My stupid Twitter bot, @grauniad, got its first complaints during the last couple of months. One of them expressed frustration that whoever was behind the account persisted in mangling the Guardian’s every tweet despite the result manifestly not being all that funny. This critic was under the impression that some poor fool was running the account by hand and was still incapable of coming up with anything amusing, a misunderstanding that I think is far funnier than anything the bot has ever tweeted.

The other complaint was a bit weightier: it came from someone who felt it was inappropriate for the bot to mangle tweets about the Sandy Hook shootings. I’m not sure whether this critic had the same mistaken idea about how the account worked, or why he hadn’t taken issue with any of the other tragedies @grauniad had tweeted about, but I don’t like to see people upset, so I gave it some thought.

I don’t moderate @grauniad, and I don’t intend to start. As well as being a lot of work for little gain, I think it would be self-defeating. At the moment, the account is a mindless automaton, and being upset by what it produces is rather like getting annoyed at your Scrabble rack if it calls you ‘dickhead’ for pulling out too many tiles. To my mind, moderation by a human can only increase its capacity to offend.

This only holds if seeing the bot’s tweets is opt-in, which it generally is, because that’s how Twitter works. The exception comes when @guardian mentions someone, usually the author of an article. I don’t want an unpleasant @grauniad tweet launched into the mentions of someone who didn’t ask for it. So from now on, @grauniad has forgotten how to use mentions, and will just tweet un-@ed usernames in whatever mangled form it decides upon. That means the only way you can see an upsetting tweet from @grauniad now is if you asked for it – in which case it’s your fault – or if someone else showed it to you – in which case it’s their fault.

Sadly, this change won’t make it any funnier.

Categorized in: