amanda southworth

I keep continuously seeing the argument that generative AI agents will improve people’s ability to write and ship more code. Therefore, through some odds and ends, we should theoretically see more non-profits or individual developers be able to manage and maintain social good software products.

Armed with good and our large goals, we go out to attack the harms and evils of the world. We do sometimes win. We sometimes do not. We look around at all of the evil that still swirls in the atmosphere above, and wonder how we could ever breathe well enough to survive through it. You have to wonder if good will ever be enough. As if good is a limited resource that we must convince people to do and to apply, or something we actively need to conjure in the right amounts as a salve to the evil of the world.

It seems unconscionable in this day and age where we keep on finding serial predators and wondering why we never uncovered them sooner, that a technology that does just that is gone.

A while ago, I wrote an essay called ‘building software that “helps”’. In it, I attempted to formulate and distill a more operational format of the stuff I write about all of the time: software, technology, what it does to us and how to make that ‘doing’ more gentle and beneficial. As I was writing it, I came up with more and more caveats and other information I wished to pull in, but didn’t. There’s a number of core lessons that I’ve come to and written about, but I’ve never formatted it in a way that consolidates those learnings. In starting — I learned about how much was left to be done.

In startup terms, the whiteout is the constant envelopment of mental model gathering and shifting that someone must do to exist and excel in a startup environment. I used to think the hardest part of a startup is gathering the skills, and I now feel that it is managing the whiteout.