Darren Russell, AI enthusiast and former AI development director at Mott MacDonald, says the arrival of generative AI marks the biggest digital transformation he’s ever seen – akin to the arrival of computers into the workplace and with even more potential impact than the invention of the internet. Hold on to your hats is his advice to others.
Russell is “in the profession that I always wanted to be in – civil engineering”, but he has had a lifelong obsession with technology and digital. “I’ve just loved it and been fascinated by it. So I’m lucky to have a foot in both camps,” he says.
Following his studies for a civil engineering degree and then a computer science degree, both at Manchester University, he completed a PhD in Sheffield in geotechnical engineering.
He joined Mott MacDonald in 1992 as a geotechnical engineer, but even then, he was involved with digital technology. “Around a decade later, I became the IT director, where I started the company’s digital transformation. In that time, we built the foundation of our technology platform and set up partnerships with Microsoft, Autodesk, Bentley and others,” he explains.
After 10 years in IT, he took on the role of group chief digital officer and spent the next decade delivering solutions for clients. At the end of 2023, he shifted to focus purely on AI, as the development director in this area. He left Mott MacDonald at the end of 2024.
BIMplus: How has generative AI changed the landscape?
“This is change at least on the scale of the internet, or even the arrival of computers in the workplace. The pace of change since 2022 has been phenomenal and it’s just getting faster.”
Darren Russell: AI has been around for decades and companies in our sector have taken advantage of predictive AI for years. This is fantastic technology that can be used in many ways, for example, to predict the performance of infrastructure (machine learning), identify defects in structures or types of vegetation or animal species (computer vision), or to rapidly model engineering systems such as flood risk (surrogate modelling).
But everything changed with the launch of ChatGPT and the rise of generative AI at the end of 2022. It’s the most transformational technology I’ve seen in the 30-plus years I’ve been working in the sector.
This is change at least on the scale of the internet, or even the arrival of computers in the workplace. The pace of change since 2022 has been phenomenal and it’s just getting faster and faster.
While you can think of predictive AI in much the same way as more traditional engineering software, it’s much easier to think of generative AI as an assistant or co-worker. The author and academic, Professor Ethan Mollick, suggests thinking of the current batch of advanced GenAI models as “an infinitely patient new co-worker who forgets everything you tell them each new conversation, one that comes highly recommended but whose actual abilities are not that clear”. Thinking of GenAI like that allows you to start experimenting with the ways that it can help you in your work and your company.
How are companies implementing AI?
GenAI can do loads of the boring, mundane, repetitive tasks that machines are really good at. That might be summarising or starting off a document for you, or doing research or producing minutes of meetings. It buys you time. And that’s where companies are making first use – freeing up people’s time so they can focus on what they joined their companies to do. Typically companies in our sector are using Microsoft 365 Copilot to realise these productivity improvements and, typically, individuals are saving one or two hours per week.
What about AI’s vast energy consumption?
This is becoming an area of real focus, because much of the work we do as engineers is thinking about sustainability. It’s best practice to measure the carbon spent in using AI solutions and thinking about whether the application is delivering more value in carbon terms than it’s costing.
Part of being responsible in the use of AI is ensuring people are using it as efficiently and as effectively as possible.
Over the remainder of 2025 and beyond, all of the enterprise software providers serving our sector will build this type of AI into their solutions. This sort of solution is table stakes: any company that is not making this available to all of their staff is behind the pace.
Leading companies are moving beyond pure productivity improvements and experimenting with more advanced solutions such as tuned models for their organisations and agent-based systems that don’t just provide advice, but start to take actions. The companies that will benefit most from GenAI will allow their people to experiment and will be prepared to innovate to find what adds most value to their organisation and their clients – they won’t be the companies that demand a return on investment analysis before trying something.
The other key area is ensuring that AI is used responsibly. That means managing a range of risks including, but not limited to, ethics, information security and IP. This relies on putting in place processes and people to assess the risks and recommend how to safely maximise the value.
What do you say to those concerned that AI will lead to job losses?
