Algospeak: The Secret Language of Social Media Creators

Algospeak: The Secret Language of Social Media Creators

About two months ago, I really got into TikTok with my other online persona, The Parenting Patch. In addition to being a linguist, I am also a mother, homemaker, and homeschooler. I joined TikTok last fall with my Form-Function Grammar website, but home renovations took over and creating language content took a backseat for a bit. I plan to get back to creating my grammar videos that accompany the workbooks in my grammar series soon.

In the meanwhile, I have been working on decluttering my house and getting my home back into order. I joined TikTok as The Parenting Patch because I felt inspired by the cleaning and organization videos. I also started making my own. Then, as happens on TikTok, some other genres of videos showed up in my feed. One was a comedian who talked about fertility fraud. Another was a drag queen who shared disturbing DMs that they received. I quickly noticed an interesting phenomenon: Creators substituting certain words with other words and phrases. For example, instead of gun, creators said boom boom stick. Instead of sperm, creators used swimmers. Instead of kill or dead, creators used unalive or unalived as replacements. I was intrigued by the phenomenon, and posted a tweet about the topic:

I learned that the term for this phenomenon was algospeak, a blend of algorithm and speak. Wiktionary defines algospeak as a “form of cant for evading social media content filters.” Cant refers to the jargon or slang used by a specific subgroup, in this case, social media content creators. Algospeak seeks to thwart the language filters of the algorithms of social media platforms.

Algospeak is not a new concept. In the early days of the internet, people used acronyms and abbreviations to communicate with each other in chat rooms and online forums. I am old enough to remember a/s/l, which was an abbreviation for “age, sex, and location.” As social media became more prevalent, users began to use emojis, hashtags, and other symbols to express themselves and to evade content filters.

However, algospeak has become more sophisticated in recent years as social media platforms have become more aggressive in their efforts to filter out certain types of content. For example, platforms like Facebook, Twitter, and Instagram use algorithms to identify and remove hate speech, graphic violence, and other types of offensive content. Algospeak allows creators to avoid these filters by using coded language that is not easily recognizable by the algorithms — at least until the coded version is added into the filter. Unalive instead of kill, swimmer instead of sperm, and boom boom stick instead of gun are just a few examples of this type of coded language.

In addition to avoiding content filters, content creators may also use algospeak to avoid triggering sensitive topics or to create a certain tone or style in their content. For instance, a creator might use the phrase spooky season instead of Halloween to avoid triggering any cultural or religious sensitivities. Similarly, creators might use the phrase hot leaf juice instead of tea to create a playful and whimsical tone in their content.

Here are some examples of algospeak:

  • “Accountant” instead of “sex worker” or “Only Fans creator”
  • “Addy” instead of “address”
  • “Alti” instead of “alcohol”
  • “Bakers” instead of “drug dealers”
  • “Beat stick” instead of “baton”
  • “Big bang toy” instead of “bomb”
  • “Bite stick” instead of “weapon”
  • “Blink in lio” instead of “link in bio”
  • “Boom boom stick” instead of “gun”
  • “BxMx” instead of “bomb”
  • “B0ng” instead of “bong”
  • “Cornucopia” instead of “homophobia”
  • “C@sh” instead of “cash”
  • “C0c@!n3” instead of “cocaine”
  • “D1v0rc3” instead of “divorce”
  • “E-p1lls” instead of “ecstasy pills”
  • “Firecracker” instead of “explosive”
  • “F0llicle friends” instead of “pubic hair”
  • “F@sh1on” instead of “fashion”
  • “F&$k” instead of “fuck”
  • “Glorp” instead of “vomit”
  • “Grape” instead of “rape”
  • “Greenery” instead of “marijuana”
  • “G1ft” instead of “gift”
  • “H1gh” instead of “high”
  • “H@x0r” instead of “hacker”
  • “I3gal” instead of “illegal”
  • “Janes” instead of “heroin”
  • “K!ll3r” instead of “killer”
  • “Le dollar bean” instead of “lesbian” or “le$bian”
  • “Leg booty” instead of “LGBTQ+ community”
  • “L0L” instead of “laugh out loud” or “LOL”
  • “L3m0n” instead of “marijuana”
  • “Mukbang” instead of “food porn”
  • “M0b” instead of “mafia”
  • “M0d” instead of “modify”
  • “Nip nops” instead of “nipples”
  • “N0ob” instead of “newbie”
  • “N@ked” instead of “naked”
  • “Onesie” instead of “lingerie”
  • “Opposite of love” instead of “hate”
  • “Ouid” instead of “weed”
  • “Panini” instead of “pandemic”
  • “Pepper popper” instead of “pepper spray”
  • “Piece” instead of “gun”
  • “Pow pow juice” instead of “bullet”
  • “P0rn” instead of “pornography”
  • “P1ff” instead of “puff”
  • “Qw33n” instead of “queen”
  • “Rooty toot toot” instead of “firearm”
  • “R@ve” instead of “rave”
  • “SA” instead of “sexual assault”
  • “Saltines” as in “white people”
  • “Seggs” instead of “sex”
  • “Sh1t” instead of “shit”
  • “Slice and dice” instead of “knife”
  • “Snip snap tool” instead of “scissors”
  • “Spicy” instead of “sexy”
  • “Spicy eggplant” instead of “vibrator”
  • “Swimmer” instead of “sperm”
  • “S3x” instead of “sex”
  • “S@fe” instead of “safe”
  • “Tea” instead of “gossip”
  • “Thunderstick” instead of “gun”
  • “Tiddy” instead of “titty”
  • “T3rrorist” instead of “terrorist”
  • “Uggs” instead of “ugg boots”
  • “Unalive” instead of “kill”
  • “Unalived” instead of “dead”
  • “V@pe” instead of “vape”
  • “Wavy” instead of “drunk”
  • “W33d” instead of “weed”
  • “W@rez” instead of “pirated software”
  • “XxX” instead of “sex”
  • “X0x0” instead of “hugs and kisses”
  • “Yt” instead of “white people”
  • “Y33haw” instead of “yeehaw”
  • “Z0mb13” instead of “zombie”
  • “ZzZ” instead of “sleep”
  • “1337” instead of “elite”
  • “3v@de” instead of “evade”
  • “4tw” instead of “for the win”
  • “8===D” instead of “penis”

