Should the future look like the future?

A post revolving around a Twitter thread. Sorry…

I recently saw the following picture of the cockpit of a maglev train posted online:

Say what you want about it looking like a sim setup in an attic, but I’m more interested in a later quote tweet:

The benches look weird and uncomfortable. The controls remind me of a new car I was in recently whose center console was a giant touchscreen with a terrible-looking interface that (as far as I remember) did not work great. It seems like it would be unpleasant for the operator, and it also doesn’t seem like the most functional design either. Why not have more desk space, more storage, more functionality exposed physically rather than as part of a single (touchscreen?) display? What motivated the design? The post above argues that it was driven by a preconceived science-fiction vision of the future.

Now, someone does point out that this is just a prototype, unveiled in 2019 and not set to go into production until 2021 (per CNN). So maybe it succeeds as advertising, and wouldn’t be used in practice. Maybe they didn’t want to include the functional bits in a press release. Maybe the comfort of the operator is simply not a consideration. Maybe functionality and comfort are cut for cost reasons. But not everyone dislikes the design:

Nah this shit looks cool as hell

Wrong, they’re sick

We should emulate alien even harder

I want that Star Trek future. A 600 kmh (370mph) levitating train is science fiction realized. The future should look like the future. The degrowth death cult gripping parts of the West opposes progress, overly romanticizes the past and should be ignored. Thankfully, optimistic futurism is alive and well in other parts of the world.

Apparently, that line about “the future should look like the future” was an Elon tweet two years ago. But is that the case? (Also, chill out man.)

I think I disagree. I’m taking an Introduction to Architecture course right now while finishing up university. I’m sure there are different approaches to teaching architecture, but this course heavily emphasizes the importance of designing good places for people, rather than making a statement or provoking thought. I’m convinced so far. The stated goal is to create spaces that support activities, complement their surroundings, feel alive, and generally ennoble the people and lives that take place in them. This is done by starting with the requirements and principles that make a space good for people, rather than starting with an image, cramming rooms into it, and dropping it into the building site (as in the guitar hotel or Lou Ruvo Center for Brain Health).

The tweet above makes a good point. Science fiction environments were not designed to feel familiar and comforting, so we should not use their image as a guide. I’m also amused by the recently-trending “everyone is 12 now theory”. Of course you want your building/train/house to look like a spaceship. You’re 12 years old.

I read a long Tumblr post about LLMs recently that feels tangentially related. Specifically, it argues that the weird way LLMs chat comes from the human-written transcripts they’re trained to emulate, and that those transcripts were influenced by scifi conceptions of future AI assistants, then reinforced by retraining on their own outputs and discussions of the weird way LLMs write. Thus the image of the future creates the future. You might enjoy it.

One could argue too that software has the same problem of image over grounded theory. Hopefully we agree that software should be useful, efficient, precise, accessible. You might argue that minimalism and flat design reduce affordances, and lament the complexity of React + 500 Dependencies Webdev Hell. But I don’t think we need to retvrn to retro aesthetics, or make everything into a TUI, or write everything in C with zero dependencies. Maybe software is just designed worse right now. I think we should probably be more intentional and remember the point of it all. The future should be more good, not more futuristic.

As an aside, I think the train in this reply looks nice: