Why the rise in ‘algospeak’ should worry you

Its been a while since I wrote this (it was about this time last year) but everything I said here remains true. Its just that some of the 'algospeak' might have changed.

What on earth does it mean?

On their own, 'cryptolects' such as 'polari' and 'leetspeak' are not overwhelmingly concerning – they simply act as a way of encoding someone’s language so only people within their social group can understand. Most, like cockney rhyming slang, are used to escape the detection of law enforcement. However, ‘algospeak’, as it has recently been coined, is the beginning of a dangerous trend.

‘Algospeak’, a portmanteau of algorithm and speak, has been rising in usage on certain social medias, particularly TikTok. It’s used as a way of preventing algorithmic censorship. Instead of risking key information being hidden by an AI, people are instead self-censoring. It does show how people’s language is evolving around censorship, but it also more concerningly shows the lengths that some companies will go to in order to only have ‘palatable’ content on their websites. This can mean governmentally enforced censorship online, particularly surrounding certain events. However, it can also be completed solely by the company themselves, and instead be used to prevent potentially educational content from making it to the masses. This is not necessarily a bad thing on its own, as it can prevent harmful content, such as white supremacist speeches, from making their way towards young people, but it can also stop people viewing content around certain topics that are not so harmful.

So how does this affect me?

Well, simply put, if you aren’t careful, it’s likely that your language online will be censored too. TikTok is one of the biggest platforms that you can see this happening in real time. But, if you aren’t the greatest fan of TikTok, then you can also see it on the big three – Facebook, Instagram, and Twitter. As much as these platforms talk about preserving free speech, they also attempt to actively restrict harmful content from the masses. Ever see one of those fact checking labels on social media under something about Covid? That was put there by an algorithm. Fact checking algorithms on social media are incredibly beneficial. We aren’t here to talk about those ones though, rather the ones that censor harmless content. These why this cryptolect is forming, although Covid misinformation also benefits from it.

Obviously, not everyone uses ‘algospeak’, that’s why you still see people mentioning taboo topics publicly. Alas, it is not always the best option to speak plainly. One unfortunately common example of this is ‘lesbian’ which is frequently censored online due to its sexual connotations. However, all that serves to do is stigmatise the community. That’s why you might see ‘le$bian’ or ‘le dollar bean’ online. Its so the algorithm doesn’t see it as bad. ‘Death’ and ‘die’ is also a word that has been censored, and so ‘unalive’ and ‘unalive’nt’ are used instead. It might look absurd, but it is a way of making sure that educational content is reaching the right people. I’m sure we’ve all seen posts at some points about how only women’s nipples (or ‘nip nops’) are censored online, so posts about how to self-check for breast cancer are made using a man’s. This has been happening for years, and whilst some social media sites are rebelling (Tumblr and Reddit) many are turning to AIs and algorithms to replace jobs that people should be doing. There’s no way of telling how bad it could get, but keep an eye out for the next time that you see a word censored somewhere. It might just be a new example of ‘algospeak’