This weblog started at a time when data science was an up-and-coming field, deep neural networks could recognize numbers and cats, and various platforms allowed us to connect with friends. There was no definite-article Algorithm, and machine learning was a more ubiquitous term than AI. I was very optimistic about the advances we could make, given the growing amount of data and compute power coming online.
We’re now at a point where data is used to foreground the most rage-inducing content to stimulate engagement, platforms lock users into a lonely dopamine cycle of posting on a schedule, and interacting with an LLM can induce AI psychosis. And let’s not even think about the ethical and ecological problems with training LLMs. Touching a computer just isn’t as fun as it used to be!
Perhaps it’s these trends, or perhaps it’s a natural result of getting older, but I find myself thinking about the positives of a previous iteration of the internet.
- It was a smaller internet, where communities were human-scale and interaction was meaningful because it came from people you knew, or at least were friends of friends.
- It was a quirkier internet, where individual weblogs showed personality instead of fitting into the mould of a given platform’s user page and content expectations.
- And it was a slower internet, where people produced thoughts and art on their own schedule, instead of pumping out content on a cycle dictated by The Algorithm.
Be the change you want to see in the world
As a result, I’m going to be less strict about the type of content I post here. I’m going to highlight neat things I find in the world, whether or not they’re data science-related. As I also pull back from posting on the big platforms like Instagram, I hope that this flexibility will entice me to post again (although absolutely not on a regular schedule!).
My first neat thing to highlight is: look back on some of your old work! As I was tidying up this site and moving it from GitHub to Codeberg, I realized I had made a whole separate portfolio site that highlights some of the results I’m most proud of from my time working in power engineering. I never linked to it before because it’s cringe to like your own posts on the modern internet. But I was surprised at both how broad spanning and still interesting my past work was!
So maybe revisit some of your old work? It’s probably better than you remember, and you might find it inspiring!