David Bushell

RSS Club #010: Chatbots

Read the original on dbushell.com ↗

This is an RSS-only post, thank you for subscribing :)

My RSS-only posts have a permalink for those who prefer reading from my website. That link is hidden and unindexed. URLs can be linked to but in the spirit of “RSS Club” I’d prefer if my RSS-only posts were only linked to from other RSS-only content.

Warning: this post contains heavy slop-talk!


I’m trying my best to leave them behind but everywhere I go is inundated with chatbot bullshit. Below are some scattered thoughts on the topic that I’m keen to leave behind. So I’m dumping them on you! Aha!


I’ve noticed a change in how people comment on my blog posts. I think the reason is they are only reading a chatbot summary. They question stuff that was meticulously answered in the post. They misinterpret quite obvious jokes as serious statements. They jump to extreme conclusions. They latch on to small tangential details that bear little consequence to the main content. They fail to understand the mood and tone of my post.

I don’t always know if a commenter read and understood what I wrote. It’s quite apparent though when they’ve read something that I never wrote.

I’ve been blogging long enough to know when I failed to make myself clear. Nope, this is a new phenomenon. There is a distinct change in how a few people comment. Far from everyone, and it’s those “normal” comments that convince me there is something off with the others.

A comment based on a wrong chatbot summary is far more discouraging than any negative or insulting criticism. I have a thick skin for the latter.

I’m seriously considering getting my website banned from Lobste.rs for this reason. On sites like Hacker News the dumb comments get lost in the noise. Lobsters is small enough that the bullshit sticks. Suspected comments based on a chatbot summary have a habit of arriving earlier than others. It’s exhausting to deal with. Maybe I just have to ignore it?


One of life’s small joys is referring to all chatbot output as “slop” and “vibe-code”. It really riles people up because their chatbot usage is different. You must understand this. They can get their chatbot to sing, it’s others who are failing to finesse.

Many in the web industry have spent the last year (or more) trying to predict the output of their chatbot. The fact that non-determinism† is fundamental to chatbots working at all doesn’t register with slopaholics. “You’re prompting it wrong” is the entire conversation.

† Technically PRNG but it’s sufficiently random to mangle human knowledge.

Everyone likes the taste of their own slop. No one seems to grasp that perhaps they too have stopped learning and started renting a sorry excuse for brain activity from a bunch of fash-tech broligarch grifters. These billionaire psychopaths invented “AI” by beating all meaning out of the term. If you watch Ghost in the Machine you’ll learn that the term “Artificial Intelligence” has deeply dehumanising origins. However, most of us think of the cool future promised by sci-fi books and movies. Regardless of what “AI” means, it’s quite the indictment that people have settled for the saddest most pathetic definition.


Once distinguished web developers are now advocating for cognitive surrender to Google and their “Modern Web Guidance”. These people have all but killed web education — but don’t worry, Google is here! They have the gall to cry foul when you dare challenge their agenda. “Why can’t we have a normal conversation?” they ask insincerely. Because they’ve chosen to ignore all moral and ethical issues. They killed web education and now they’re trying to educate an f-ing chatbot. That ain’t normal.

The deskilling of the web industry and destruction of related professions has been rapid. I won’t be forgetting those who participated.

The quality of chatbot output is the least of my concerns. Although it does reveal how low the bar is (lower than I assumed necessary). And how sorely lacking in taste vibe-coders are.


The funny thing is that vibe-coders can’t escape the negative effects of all this slop. They too see the unreadable emails, the worthless content, the broken websites and apps. And they have the audacity to moan about it! The world is burning around them but their fire is under control.

When the AI industrial complex collapses so will the economy. If Anthropic and OpenAI are allowed to IPO, the door is open for “safer” investment to flow in. That significantly increases the likelihood of bailouts. Whatever happens, it’s inevitable that more lives will be irrevocably ruined. Vibe-coders will be quick to point fingers at everyone but themselves.

“It was just a tool!” they’ll plead.


Thanks for reading! Follow me on Mastodon and Bluesky. Subscribe to my Blog and Notes or Combined feeds.