The AI Development Dilemma: Who Really Captures the Value?

In 2024, a developer sat down with TestFit, a generative design tool that’s quietly reshaping how feasibility is done in real estate. Instead of calling three architects, waiting weeks, and paying for concept sketches, they ran the numbers themselves.

Within a few hours, they had dozens of site layouts, parking configurations, and yield-on-cost models. The verdict was brutal: seven out of ten schemes that looked fine on paper wouldn’t work out in reality.

That single day of analysis saved the developer from wasting months of consultant fees and millions in soft costs. More importantly, it redirected their capital to sites that could actually deliver.

As Bisnow recently put it, “Generative design software like TestFit has turned feasibility studies into real-time decisions, not multi-week exercises.”

The value proposition is simple: move faster to contracts, financing, and control of the project. And in today’s market, that’s the real game.

The Wrong Question, the AI Dilemma

Most panels and articles keep circling the same question: “Will AI replace architects?”

It’s the wrong question.

History shows us that new tools rarely erase professions outright. The change is subtle, gradual and unavoidable at the same time. The technology shifts leverage. CAD didn’t kill architecture, but it gave firms with a quick implementation mindset the upper hand. BIM didn’t eliminate designers, but it privileged those who could manage complex data environments.

As Richard Sennett argues in The Craftsman, tools don’t just extend human capability; they restructure the relationships between those who use them. AI is no different.

The right question is: in an AI-driven industry, who captures the value, the developer, the architect, or the investor? The answer is unfortunate for architects (an industry currently plagued by tech sceptics): the one who moves quickest.

Procurement: The Quiet Power Centre

Procurement has always been the invisible lever. It decides who leads a project, who absorbs risk, and who pockets profit. AI supercharges procurement in three ways:

1. Speed: AI collapses feasibility assessments from months to minutes.

   
   

2. Pattern Recognition: AI spots correlations across datasets (planning history, material cost volatility, rental yields) that humans simply can’t filter at scale.

   
   

3. Negotiation Power: AI-assisted contract analysis reduces ambiguity, compressing negotiation cycles.

   
   

McKinsey’s Reinventing Construction (2024 update) estimated that AI-driven procurement could cut project costs by 10–15% across the lifecycle. MIT’s Tech Review reported developers slashing negotiation times by 30% using contract optimisation software.

These are firms in Europe and Asia already building faster, leaner, and with tighter financial control.

And here’s the uncomfortable truth: developers, not architects, are the natural beneficiaries. They already own procurement.

The Developer’s Edge

Developers win by default because AI plugs straight into their workflows: site selection, risk modelling, cash flow forecasting.

Imagine this scenario:

• A developer uploads a design into an AI system.

• The system benchmarks it against 1,000 built projects.

• It flags that your scheme is 18% over budget compared to market comparables.

• It predicts, with 92% confidence, that a competing scheme will yield stronger rents.

By the time you walk into the design review, they already know your weaknesses.

When procurement becomes algorithmic, developers avoid starry-eyed architects with promises that never materialise. And as Geoffrey Hinton warns, the danger of AI is not sentience, but centralisation of power. For real estate, that power sits firmly with capital.

The Architect’s AI Dilemma

Architects risk being relegated if they stay “inside the drawing.” The risk isn’t replacement, it’s reduction: from strategic partner to service provider.

BIM showed us how this happens. Firms that embraced BIM early gained a seat at the table in coordination and delivery. Those who resisted found themselves at the mercy of contractors.

AI will replay the same pattern. But faster.

Some practices are already pivoting:

• Zaha Hadid Architects use AI-driven generative models for massing and rapid iteration.

• BIG has invested in in-house data analytics to test schemes both financially and spatially.

• Start-ups like TestFit automate feasibility modelling, allowing developers to bypass architects for early-stage layouts.

The message is clear: unless architects expand into procurement literacy, their role narrows. Unless we see architecture as a means to make money, we will lose control again and again until AI will definitely be able to replace us.

The Three Possible Futures

From here, I see three plausible models for the architectural studio:

1. The Hybrid Studio-Developer

Architects bring procurement and data in-house. They offer feasibility, financial stress testing, and design as an integrated package. Think “mini-developer with design DNA.”

2. The AI Partnership Model

Studios ally with AI consultancies. They walk into pitches not just with renderings but predictive financial models. They shift perception from cost-centre to value-protector.

3. The AI-Native Studio

Entirely new firms emerge, built around AI workflows. Few sketches, lots of machine learning. Procurement and design collapse into one. These firms compete aggressively with both traditional architects and developers.

Across all three, the common denominator is this: procurement literacy becomes as critical as design literacy.

Lessons from Other Industries

Architecture isn’t the first sector to face this shift. Look at medicine: radiologists feared AI would make them obsolete. Instead, AI tools amplified the value of those who knew how to interpret them in a clinical context. The losers were those who ignored the technology.

Look at finance: algorithmic trading didn’t erase traders, but it destroyed those who couldn’t adapt. The winners learned to partner with machines, not fight them.

Architecture will rhyme with these stories. The profession won’t vanish. But the distribution of influence will change.

Practical Playbook for Architects

So what can architects do now? Three moves:

1. Get fluent in procurement. Read as much about contracts and financial models as you do about design. The Language of Architecture won’t save you here but The Language of Real Estate Finance might.

   
   

2. Experiment with AI tools beyond design. Don’t just play with Midjourney. Test TestFit, Spacemaker, or Archistar. See how they collapse feasibility into hours. Understand what your clients are already seeing.

   
   

3. Reframe your pitch. Stop selling “design” as a standalone. Sell design that maximises and protects value. Developers don’t buy aesthetics. They buy reduced risk and enhanced returns.

The Challenge

The real danger isn’t AI replacing architects but developers asking: “If the machine tells us what’s viable, why do we need an architect before the Concept Design stage?”

That is the AI Dilemma, and unless architects climb up the procurement ladder, they’ll be called in only after the value has already been captured elsewhere.

Final Thoughts

AI is a procurement story. At this moment in time, we can start writing ourselves into the chapters that will be read by future generations. Because 'to the victors belong the spoils' we need to make sure we are not reduced to a footnote.

Procurement is where the future of architecture will be decided.

This newsletter exists to explore these collisions: between design and capital, between technology and procurement, between who builds and who profits.

If this resonated, do two things:

1. Share it with a developer, architect, or investor who needs to be in this conversation.

2. Grab the free AI x Development Cheat Sheet (key tools, reports, and resources) I’ve created for subscribers.

AI_x_Development_Cheat_Sheet.pdf

Download PDF • 239KB