There is so much for engineering companies to do with all the challenges that our clients face around the world: climate change, geopolitical pressures, all the economic upheavals we’re seeing. AI is changing how we work, but that is going to lead to improvements in productivity and enabling people to deliver a lot more by using AI to liberate their time and augment, rather than replace, what they do.
I like how Nathaniel Whitmore, founder and CEO at Superintelligent, describes things “…rather than thinking about AI as a way to cut costs and just do the stuff more efficiently… It’ll be the companies that turn all of their employees into supercharged versions of themselves… and generally have a bias towards doing more, better things rather than being content with the same.”
I think particularly creative people are so incredibly clever that they will find ways of using this technology that we just can’t imagine yet. I completely understand that AI makes people worried, but I think it is going to unleash a whole new wave of creativity. Think what synthesisers did for music. If you want an artist’s view of this, just listen to the interview with Claire Silver. She says: “With the rise of AI, for the first time, the barrier of skill is swept away and in this evolving era, taste is the new skill.”
What do you see as the downsides of generative AI?
“Organisations need to move as quickly as they can, balancing risks and opportunities as well as the cost of investing in technology and training.”
Generative AI brings a new set of risks that are a bit unfamiliar. For example, with all the technology we’ve seen, if you give it a set of inputs, it will always give you the same output.
But with generative AI you can give it the same input and you will get a different output. So, if you ask the same question of ChatGPT a couple of times, you will get a different response, because it’s based on probabilistic algorithms. The way this can be managed is much more similar to managing people, and many of the quality processes that companies have in place will help.
There’s also the risk of bias through the input data, information security and IP, among others, and the level of risk is often dependent on the individual use case. Companies are becoming aware of these challenges and emerging legislation, such as the EU AI act.
Companies are now starting to address these challenges by setting up organisations and processes internally to address them. Some are publishing their approach, such as Mott MacDonald’s responsible AI policy. This will become more common and will provide protection and reassurance as more and more GenAI solutions are deployed.
And other challenges of AI?
How will engineers’ jobs change?
Over the next few years, I hope we see more and more use of predictive AI to improve the delivery and performance of infrastructure. There’s so much untapped potential to improve productivity and deliver better outcomes for clients.
And we will also see GenAI taking on the routine tasks, freeing up people’s time. And I think we should really welcome that. And 2025 will be the year that we start to see AI agents that will not just give advice, but also start to take actions on our behalf.
Beyond that, there will be some potential for really clever applications of GenAI. Last year, we saw Geoffrey Hinton win the Nobel Prize for physics and Demis Hassabis win it for chemistry. Both are computer scientists but have come up with something that is absolutely gamechanging in the world of physics and chemistry. So if we as an industry don’t have the super-clever ideas, then maybe someone else will.
I’m naturally positive and optimistic, but I am aware of the challenges. One is that the rate of development is so amazingly quick. Most organisations can’t move at that speed: they have to check that AI does what it’s meant to do, and whether it’s safe. Organisations need to move as quickly as they can, balancing risks and opportunities as well as the cost of investing in technology and training.
There are other challenges around data and IP and how we can use our work to generate models. Engineering companies often don’t own the IP for the work they do for clients. The client owns that, so you’re not left with much to build your own models from.
For example, an engineering consultancy may have designed thousands of structures around the world. You might ask, why don’t they take all of the data from those, feed it into a large language model, and then the next one they do will benefit from all of that thinking? But they can’t do that because often the client owns the design.
The irony is that their experts have that knowledge and they take it with them when they come to design the next structure – no one has a problem with that. I think this is an issue that the sector needs to resolve if we are to maximise the value of AI for all involved.
What’s your parting shot?
I’ve seen massive technology change over the past 30 years, but I think there’s a potential for the same level or more in the next three years alone. People should really hold on to their hats and get on board – as the Wharton professor Ethan Mollick says, you don’t understand AI until you’ve spent 10 hours with it.
Don’t miss out on information management and digital construction news: sign up to receive the BIMplus newsletter.
The post AI state of the nation: how far have we come and what’s next? appeared first on BIM+.
