我不再喜欢上网了。

I don't enjoy the Internet any more

August 17, 2026

I grew up on the Internet. I spent a significant proportion of my childhood, and later my adult life, on message boards, blogs, social networks, chatrooms, and other online communities. I made friends, found jobs, and was exposed to so many new concepts thanks to the Internet. Many of my core beliefs were sharpened by ideas and discussions I followed online. It’s been a huge part of my life.

But today, the joy is largely gone. This wasn’t a sudden change: my experience with the Internet has been getting worse over many years. As it became increasingly monetized, thoughtful knowledge sharing and new ideas started being replaced by an ever-growing volume of attention-seeking, low-quality content. And today, largely thanks to LLMs, the ratio is so skewed that it doesn’t feel worth it.

I’m reminded of the video game Disco Elysium (spoilers ahead). In the game, the world is made up of distinct regions called isolas, separated by a nothingness called the pale. The pale is a gradually growing area that degrades information. Venture far enough into it and you’ll be left with a damaged mind. The pale covers most of the world, and some believe that it’ll continue to expand until it envelops everything. This sounds a lot like AI-generated content! On the Internet, we still have a few isolas here and there, but the pale is growing faster than we can handle.

I’ve occasionally managed to find isolas of interesting people on social media like Bluesky, but I’m increasingly skeptical of short-form content. Learning about the world through 500-character snippets, or 30-second videos, is not a serious way to engage with knowledge. It triggers our worst instincts in terms of tribalism and groupthink and makes us intellectually lazy. This is true of short-form content whether you’re experiencing it on Instagram, TikTok, X, Mastodon, or something else.

Longer-form content like blogs isn’t safe, either. I’ve stopped reading Hacker News because too many stories on the front page are clearly written by an LLM. The writing is bad, and while there might be an interesting idea in there, it’s equally likely to be empty filler content. The effort of writing an article used to indicate that the author felt they had something worth saying. That’s no longer true. If the author couldn’t be bothered to write it, why should I bother reading it?

The final nail in the coffin was work. It’s incredibly hard to get to know the people you’re working with when everyone is slinging LLM outputs to each other. And this isn’t some anti-AI screed — I’m all-in on agentic software engineering! My issue is with communication more broadly. When you have no sense of what your coworkers believe, or how they think, all the joy and connection of collaboration disappears. I hope that we’ll collectively decide that it’s rude to pass off LLM writing as your own, but I wouldn’t bet on it.

The incentives on the Internet have been skewed for a long time, but there were always isolas of lovely people and high-quality thinking. These isolas are now much harder to find in the pale of slop, and I don’t see a way back.

Get new posts delivered to your inbox.