to the question that you ask, which leads to a lot of misinformation at times, doesn't it?

Jamie Bernthal Hooker (08:48):

Yeah. Well, there's lots of different types of AI, I guess. So if we're talking about a generative or if we're talking about LLMs and a kind of chatbot interface, yeah, you get that range of from helpful but uninformed to confidently wrong. And we all know people like that too! But then you can have other forms of AI that are super good for dealing with data in many ways. And we'd be absolutely bananas not to be thinking about those forms of AI as everyone in this space is. But what we can't be is slavishly devoted to them. And something maybe weird about me is that I love detective fiction. I write detective fiction, I read detective fiction, my colleagues are sick of hearing about Agatha Christie! And one thing, so I always have treated the process of research as being like a detective in a detective novel, and that just helps me get through the day.

But one thing that I'm noticing with the current talk around AI is it reminds me of when forensic science really took off in detection, and criminal profiling, and that ability to use data to actually pinpoint with accuracy things about what had happened at a crime scene, what the criminal had done. And people were saying, "Well, there's no need for the police anymore, there's no need for detectives anymore, because we've now got data which tells us things and can't be wrong!" And, of course, it hasn't worked like that at all. This makes detectives' jobs different and gives them other things to do and to interpret. And it's a similar thing with any technological development. So, I relate everything back to crime fiction because I love crime fiction, but I also love offshore wind. And what we're seeing now is a lot of exciting opportunities and possibilities being opened up, but I think some of the discussions that you hear around of life as we know it changing completely and never appearing again are greatly overstated. I'm very excited about whatever we can do in this amazing industry going forward.

Dan Wiseman (11:23):

Yeah, absolutely. And it is very much a time and a place sort of approach with AI. And especially when it comes to things like kind of longer forecasts and forecasts further into the future. Because obviously we as analysts, when we release our overview reports, we kind of have our forecasts for certain milestones, like 2030, 35, and 40. But with the analysis that we're able to do, and even small things like having individual contacts within certain government bodies in our markets, means that we're able to predict much, much further than that, which then becomes very difficult again for AI to do because it's working based on released numbers. And I think my, I guess like my feelings on AI and its impact in what we do as analysts is kind of it's very good at telling you what the market is saying, but it's not telling them what the market is trying to do and what the market is telling us. So, Chris, in terms of forecasting and how you predict things happening kind of further on down the line, what do you feel is like the strengths from your markets to be able to inform you of what the future is like for countries?

Chris Ayers (12:50):

Yeah, I mean, I don't have a crime fiction passion. The closest thing to passion I have would be around pale ale, but I suspect there isn't a long-term analogy to be made there! I think in terms of, you know, I always think back to whenever I've spoken to customers or I'm hearing a speech or from someone in industry or an announcement, so often, at risk of sounding trite, so often the phrase I hear is "but we just don't know" at the start or at the end of that speech or that conversation. And I always go back to that. And I think, like you were touching on, I think what my markets do, having actually having quite a range of different markets as I have had in terms of, you know, I've had Iberian Peninsula and I've had Nordic countries, is it really, really hones in and makes you appreciate just how, A, just how different every market is, how every legislative makeup each market has.

And I think that appreciation and acceptance—well, not acceptance, but that recognition of that—it almost pushes you to delve further and deeper, and with that you can identify trends and patterns in a way that I think a surface level kind of an analysis up to 2030 perhaps may lack, identifying the key the bottlenecks that are unique to every market that an in-depth investigation into that market would allow, allows you to go beyond 2030 to 2035 and 2040. And that harks back to the structural challenges that all markets have. So, what I'm trying to say is I suppose my, the range of markets I have and everyone having, all of our TGS analysts having their own little geographies, their own areas of the world, I'm trying to, that's another better way of putting it...

Jamie Bernthal Hooker (15:34):

Fiefdoms.

Chris Ayers (15:37):

Fiefdoms! Fiefdoms, I like we'll use fiefdoms.

Dan Wiseman (15:39):

Fiefdoms sounds good!

Chris Ayers (15:41):

Everyone has a, almost takes ownership, they almost get a sort of intimate kind of...

Dan Wiseman (15:44):

Yeah, you have like a favorite country.

Chris Ayers (15:47):

You have a favorite country! I don't wish to say that, but yeah, you have a favorite country and you're invested in it, and that's where the passion comes in.

Dan Wiseman (15:50):

And like we were saying earlier, because like when you were speaking with Jamie earlier on in the office and we were talking about when analysts start, they always have a one country that they put all their faith in and like "oh, this country's going to do really good, it's going to be a really good market" and sometimes it doesn't, and like you realize like "oh okay, I need to be a bit more realistic about things like this."

