As A Language Teacher, AI Is Both Brilliant And Dangerous – Here’s Why
Updated: Jan 22, 2024
Written by: Megan Miller, Executive Contributor
Executive Contributors at Brainz Magazine are handpicked and invited to contribute because of their knowledge and valuable insight within their area of expertise.
Ever since the advent of the Internet, foreign language students have been poised to make life a little easier and get help where they can: either by using tools like Google Translate or, more likely these days, having a tool like Chat GPT create something in their target language.
The Internet, among many other things, is a very powerful tool that can democratize the world. It democratizes learning, meaning no one has to feel limited based on the ability to pay for college tuition or carve out time for a school schedule. While Internet access still lacks equity, it does help connect humans to humans.
Shiny happy brilliance
While artificial intelligence (AI) definitely isn’t new as people have been using it for years if not decades, having such a powerful tool is such close proximity to learning and speaking is still novel. This means that language access is now widely available: using prompts, a student can simulate a conversation (one of the cornerstones of practicing a language), and the AI tool can immediately give a variety of responses, translations, or prompts.
Additionally, the immediacy effect is already built-in to our brains, making sure that we prefer an instantaneous reward over a future-oriented solution. This makes AI very sexy in the language space: simply rely on a tool that doesn’t need to rest, doesn’t have working hours, and can spit out an answer within milliseconds.
Using AI tools to learn, practice, or use a language also means that anyone: young, old, formally educated or not – can travel and communicate, hopefully also connecting and relating to others where English isn’t as popular or widely spoken. This is great news for everyone’s time and wallets: no longer is there a years-long battle to have a simple conversation; the conversation can be aided and translated by AI.
With great power comes great oversight
Anecdotally, however, AI is rife with mistranslations. From an Italian football game in which the term “punizione” was taken out of its football context, to an immigrant family being terrified that a COVID-19 booster was a “disparo” (shot from a gun) instead of a “vacuna” (vaccination), the risk of offending someone by using AI is high.
As much as the earlier examples pain me personally, a lot of our personal or anecdotal uses of AI to translate sentences don’t have much inherent risk. If you say “dime” instead of the more formal “dígame” (tell me) command in Spanish, for example, the relationship can still recover. Using a tool to translate something directly and perhaps literally can mean being met with a blank (or worse, terrified) face. Not similar to images, AI-translated sentences still feel stilted, which is how us teachers always seem to know when something has been taken through a direct translation tool.
The translation and interpretation spaces are still not widely regulated. When you put an unregulated AI tool on top of that, the risk of mistranslation or misinterpretation grows. As language professionals (and there are many), we’ve spent years of our lives making sure that the cultural nuances and meaning don’t get lost in translation. Regional idioms and dialects, not to mention the different types and tones within each individual text, must be taken into account. AI simply does not have a grasp on the context and nuance of word choices to create a fair translation.
Speaking of risks, the big issues with AI in language come with higher stakes. In a business context, do you want to accept the reputational risk of not hiring a professional translator? In a diplomatic context, would you dream of not having a credentialed interpreter? (If the answer is “no”, check out ProZ.com or the LITranslator hashtag on LinkedIn to find your best fit).
Perhaps the biggest danger of them all
Personally, the more I read about, utilize, and think of AI in the language space, the more divided I feel. On one hand, tools, technology, and innovation will continue to proliferate. Perhaps AI will become more regulated, with the proverbial brakes pushed once in a while to ensure that the source data is as fair, unbiased, and correct as it can be. Simply sticking our heads in the sand or flat-out refusing to utilize a tool means that we’ll be left behind. Attempting keywords and sentences is better than miming our ways across continents and foreign lands.
On the other hand, at its core, language is how us humans make sense of our feelings and emotions. We put labels on them and create words with specific sounds, ready to speak, write, record, sing, hum, or yell. Our language is just as informed by our popular culture as it is by the societal norms, acronyms, and expectations in all of our worlds: whether they are white collar or blue collar, left or right, traditional or contemporary, remote or in-person. If we offload some of that creation to an insentient tool and feed it information related to historical events, music, and literature; as well as common subjects, verbs, and nouns: would we even recognize the sentences when it’s done?
To me, the biggest danger is not having AI tools. It’s how we manage them – and how we manage ourselves.
Follow me on Instagram, LinkedIn and visit my website for additional motivation and education tips. Reach out to let me know you’ve read my article; I’d love to hear from you!
Megan Miller, Executive Contributor Brainz Magazine
Megan Miller is a leader in Spanish <> English teaching and bespoke habit-based language learning. Ever since discovering the worlds beyond words as a child, Megan has dedicated her efforts to mastering Spanish, English, and how to create lasting habits to improve and maintain language skills. She is the CEO of Aprovechar Language Solutions whose mission is to empower those in need of a bilingual voice.
