From Hollywood to Hallucinations: Is AI the Final Act for Regional English Dialects?
In the 1950s, something curious happened. As post-war economies boomed, America exported more than just Coca-Cola and cars—it shipped an entire culture. Movies, TV shows, and music began to fill living rooms and cinemas worldwide. Marilyn Monroe smiled; Elvis shook his hips; people in places as far-flung as Tokyo and Timbuktu started humming songs in a language they did not even speak. English became cool, but not just any English—American English.
Fast-forward seventy years, and the dominance of American English has evolved from the crackle of vinyl and the glow of black-and-white TV to the digital chirp of Siri and the cold clarity of AI chatbots. Where Hollywood once taught the world to say, “gee whiz” and “howdy,” artificial intelligence could be the final frontier in homogenising how we speak, write, and even think. But is this the start of a linguistic monoculture? And if so, is it time to panic—or laugh?
When Entertainment Became an English Teacher
English became the de facto language of global entertainment in the 20th century for two reasons: accessibility and influence. The Beatles did not just sell records; they sold phrases like “Can’t Buy Me Love.” Meanwhile, American movies highlighted a world where “action” and “cut” were not just words—they were commands in the universal language of storytelling.
But here is the catch: even as English-speaking media spread, it did not flatten the diversity of the language. You could watch a John Wayne western in Germany, and still come away with “Howdy, partner!” rather than “Hello, mate!” Likewise, regional English accents—Cockney, Scots, Australian, Jamaican—persisted, even flourished, often shaping local music and culture in return. For a long time, English was adaptable, elastic, and gloriously messy.
AI and the Tyranny of Standardised English
Then came artificial intelligence.
Modern AI systems, from ChatGPT to Alexa, are trained on enormous datasets of text, mostly in English. And not just any English—largely American English. It is the language of Silicon Valley, tech companies, and international business. While these systems are remarkably multilingual, when it comes to English, they default to rules and spellings that would make any Midwesterner proud. “Color,” not “colour.” “Sidewalk,” not “pavement.”
Unlike Hollywood, which embraced linguistic quirks for character development, AI systems crave uniformity. If you prompt an AI with “What’s a braw way tae cook tatties?” you’ll likely get a confused response or a suggestion that you meant to say, “What’s a great way to cook potatoes?” The machines, it seems, prefer standardisation —perhaps to make their jobs easier.
And that raises an interesting question: if AI tools dominate more of our lives, could they subtly teach us to conform to their preferred version of English?
Linguistic Creep: The AI Effect
Language is not just a tool for communication; it is a reflection of culture, identity, and history. Scottish dialects like Doric or Glaswegian, for instance, carry centuries of tradition in their phrases. Will those traditions slowly fade if global interactions increasingly require “AI-friendly” English?
There are already signs of this linguistic creep. Consider how we talk to our devices. When was the last time you used colloquial slang or a regional idiom in a Google search? Probably never. We have been quietly training ourselves to speak in stripped-down, grammatically correct sentences that Siri or Alexa can understand. It is transactional, not conversational.
The risk, of course, is that this process might narrow the rich tapestry of English into something blander. AI does not “understand” context like a human does; it thrives on predictability. And predictability is the enemy of linguistic variety.
A (Humorous) Look at the Future
Let’s imagine it’s 2050. AI assistants are ubiquitous, guiding everything from grocery lists to interstellar travel. Conversations might go something like this:
Human: “Hey, AI. How should I take care of my bairn’s teething?”
AI: “Did you mean ‘baby’? Also, the word ‘bairn’ is deprecated in our dictionary. Please update.”
Or worse:
Human: “AI, could you order me a flat white and some biscuits?”
AI: “You meant ‘coffee with milk’ and ‘cookies.’ Adjusting your order to reflect this superior terminology.”
Will we all eventually end up speaking like tech manuals?
Why the Future Isn’t (Entirely) Bleak
Before we declare war on our Roombas, it’s worth noting that language is surprisingly resilient. Efforts to standardise English are nothing new. The Victorians tried it with dictionaries. American spelling reformers tried it with “honor” and “plow.” And yet, English is still delightfully chaotic.
AI might encourage us to use simpler, more universal forms of English in certain contexts, but people love to rebel against linguistic authority. (Have you seen how teenagers text?) Dialects and accents evolve, resist, and adapt in ways that often surprise even the most seasoned linguists.
Conclusion: A New Chapter in the Story of English
AI won’t destroy regional English, but it will shape how we use the language. The real danger isn’t losing our words—it’s forgetting how to laugh at ourselves as we struggle to make machines understand us. After all, what’s more human than refusing to standardised?
So, the next time you’re tempted to scold your voice assistant in broad Scots, go for it. You might confuse the machine, but you’ll remind yourself—and everyone listening—that English is far too big and beautiful to be flattened into one accent, one spelling, or one dialect.
Because at the end of the day, AI might be smart, but it doesn’t stand a chance against the creativity of humanity’s words.
What do you think—are you team “local dialect” or ready to welcome our AI overlords?
References:
Crystal, David. English as a Global Language. Cambridge University Press, 2003.
Ghosh, Pallab. “How AI Is Learning English—and Teaching It Back.” BBC News, 2023. https://www.bbc.co.uk/news/business-65849104.amp