Message in a Bottle

Mar 10, 2023

          Honestly, there's way too much I could say about studying/working with AI, especially with everything coming out. Advice, opinions, reflections, I don't feel particularly qualified to give any of them, so perhaps this is a list mainly of advice to myself that maybe others will resonate with too.

1. Ask yourself if this needs a screen

Half the startups I hear about today seem like someone deciding to insert themselves as a middleman into some first world problem that really doesn't need an app or a screen or AI or maybe even a solution at all and simply allows that someone to put "CEO & Founder" in their LinkedIn heading. Maybe that's pessimistic. I honestly believe that most problems can be solved by just going for a walk or talking with another human being face-to-face, and that if more problems were solved amongst trees or with a collaborative face-to-face solution instead of the hundredth app on my iPhone or the hundredth ChatGPT API interface, we'd be way better off. When studying AI and technology, it's super tempting to think "oh my god I have the perfect app idea to fix this problem," but I think the most truly human-centered solution probably isn't going to add another hour to someone's weekly screentime counter. With that being said:

2. Think deeply about the term "human-centered"

My Instagram explore page is filled with UI/UX design tidbits about how you shouldn't use black text on a white background but rather very very dark grey text on a very very light grey background to prevent user eye strain. While this and all the other tips I see are valid design heuristics and important details to remember, I feel like calling this "human-centered design" is a little overkill. It makes it too easy to think that creating the best design for humans means padding your button differently or choosing a different font and not whether or not you should be making this product at all. Rather than getting sucked into these small details, I want to try to remember as much as possible to question the premise and ask whether my design projects are truly improving way of life or just improving the efficiency of something that might be unnecessary at best or destructive at worst. Is your AI truly "human-centered" when you've minimized the number of clicks in a task flow if you're actively displacing hundreds of people from their work?  

3. Balance your time:
Reading, Experimenting, Discussing, and None of the above.

This one goes a lot of different ways.

Reading: I think sometimes I (and other people) get sucked into making and coding and rarely stop to look up and see what's going on around them: keeping up with the news, reading long-form articles and books, slowing down and just generally enjoying a multidisciplinary and horizon-broadening experience that's more than just VSCode. I think forcing yourself to open up your mind to different ways of thinking prevents you from getting sucked into a world of AI and code that prioritizes speed and optimization and cool gadgety effects without thinking about secondary effects, and I think the best way to get a feel for those effects is to spend time doing things that are not AI. Note that this does not include doomscrolling on Twitter or Instagram.

Experimenting: At the same time, I think it's easy for me to get sucking into reading and thinking about things I want to do instead of actually doing them, so I feel like it's important to spend time actually experimenting with the ideas that excite you. Coding up demos, making music, creating art, etc. If you want to make cool things, you have to learn how to make cool things, and you learn that by trying it yourself, and in the process learning generally about the capabilities and limitations of different tools.

Discussing: I think an import caveat to this, though, is that you don't necessarily need to release all your experiments into the public void for people to start interacting with, because I feel like there's definitely already one two many of those projects that went off the rails. However, I'm very pro sharing projects in smaller environments (like class!) and discussing ideas.

None of the above: Go walk through a rose garden. Swim in the ocean. Meditate. Climb a tree. I've always struggled with being indecisive, so of course I couldn't say any of the above without yet another follow-up-caveat-footnote-asterisk. As much as there is to be learned in reading and making and talking, the best way to be able to think critically about anything is to have space in your brain to do so, and sometimes the best way to clear space for fresh ideas and perspectives is to do. nothing.

Maybe you're already doing all of these things, but unfortunately I'm still figuring it out.

Also, sorry for ~maybe~ going over the 500 word limit again 😅