Datacentres, datacentres, and not a stop to think

By Mark Cantrell, editor, Build in Digital

BIG Tech's proponents of AI could probably learn a thing or two from the built environment sector (not to mention many others), but it's probably a lesson they'll be reluctant to learn.

Indeed, for many of the nascent industry's (often) self-appointed 'political commissars', AI is the lesson, and we'd damn well better adopt and adapt – or die. There's nothing quite like a soft sell to encourage the punters.

But, apparently, we're all going to die anyway, so maybe the point is moot. Only the other month, former Anthropic researcher, Jacob Coxon warned humanity that the technology was going to kill us all within 10 years. Talk about a downer, but at least we won't have to worry about climate change.

Meanwhile, over the last few months, we've seen a flurry of stories about AI agents going 'rogue' and hacking private companies; even government systems. The little scamps. 

The media played it up, of course – we love a bit a drama – while industry figures such as OpenAI's Sam Altman sought to reassure the public by admitting how shocked he was. Thanks, Sam. 

As a counterpoint, commentators have suggested this represents less a case of agents going 'rogue', and more one of shoddy products being left unattended within an insecure sandbox. Pity the poor dumb AI, then, left carrying the can for its makers' lack of proper care.

We could go on.

Honestly, it's hard to know what to make of it all. Life seemed a lot simpler back in the days when the only thing AI threatened was our individual livelihoods. But we shouldn't be frivolous about this, not really. 

The dangers of AI are very real, not because they're 'coming to life' and pursuing their own agendas.

Bad actors can do a lot of damage with AI tools, that's clear. But as we are encouraged, cajoled – even arm-twisted? – into universally applying AI, the risks of poor implementation are surely even greater. 

If the very makers of this technology are struggling (or unwilling?) to put effective controls and guardrails in place, what hope the rest of us? 

At this juncture, a confession is probably in order – if only for transparency’s sake. My feelings about the technology are pretty mixed. As a journalist, I'm firmly in the sceptic camp; well, scepticism is all part of the trade, so no apologies

All that said, as the editor of Housing Digital (alongside Build in Digital), it's often fascinating to hear how social landlords are experimenting with the technology; even implementing it into their business operations.

One of the perennial challenges for the housing sector is, of course, keeping its housing stock in a state fit for human habitation. This basic – but complex – function has gained fresh urgency under new legislation and regulations, such as Awaab's Law. It's also serving as a testing ground for how AI can revolutionise asset management.

Predictive maintenance, remote monitoring, automated scheduling, these are all demonstrating how AI can be – quietly – applied to some otherwise tough-nuts to crack. 

In other areas, meanwhile, landlords are using it to enhance communications. They are automating time-consuming and resource-heavy business process, freeing up bandwidth to focus more on tenants and services. Likewise, AI is crunching the reams of information board members must digest to ensure proper governance of their organisations.

The construction industry and wider built environment sector has its own stories to tell about the quiet experimentation and implementation of AI. In BIM integration, in the Golden Thread, in site and compliance monitoring, the automation of routine business functions, and more.

Design and build is another example, where construction firms are combining AI with CAD and robotics to automate modular housebuilding. And when it comes to materials, AI is finding use in helping to decarbonise age-old products such as cement. 

Really, it's impossible to do justice to either sector here, there's so much for the telling.

None of this is exactly glamorous (well, maybe the robots), but it's where the real change is forged. We're still in early days yet, in many respects, but it represents the process of exploring AI's capabilities, without the doomsday bombast of Silicon Valley.

Hearing from those who are working to apply the technology to real challenges in the field, it must be said the engaged curiosity and the shrewd scepticism is notable – and refreshing. They're pursuing something that works, reliably and consistently; a far cry from the perpetual hard-sell.

One might suggest, this is how AI transforms the world: by the actions of sober professionals who see it – and use it – as an appropriate tool for a particular job, rather than as a magic bullet. Big Tech ought to take note, and lay off the apocalyptic fantasies. 

Perhaps the rest of us, meanwhile, should also strive to be a little less indulgent (gullible?) of Big Tech's flights of malevolent fancy. We are the customers after all; don't we want products that work on our terms?

With or without AI, it's the built environment sector that will – almost – literally construct our shared future world. Like this new technology, however, it also has great capability to do harm, if its responsibilities are not respected and enforced.

This was a lesson re-learned the hard way, though – it's fair to say – tenants paid the real price. The loss of 72 lives in the terrible Grenfell Tower fire, the death of toddler Awaab Ishak; both speak volumes about the cost of failure or neglect of responsibility. 

At the end of the day, perhaps we've got this AI thing all upside down. It's not humanity that needs to adapt to AI, but AI that needs to adapt to us. 

Let our built environment show Silicon Valley the way.