Darren Russell, AI enthusiast and former AI development director at Mott MacDonald, says the arrival of generative AI marks the biggest digital transformation he’s ever seen – akin to the arrival of computers into the workplace and with even more potential impact than the invention of the internet. Hold on to your hats is his advice to others.
Russell is “in the profession that I always wanted to be in – civil engineering”, but he has had a lifelong obsession with technology and digital. “I’ve just loved it and been fascinated by it. So I’m lucky to have a foot in both camps,” he says.
Following his studies for a civil engineering degree and then a computer science degree, both at Manchester University, he completed a PhD in Sheffield in geotechnical engineering.
He joined Mott MacDonald in 1992 as a geotechnical engineer, but even then, he was involved with digital technology. “Around a decade later, I became the IT director, where I started the company’s digital transformation. In that time, we built the foundation of our technology platform and set up partnerships with Microsoft, Autodesk, Bentley and others,” he explains.
After 10 years in IT, he took on the role of group chief digital officer and spent the next decade delivering solutions for clients. At the end of 2023, he shifted to focus purely on AI, as the development director in this area. He left Mott MacDonald at the end of 2024.
BIMplus: How has generative AI changed the landscape?
“This is change at least on the scale of the internet, or even the arrival of computers in the workplace. The pace of change since 2022 has been phenomenal and it’s just getting faster.”
Darren Russell: AI has been around for decades and companies in our sector have taken advantage of predictive AI for years. This is fantastic technology that can be used in many ways, for example, to predict the performance of infrastructure (machine learning), identify defects in structures or types of vegetation or animal species (computer vision), or to rapidly model engineering systems such as flood risk (surrogate modelling).
But everything changed with the launch of ChatGPT and the rise of generative AI at the end of 2022. It’s the most transformational technology I’ve seen in the 30-plus years I’ve been working in the sector.
This is change at least on the scale of the internet, or even the arrival of computers in the workplace. The pace of change since 2022 has been phenomenal and it’s just getting faster and faster.
While you can think of predictive AI in much the same way as more traditional engineering software, it’s much easier to think of generative AI as an assistant or co-worker. The author and academic, Professor Ethan Mollick, suggests thinking of the current batch of advanced GenAI models as “an infinitely patient new co-worker who forgets everything you tell them each new conversation, one that comes highly recommended but whose actual abilities are not that clear”. Thinking of GenAI like that allows you to start experimenting with the ways that it can help you in your work and your company.
How are companies implementing AI?
GenAI can do loads of the boring, mundane, repetitive tasks that machines are really good at. That might be summarising or starting off a document for you, or doing research or producing minutes of meetings. It buys you time. And that’s where companies are making first use – freeing up people’s time so they can focus on what they joined their companies to do. Typically companies in our sector are using Microsoft 365 Copilot to realise these productivity improvements and, typically, individuals are saving one or two hours per week.
What about AI’s vast energy consumption?
This is becoming an area of real focus, because much of the work we do as engineers is thinking about sustainability. It’s best practice to measure the carbon spent in using AI solutions and thinking about whether the application is delivering more value in carbon terms than it’s costing.
Part of being responsible in the use of AI is ensuring people are using it as efficiently and as effectively as possible.
Over the remainder of 2025 and beyond, all of the enterprise software providers serving our sector will build this type of AI into their solutions. This sort of solution is table stakes: any company that is not making this available to all of their staff is behind the pace.
Leading companies are moving beyond pure productivity improvements and experimenting with more advanced solutions such as tuned models for their organisations and agent-based systems that don’t just provide advice, but start to take actions. The companies that will benefit most from GenAI will allow their people to experiment and will be prepared to innovate to find what adds most value to their organisation and their clients – they won’t be the companies that demand a return on investment analysis before trying something.
The other key area is ensuring that AI is used responsibly. That means managing a range of risks including, but not limited to, ethics, information security and IP. This relies on putting in place processes and people to assess the risks and recommend how to safely maximise the value.
What do you say to those concerned that AI will lead to job losses?
There is so much for engineering companies to do with all the challenges that our clients face around the world: climate change, geopolitical pressures, all the economic upheavals we’re seeing. AI is changing how we work, but that is going to lead to improvements in productivity and enabling people to deliver a lot more by using AI to liberate their time and augment, rather than replace, what they do.
