When I go on LinkedIn to post, I never read beyond the first post. Usually, the first one I see is something someone I know liked. The next, which I only get to see half of, is an ad. I exclusively use the web version on my computer because I don’t want to forget that I am working when I browse social media.
They’re trying to emulate the academic incentive system. But as we know from all the teachers on LinkedIn complaining about how their students are cheating on exams with AI, such a system will only turn their platform into a septic tank of AI slop.
The entire academic incentive system is built around encouraging students to prove that they know something. The actual subject matter may not actually matter, but the system basically rewards students based on how well they can demonstrate their knowledge of a certain concept or action. So, for social media platforms that take themselves too seriously—like LinkedIn, Substack, and X (maybe? It's been a while since I was there)—they borrow this same academic incentive structure. They attempt to reward creators for content that, in one way or another, functions just like a high school exam: Do you or do you not know what you’re talking about?
That is why the AI slop on LinkedIn and Substack, especially, is just lame. It feels like plastic vomit. Meanwhile, similar slop on platforms like Reddit feels a lot more like organic garbage. Overall, this kind of system that optimizes for some kind of objective "Truth" or standard actively discourages more speculative and creative kinds of content creation.
That is why I am building Pyepa: the algorithm that is just the best. Really, the very best (in Trump’s voice). If you know people with money, connect me—my company is looking for funding.
Otherwise, if you know whether LinkedIn is hiring for their product team, send the job post my way. I have some ideas on how they can make their platform better for people like me.