I am an architect and I do a lot of ai work. My specialty is actually in generative parametric design and I’m just taking a break from writing some code for an AI project I am working on.

I’m seeing a lot of bad takes here, I’m assuming from a mixture of not reading the article and not knowing what actually goes on in the field. No, we don’t just make pretty pictures. No, a trend you don’t like of boring or shitty buildings you’ve seen doesn’t mean the profession is dying (and for a lot of those you can look at developers to share the brunt of your and my irritation).

People working as “architects” do a huge variety of work and no two you talk to are going to have the same workflow or process so I cant speak for everyone. For me, ai tools aren’t ever going to take my job, just remove more time consuming tasks and, in the long run, increase complexity and expectations. Same as when we moved from hand drafting to CAD, and again when we moved to 3D BIM design.

Each step drastically reduced busy work but over time increased the base level complexity in the design work. When architecture was all analogue, we weren’t doing statics modeling and parametric studies. And now with BIM, I have to consider and model equipment and MEC feasibility. Even compared to a couple years ago, now I’m doing solar and environmental modeling to track energy performance and inform the designs and suggest changes early on.

There was a doctorate researcher I spoke with recently that mentioned that the direction the profession is going is that we will no longer make individual choices for every design element. Instead, we will manipulate the data and direction that end up at the final choice. And I think he’s right. I think in the last year I’ve hand modeled maybe one project? Everything else has been purely data driven generative design.

I use AI image generators to do early design inspiration alongside sketching. I have a local Stable Diffusion AI instance trained on my wireframe modeling that I use to create scenes for presentations faster. I build small tools that help me recursively optimize structural elements. The last few months I’ve been working on my own big AI project that could really help a lot of my peers as it develops, too. I can’t talk about it just yet but I will after the funding period ends. The future is looking bright.

Tldr: the whole field of architecture isn’t responsible for those shitty city apartments you don’t like, AI tools are helping us because architecture is much more data driven and complex than you think it is, architecture isn’t a struggling or dying field like the article quotes- what’s killing the joy is greedy cheap developers.

Happy to chat or answer questions

Not if that first picture is anything to go by.

Put AI in the headline and people will click on it.

Include “…will replace _____” and even more will click on it.

Who is the one being programmed here? The computer or us?!

/facepalm What kind of Escherian hell will AI architecture come up with? The worst kind.

This is the best summary I could come up with:

One second they form rows of towers, next they morph into low-rise courtyards, then they flip back into long slender slabs, before cycling through hundreds of other iterations, in a hypnotic high-speed ballet of bristling buildings.

Image-making tools such as Dall-E, Midjourney and Stable Diffusion have allowed the effortless creation of seductive visions: skyscrapers in the style of Frank Lloyd Wright, fantasy mash-ups of sci-fi and art nouveau, squidgy marshmallow staircases, buildings made of rubbish.

Tsigkari’s team has also developed a simulation engine that allows realtime analysis of floor plans – showing how well connected one part of a building is to another – giving designers instant feedback on the implications of moving a wall or piece of furniture.

Just as deepfake technology makes it possible for dead actors to be resurrected in new roles, could we see a world where the back catalogues of deceased architects are used to generate cover versions of their work, regurgitating counterfeited classics in a nauseating feedback loop?

Having since worked for central government on digitising the planning system, Mills has now co-founded Blocktype, an AI-powered tool for developers and planners, aimed at simplifying the process and ultimately reducing land speculation.

“The idea behind Blocktype is that it can give you a ballpark sense of what’s possible on a site, providing sketch layouts and viability appraisals.” Mills stresses it is not a replacement for architects, but a tool to help developers think spatially when trying to determine what to pay for land.

I’m a bot and I’m open source!

How about we have machines deliver packages while people make art?

Arcitechs design some buildings that are pretty.

But i rarely see modern buildings that are designed by an arcitect, that are also practically designed for the people who work or live in it, on a daily basis.

A lot of them dont seem to think: if i worked/lived here… how would that work out in the long term?

My wifes workplace was designed by an arcitect and its dreadfully idiotic in every way. But it LOOKS good.

If it was trained off of internet data every building will be full of ball pits, impractically large kitchens, a hidden swastikas.

People should have ways to sustain themselves and interesting things to do. So, each time computers (& machines) beat us at something we tend to deny it until it is really obvious. Let’s face reality : in the near future we will have to admit they beat us even in architecture.

There is a use case, for example, running pipes and cabling in an efficient way.

But I really don’t want to see more soulless design.