I see a number of problems with algospeak. First, I do not like the idea of replacing correct terms with cutesy terms, especially in regards to human anatomy and physiology. As a parent, I teach my children correct medical names including penis, scrotum, testicles, vagina, vulva, and breasts. Groups that work to fight child abuse including the American Academy of Pediatrics support the teaching of correct medical names. Incorrect and cutesy names send the wrong message about the body and bodily functions. Sexual abusers use secrets including code words for body parts to hide their abuse. Words such as penis and nipples are correct medical terms. Inventing cutesy code words to bypass content filters is such a slippery slope. I also found news articles about the ways in which bad actors hide harmful material in plain sight using tricks like algospeak.

Content filters that cannot understand context are equally, if not more so, problematic. Speaking out against gun violence is not the same as threatening to use violence with a gun against someone. But content filters will flag both equally for the use of the words gun and violence. A history channel speaking about the death of millions who were killed by the black plague is not the same as an extremist threatening to kill people. But, again, a content filter will flag both because of the word kill (and death) regardless of context.

Algospeak terms are easily added to content filters. If I can look up lists of algospeak terms, then so can people and even AI in charge of the items screened for by filters. While seeing language change in real time is quite exciting for a linguist, where does the madness stop? And, honestly, the people with bad intentions will always find a way around the filters. Unconditional filtering without considering context harms access to information and free speech. Consider the ways in which a filter on the word breast impairs access to content on breast cancer or breastfeeding.

Another issue with algospeak is the creation of confusion and miscommunication between individuals. Using different words to refer to the same thing can cause misunderstandings and even lead to conflicts. Additionally, using algospeak can reinforce negative stereotypes and stigmatization of certain topics. For example, using seggs instead of sex may suggest that sex is inherently bad or shameful, perpetuating a negative connotation. Finally, algospeak can also lead to a lack of accountability and responsibility for the words people use online. By being able to hide behind cutesy or coded language, individuals may be less likely to take responsibility for the impact of their words on others.

Although beneficial for creators looking to bypass censorship and evade content filters, algospeak also poses a number of issues. Coded language can foster a sense of community but can also be used to manipulate followers. As social media platforms continue to evolve, creators and platforms must approach this phenomenon with caution and responsibility. The use of algospeak can lead to the dissemination of harmful content including hate speech and misinformation. Additionally, coded terms can cause confusion and miscommunication as different groups may use varying coded language to refer to the same thing.

Language changes. As a linguist, I find the ways in which language changes fascinating. I love delving into the changes that have happened in the English language. For example, I have researched the history of the pronoun you and I am planning an upcoming post about the singular they and its possible future. Language change is awesome and inevitable. Algospeak even reminds me a little bit of Cockney rhyming slang. But something about the intent to manipulate of algospeak gives me creepy vibes. As I said in my original tweet, this seems like a dangerous slope.

See also I Type How I Sound: E-Language and Phonetic Spelling.


Kao, E. (2021, January 21). The Dangerous Seduction of the Algorithm in Social Media. Wired.
Levine, A. (2022, November 11). TikTok Private CSAM: Child Sexual Abuse Material. Forbes.
Lyon, J. (2021, January 28). How social media algorithms reinforce extremist views. BBC News.
Madrigal, A. C. (2021, March 3). How Algorithms Can Sustain Inequality. The Atlantic.
Newton, C. (2022, January 20). ‘The ultimate weapon for mass persuasion’: Facebook’s ad algorithm is creating a ‘social problem’, researchers say. The Guardian.
Persaud, C. (2022, March 18). How TikTok’s algorithm can manipulate our behaviour – and what we can do about it. The Conversation.
Roose, K. (2021, April 8). The Facebook ad boycott didn’t work, but that’s not the point. The New York Times.
Thompson, D. (2022, April 8). Internet ‘algospeak’ is changing our language in real time, from ‘nip nops’ to ‘le dollar bean’. The Washington Post.

One Comment

Leave a Reply

One Ping

  1. Pingback:

Leave a Reply

Your email address will not be published. Required fields are marked *

Lies Your Grammar Teacher Told You: English Verb Tense

Lies Your Grammar Teacher Told You: English Verb Tense

Can Be a Vowel in English: Letters and Graphemes

Can ‘w’ Be a Vowel in English: Letters and Graphemes