Social Media today is a perfect reflection of all the ways in which the world is broken

I've been around computers long enough to remember the earliest forms of online social networking: independently-run Bulletin Board Systems (BBSs) in the 1980s, and Usenet in the 1990s. Despite the fact these early interaction tools relied on dial-up modems (my very first modem ran at 300 baud - you could watch text flow onto the screen like a mason laying rows of bricks), conversations were rich and meaningful and full of human personality. In some ways, I think the friction in those systems might have enhanced the quality of interaction. After all, when getting online involves a series of ear-splitting carrier signal shrieks, and staying online means tying up the phone line, the mind tends to focus on making the most of that effort, that friction. Or at least mine did.

That doesn't mean I never encountered toxicity or disordered personalities in those early online forays. I still vividly remember a guy on the CompuServe dialup boards who was using an IBM PC when most others were using Commodore, Atari, TRS-80, and similar "home computers". He somehow discovered that when he inserted certain control characters in his posts, it would issue a clear-screen command to everyone else. When called out on it, instead of stopping with the special characters, he edited his screen name to include those characters, so literally every time he said anything at all it would clear everyone's screen. So yeah, we had trolls back then too.

But what we didn't have was algorithms and algorithmic amplification of some voices over others. Once I figured out how to filter and ignore stuff based on headers (CompuServe boards didn't have this, but a decade later any decent Usenet client did), I'd never see that troll again. Most Usenet newsgroups had a regular cast of characters, and over time you'd get a handle on each personality. If someone posted a lot, you'd see a lot of their stuff (unless you'd filtered them). If someone posted very little, you'd see very little of their stuff. Nobody was tracking engagement. I'm pretty sure no Usenet clients back then were even capable of tracking clicks and views.

Things are very different today, living in the attention economy where every click and hover is tracked, measured, and optimized. In the attention wars, sensationalism wins. Outrage wins. Even if you mute a particular voice, the algorithms will surface ten more like it because that's what other people around the world are engaging with. It's a system designed to reward the superficial and punish the deep. I've lost count of the number of times I've seen a LinkedIn post with 8 carefully-considered and painstakingly-crafted paragraphs stall at 100 views, while another post with 8 cute-yet-hollow carousel slides receives hundreds of likes and thousands of views.

The advent of LLMs such as ChatGPT has only deepened this dynamic. Trying to hook readers but don't have a story to share? Invent a story. Need some business insights from a field where you've never worked? It's all just a prompt away. I want to be clear about one thing: I am not hating on LLMs. In fact I plan some future blog articles to talk about the ways in which I am using ChatGPT to unleash creative ideas I would never have thought possible before. I am simply highlighting the confluence of two trends at this moment in history:

1. Hyper-optimized social media engagement algorithms that reward triviality and "hooky" content in an ever-deepening spiral down the intellectual toilet bowl, and
2. Hyper-capable LLMs that get better and better at creating fake content perfectly tailored to attract eyeballs and generate likes.

We have, collectively as a society, seemingly decided this is the highest and best use of these incredibly powerful tools we've created.

Well, I'm out. After more than a decade on LinkedIn, I'm done trying to crack the code. My best ideas and my best writing will be found here, on this blog, from now on. I'd rather attract twenty thoughtful and patient readers here than chase engagement on a mainstream platform. I'll still maintain a presence on LinkedIn and BlueSky, and if I create video content maybe I'll put it on YouTube, but I'll treat them as outposts, not a home base.

In the coming years I believe personal sovereignty and authenticity will be emerge as the last things standing in our increasingly AI-mediated world (future blog post spoiler alert). Here's where I build my castle.