Jamie Bernthal Hooker (16:09):

But if I could, I couldn't agree more with what Chris has been saying. And one of the key things that I think is really important is politicians and political announcements are often very short-term focused, even if they include long-term targets, or the idea that "we're going to do X, Y, Z by 2040", they don't necessarily expect to follow all the way through on that because a political term is typically four to five years, and then there's often another election, not always. And so sometimes what's announced has to be filtered and understood and taken in a certain way, but also in the context of what is and isn't achievable in terms of, very often it's about the investments and infrastructure required to realize a target or an ambition.

So, we can see a very credible looking offshore wind target and a very credible looking auction schedule to go with it, but we need to know are there actually enough ports to support that, are there enough people interested in it, is there the right kind of support mechanism going on there? And nearly always, I think very few instances in the world have everything optimally. And so we're bringing our judgment to that, and I think it's really important to have that long-term holistic view.

Dan Wiseman (17:39):

Yeah, and I agree, and I guess it's a matter of reading between the lines a lot of the time. Because in one of my markets, which I shall not name, on the surface of it, when they go away to have this governmental meeting on offshore wind, it's very easy to assume that they're going to come out of that meeting with an idea of how they're going to progress forward with it. But having known that market and seeing this happen time and time again when plans don't come at the end of it, I know that when they go into have these talks that nothing's going to happen because of it. Whereas if you read it in the media, you're like, "Great, okay, well, they're serious about this, they're going to do something." But it's kind of having that knowledge and that familiarity about your markets to know that even though something's being said, you're like, "Hmm, does that actually mean that's what's going to happen?" I mean, that happens pretty much all the time in the States. Jamie, you were the previous US analyst, so you know that when things are said in America, they're not normally what's going to happen.

Jamie Bernthal Hooker (18:45):

Yeah, but that's so interesting because it's not just "this isn't going to happen like they've said", but then you have to think "what will happen?". And that's where the real fun comes in, the detective work, which, yeah, I'm going to convert everyone to Agatha Christie at some point, but yeah.

Dan Wiseman (19:04):

And it is, yeah, the detective work. Because staying with the US, obviously under the current administration there's been a lot of doom and gloom around the US and a lot of people said "oh, offshore wind is dead in the US" and that's been the major headlines in the media, whereas we know better. Because we know that, okay, just because this is happening doesn't mean that this, this, and this can't happen. And we've seen them carry on in the capacity that they've been able to, and we knew that there were going to be lawsuits, and we knew that the country was going to fight back, and we're seeing it kind of on a daily, well I am because I'm paying attention to it, seeing it on a daily basis that, you know, there is that fight back when everybody else gave up on it. And again, that's something that we're able to do that helps us further into the future, because I know as an analyst for the US, in the next couple of years, dependent on the midterms, dependent on the next administration's party makeup that comes in, it means something completely different and then things can change again. But we also know that just because it's a more favorable political alignment doesn't necessarily mean that it's a more favorable offshore wind party.

Jamie Bernthal Hooker (20:20):

That's it, it's nuance. And headlines aren't enough to inform these very big and expensive decisions that we're dealing with. Can't be informed by headlines and vibes alone! And so, yeah, we often find, don't we, and we found it a lot with the US and with some other markets in the last few years, conversations around projects being canceled. And technically those projects weren't being canceled, they were being shelved or sold or delayed or postponed. And that kind of difference is actually really important, and can be a kind of multi-million pound difference. And so it's really important to bring that nuance. And the other thing is, when looking at targets, looking at how they work, is it a generation target or a leasing target, or is it some other form of target? And sometimes that nuance and that understanding is not centered in discourse. And so it becomes really important to know that there are people looking at it, processing it, understanding it, and bringing serious, granular, and brilliant data to it. And so I'm equally proud of our fantastic data, and also our enthusiasm and detective work.

Dan Wiseman (21:50):

Yeah, I mean it just cuts through the noise, doesn't it? Because I think we've seen it in all of our markets, where there's been a minor setback that's been blown massively out of proportion and we've been there to be like "actually, it's not that bad." And we're able to kind of promote that in our database, in our reports. I know a few of us are relatively active on LinkedIn and we tend to kind of post stuff on there when things kind of come through. But I think another thing really is, is trust, I guess, you know, it's very easy to find information from various sources. I mean, we do it on a daily basis. Our job is to find out information, but it's finding who it is that you can trust. And I think especially here at TGS 4C, we've been around for quite some time now.

