"Why am I doing this?"
One of the reasons that I think I still have a job in tech is that, despite not having a college degree, I've always had a love and fascination for technology. I remember being eight or nine years old at my grandmother's house tinkering with her electronics. That tinkering continued into the era of personal computers where I futzed and fussed with wretched IRQ numbers, baud rates, boot orders, sound drivers—all in order to play Doom. In high school, I took a handful of computer-related classes; most of them were so easy that I zoomed through them and passed the time finding and playing with the easter eggs hidden in Office 97. I did some futzing around with HTML but didn't have the patience for it.
One of my fondest memories during that time was when my friend and I realized that computer performance was impacted by thermals. So, in the dead of winter, we would intentionally leave the windows open in his room all day so that when we came to his house after school and went to his frozen room to play video games, we would get +20 more FPS and sub-300ms latency in Counter-Strike 1.5 and Team Fortress Classic (this was 2002).
When I went to college, my focus was on learning history, philosophy, and theology. However, I continued to enjoy tech and being on the bleeding edge. I went on to dabble with blogging, podcasting, video production, and live streaming on my own time. All my leaning into technology typically had one of two endpoints: a) I wanted to enjoy playing or talking about playing video games and I wanted the best possible FPS in them, or b) I wanted to feel like I was creating something that meant something to someone.
In the early 2020's and before, I didn't pay much attention to the state of language models and artificial intelligence. Every once in a while I'd notice an article on the topic about something crazy-sounding like a model that was playing StarCraft. I'd take mental note and move on. But I remember my first first Neo-in-the-Matrix-"Whoa" moment I had with large language models in August 2022. Dall-E had just been released by OpenAI, and access was being doled out on an invite-only basis. Which wasn't a surprise to me. I'd been trained my whole life to wait in line for something: Disneyland, Black Friday, a Gmail invite, a Facebook invite. Kids these days just don't know.
Among the first images I generated was this one:
