Grupo Vizcaya is looking for an Arquitecto superior.
I look at this job opening and smile. Alicante, sun, sea… and an architectural firm looking for a young specialist for a “senior architect” position. I checked their website – beautiful, modern residential complexes. It’s clear these guys have taste and, more importantly, a steady stream of projects. They work in BIM, as is standard in the modern world.
So, what’s the pain point? It’s readable between the lines. They don’t need a creator or a genius. They need a reliable executor. Someone who will sit down and methodically transform beautiful concepts into working documentation. “Redacción de proyectos de ejecución, memorias y mediciones” – that’s classic. Drawings, specifications, cost estimates. The most routine, most meticulous, and most time-consuming part of an architect’s job. They are looking for a functional role, someone to cover this production stage while senior partners design new villas and negotiate with clients. They are looking for “entry-level,” meaning they want a qualified specialist for relatively little money. A classic scaling story.
Now, let’s imagine that instead of a young architect, full of ambition (and a desire to get paid every month, go on vacation, and occasionally get sick), they “hire” an artificial intelligence-based system. Fantasy? Not at all. This is already today’s reality; not all managers have simply gotten used to thinking this way yet. AI won’t replace the chief architect, but it can easily replace that very “junior-senior” documentation specialist.
Let’s play out a short dialogue in the mind of a hypothetical Grupo Vizcaya manager. Let’s call him Javier. And his interlocutor will be his inner voice, having read my blog.
Inner Voice: Javier, we’re looking for someone for routine tasks again. Maybe it’s time to stop?
Javier: What are the alternatives? Projects won’t draw themselves, specifications won’t write themselves. We need hands and a mind that knows BIM.
Inner Voice: Hands – yes. But a mind… Let’s break down the tasks of this new employee. First – “proyectos de ejecución” (execution projects). Take your own Revit or ArchiCAD, which you use. AI-based plugins like TestFit or Finch already exist today. You upload a schematic design, set parameters – Spanish building codes (Código Técnico), insolation requirements, materials. In a few hours, the AI generates a complete set of working drawings: floor plans with dimensions, sections, elevations, details. Not perfect, but 80% ready.
Javier: But it needs to be checked! What if it makes a mess?
Inner Voice: Of course, it does. But what’s more efficient: spending 80 hours creating drawings from scratch or 10 hours reviewing and correcting what the machine generated? Your experienced architect, whom you’re currently distracting with mentoring a novice, will be able to check the AI’s work much faster. He’ll become a quality controller, not a producer.
Javier: Hmm, alright. What about the second point, “memorias” (specifications)? Those are texts, legal formulations.
Inner Voice: Oh, that’s the easiest part. Take any serious language model, even GPT-4. Create a template for it: “Write a technical specification for a 5-story residential building in Alicante. Wall material – …, windows – …, energy efficiency standard – …”. The model, trained on millions of technical documents, will give you a ready-made “memoria” (specification) of 50 pages, which you’ll only need to proofread for compliance with the specific project. That’s a couple of hours’ work, not a week.
Javier: And the cost estimates, “mediciones,” will it calculate those too?
Inner Voice: That’s its favorite. Modern AI tools for BIM (like Togal.AI, for example) scan your 3D model and read every cubic meter of concrete, every square meter of tile, and every linear meter of cable. The accuracy is higher than that of a tired person who might miss something by Friday evening. Errors in cost estimates cost real money, remember?
Javier: Sounds good. But how do we implement all this? It’s a whole revolution. My people will be scared that robots will replace them.
Inner Voice: This is where your managerial wisdom is needed. You need to reduce distrust gradually.
1. Start with a pilot project. Take one small object and try generating documentation with AI in parallel with human work. Compare time, costs, and the number of errors. Numbers convince better than words.
2. Explain to the team. AI is not a replacement for them. It’s their most powerful tool. It takes away routine tasks so they can focus on creativity, complex details, and client communication. The architect stops being a drafter and becomes an orchestra conductor, where AI is one of the instruments.
3. Invest in training. Teach your key employees not to draw, but to set tasks for AI and check the results. These are new, more expensive, and valuable skills in the market.
And how to validate the result? Very simply. You already have people for this – your experienced architects. Their expertise doesn’t disappear. Only now it’s applied not to manual labor, but to final verification. The process looks like this:
1. AI generates a package of documents (drawings, texts, cost estimates).
2. A junior specialist (perhaps not even an architect, but a BIM operator) checks the basic compliance of the model and documents.
3. A senior architect performs the final check, affixes their licensed signature, and bears full responsibility.
In the end, instead of hiring another employee with all the associated costs, Grupo Vizcaya could invest in subscriptions to several AI services and slightly restructure its processes. The gain in speed is manifold. Scalability is infinite. And most importantly, their best minds would be freed from routine to create the beautiful architecture they sell. But for now… for now, they are looking for a person. Well, that’s a choice too. Just, perhaps, no longer the most efficient one.
Источник: https://www.linkedin.com/jobs/view/4404816629/
