The Bauhaus question

If you search for Bauhaus inspiration, you’ll get plenty of circles, primary colors and clean typography. Those are easy to copy. What interests me more is the question the school was built around: how should people learn to make things when the means of production are changing?

The Bauhaus existed from 1919 to 1933, and in that short time it changed its location, its directors and its emphasis, so treating it as one frozen style loses most of what made it interesting. Its 1923 publication, Staatliches Bauhaus Weimar 1919-1923, presents a program where technology and art are part of one education. These are two kinds of knowledge that tend to develop their own language and their own measures of success, and the school tried to connect them.

I see the same split in how AI tools get discussed. One conversation is about capability: speed, cost, model quality, how much work can be automated. The other is about experience: whether the result is coherent, appropriate, worth living with. I’m an engineer, so I know how tempting it is to treat the second conversation as decoration. Designers can do the same from the other side and treat the first one as an implementation detail.

A finished object shows quickly why that doesn’t work. Take a lamp designed with a generative model. It can look extraordinary in the image. To be a good lamp, its shape has to deal with how light behaves, what the material can do, how hot the components get, whether somebody can assemble it, and what happens when a part breaks and needs replacing. The image doesn’t do any of that work.

I don’t want to stretch the analogy. A company providing a model is not a school from the 1920s, and generative software comes with its own economics, dependencies and disputes about training data.

What I’d borrow is the idea of learning through connected practice, where people move between imagining, specifying, making and evaluating. I’d love to see a course where a programmer has to explain a material choice and a designer has to explain a failure condition! Nobody needs to become an expert in everything. It’s enough to have so much contact with the other side’s problems that you can have an honest conversation about them.

It’s also why I don’t believe taste is a last layer you apply to a finished system. Choosing what to optimize already decides what the system can produce, so the aesthetic judgment is there from the first decision about which results count as good.

How do you set up a practice where technical ability and a sense for experience make each other better? I don’t know yet, and I’d like to find out.