New York

Hi, this is Isma 👋 

There are so many posts on LinkedIn about AI.

Flux, Loras, Generative AI…

I keep an eye on the interesting things that I come across, and I try the coolest ones myself to understand their real capabilities.

Add it to my workflow.

Like ComfyUI, or training custom models, for example. That´s cool.

I played around with Stable Diffusion as well, with automatic11, before…

Bive coding. 100%…

I find this fun.

But at the same time, lately I have been thinking about what would be one of the “ultimate” workflows for architects.

Smartly leverage all this.

It is much faster to produce concept designs now, with ultra-enhanced realism:

With sketch to image, image to image, image to 3d, and image to video.

Being much faster. I like that.

Being in control. I like that too.

Where does “traditional” computational design fit in a workflow like that?

More control during the initial stages, plus faster iteration?

Software interoperability? From conceptual models to rationalized BIM?

…

How many architects are modifying their workflows to introduce AI?

How I see it: AI is for everybody. Like the internet.

While the user experience improves, there´s not much of a learning curve.

It is not the same for Computational Design / Parametric Design…

This is how my LinkedIn feed looks: mostly AI and software… but, at the same time.

Two weeks ago, I was on a super nice beach in the Bahamas, talking with a Spanish architect.

He designs villas in Nassau.

Almost all his job is: hand drawings and AutoCAD.

No AI. No BIM. No Parametric Design.

…

Opposite sides of the same coin.

I would love to hear your thoughts on this.

Yes yours.

PS: I will be in NY next week. If you live there and want to meet, I´m in.