We've been releasing these quarterly updates, we release market update reports as well, we're very active on social media, we're very active in media in general. We have analysts who have done interviews with major renewable media sources, and our company is gaining traction. So trust is a big thing in terms of how we do our jobs, because obviously in order for us to be able to put that out, we need to find sources that we can trust as well, don't we?

Chris Ayers (23:18):

No, absolutely. And I think, again, this is what really separates us from, certainly from AI, and from other data market intelligence providers. We have a pretty rich, almost heritage, of contacts, of source of primary sources, of people in the industry, of knowing where to look, and that comes with time and investment that all our analysts put in into understanding where to retrieve information. And that level of trust, and also our analysts I think are really, really good at scrutinizing the data that we come across. Sometimes we do come across headlines that are quite sensationalist, that do not tell the whole story. And I think all of our analysts are very, very good in very deeply interrogating everything we see and making sure we can find, if not one, two, three primary sources that we can scrutinize what we see against.

And I think that over time we've built up that reputation as the go-to place for data market intelligence around offshore wind. And, again, it takes time, but I think it's time well worth spent, and I don't think any amount of generative AI can ever really substitute that. I think, and we have it here in our, you'll be able to see it in our, I was pointing to my temples, we have it in our brains, we have become so familiar with our markets and who to speak to and the key players, developers, and stakeholders in that market, it is invaluable information. And I think everyone at TGS 4C are, and rightly should be, proud of the amount of time and dedication they put into understanding our markets and building that trust that our customers have come to recognize.

Dan Wiseman (25:23):

Yeah, no, absolutely. And it's something else really when you have a customer who comes to you and you're having a conversation about a particular market and they ask, "Well, what do you think?" And you're like, "Well, this is what I think," and they take it on board, and the fact that we're able to just give our opinion on something and people are taking a very kind of pointed notice of it and really take on board how we feel about things is kind of special. And it kind of makes us understand actually, you know, we know what we're talking about, and people know that we know what we're talking about, so they come to us for that information. And like you were saying, I think it's down to the fact that we do scrutinize everything that we see source-wise. And I think, I mean, maybe it's the typical pessimistic Brit! That means that we have to make sure that what we're reading is true. And I think in this day and age of multiple media outlets all reporting on one story, it kind of becomes very Chinese whispers at the end of the day.

So you'll read one story, but then you'll read it somewhere else and you're like "hold on, this isn't particularly correct." I think I spoke to you, it was probably a couple of weeks ago, about some finances for a developer where they were year-on-year, but they didn't really put that in place, so it made it sound like they made a massive profit. So it's that attention to detail, and that clarity, and being concise, and the context that we're able to bring, that obviously makes us as good as we are. And leadership! We can't, can't forget about leadership. I mean, Jamie, you've been a manager for the team for a couple of years now, and I think you've been wholly responsible for the team that we've got at the moment and the product that we've done.

Jamie Bernthal Hooker (27:10):

I'll take it!

Dan Wiseman (27:13):

So I mean, what is it like to manage the market research team? How does it feel?

Jamie Bernthal Hooker (27:18):

It's a dream! Of course it's a dream. It's a huge privilege. I don't want to go on too long about how much I love all this, but as a company, we've been doing this for, it will be 18 years in 2027, which is very cool. So we've been there for a huge chunk of the international offshore wind journey, and I'm excited to see where it goes next.

Dan Wiseman (27:48):

Excellent. Well, I think that kind of brings us to the end. If there's anything else you guys wanted to touch on that maybe we haven't spoken about before?

Chris Ayers (27:58):

Well, of course, I must add for three and a half months here, it's been a dream to be part of the team! So thank you, Jamie.

Dan Wiseman (28:12):

Great. Well, thank you both for joining me today and having a chat about, well, I guess it was very self-indulgent! It was very much about ourselves. It's been wonderful to have a chat with you about quite a big topic, like I was saying earlier, in a variety of different contexts. We can't escape AI, and it's always nice, albeit slightly self-indulgent, to talk about ourselves as a team. But we're very proud of what we do, and we do a very good job, and like Chris said, you know, we should all be proud of what we do. But let us know, actually, your opinions on this. It's a very divisive topic, is AI and forecasting in general. So you'll most likely see this on LinkedIn or our TGS page, so do feel free to get in touch via our socials, and we'd be more than happy to answer any questions or engage in any conversations that you may have. But until then, thank you very much for listening, and I'm sure we'll have much more to talk about next time here at TGS 4C.