I like how Nathaniel Whitmore, founder and CEO at Superintelligent, describes things “…rather than thinking about AI as a way to cut costs and just do the stuff more efficiently… It’ll be the companies that turn all of their employees into supercharged versions of themselves… and generally have a bias towards doing more, better things rather than being content with the same.”
I think particularly creative people are so incredibly clever that they will find ways of using this technology that we just can’t imagine yet. I completely understand that AI makes people worried, but I think it is going to unleash a whole new wave of creativity. Think what synthesisers did for music. If you want an artist’s view of this, just listen to the interview with Claire Silver. She says: “With the rise of AI, for the first time, the barrier of skill is swept away and in this evolving era, taste is the new skill.”
What do you see as the downsides of generative AI?
“Organisations need to move as quickly as they can, balancing risks and opportunities as well as the cost of investing in technology and training.”
Generative AI brings a new set of risks that are a bit unfamiliar. For example, with all the technology we’ve seen, if you give it a set of inputs, it will always give you the same output.
But with generative AI you can give it the same input and you will get a different output. So, if you ask the same question of ChatGPT a couple of times, you will get a different response, because it’s based on probabilistic algorithms. The way this can be managed is much more similar to managing people, and many of the quality processes that companies have in place will help.
There’s also the risk of bias through the input data, information security and IP, among others, and the level of risk is often dependent on the individual use case. Companies are becoming aware of these challenges and emerging legislation, such as the EU AI act.
Companies are now starting to address these challenges by setting up organisations and processes internally to address them. Some are publishing their approach, such as Mott MacDonald’s responsible AI policy. This will become more common and will provide protection and reassurance as more and more GenAI solutions are deployed.
And other challenges of AI?
How will engineers’ jobs change?
Over the next few years, I hope we see more and more use of predictive AI to improve the delivery and performance of infrastructure. There’s so much untapped potential to improve productivity and deliver better outcomes for clients.
And we will also see GenAI taking on the routine tasks, freeing up people’s time. And I think we should really welcome that. And 2025 will be the year that we start to see AI agents that will not just give advice, but also start to take actions on our behalf.
Beyond that, there will be some potential for really clever applications of GenAI. Last year, we saw Geoffrey Hinton win the Nobel Prize for physics and Demis Hassabis win it for chemistry. Both are computer scientists but have come up with something that is absolutely gamechanging in the world of physics and chemistry. So if we as an industry don’t have the super-clever ideas, then maybe someone else will.
I’m naturally positive and optimistic, but I am aware of the challenges. One is that the rate of development is so amazingly quick. Most organisations can’t move at that speed: they have to check that AI does what it’s meant to do, and whether it’s safe. Organisations need to move as quickly as they can, balancing risks and opportunities as well as the cost of investing in technology and training.
There are other challenges around data and IP and how we can use our work to generate models. Engineering companies often don’t own the IP for the work they do for clients. The client owns that, so you’re not left with much to build your own models from.
For example, an engineering consultancy may have designed thousands of structures around the world. You might ask, why don’t they take all of the data from those, feed it into a large language model, and then the next one they do will benefit from all of that thinking? But they can’t do that because often the client owns the design.
The irony is that their experts have that knowledge and they take it with them when they come to design the next structure – no one has a problem with that. I think this is an issue that the sector needs to resolve if we are to maximise the value of AI for all involved.
What’s your parting shot?
I’ve seen massive technology change over the past 30 years, but I think there’s a potential for the same level or more in the next three years alone. People should really hold on to their hats and get on board – as the Wharton professor Ethan Mollick says, you don’t understand AI until you’ve spent 10 hours with it.
Don’t miss out on information management and digital construction news: sign up to receive the BIMplus newsletter.
The post AI state of the nation: how far have we come and what’s next? appeared first on BIM+.
The arrival of generative AI marks the biggest digital transformation yet seen in construction: embrace or fall behind, says Darren Russell, formerly of Mott MacDonald.
The post AI state of the nation: how far have we come and what’s next? appeared first on BIM+.