AI Researcher

When it comes to recent major influences on education, there are few things as prominent as generative AI chatbots. Comprising tools like ChatGPT and Google’s Bard, generative AI has swept through schools across the country at an unprecedented rate. Their ability to find solutions to problems and create written works that can seemingly bypass plagiarism checkers (N.B. there are now AI-powered tools that can detect AI-generated work) has forced curriculums to rapidly evolve with this new “AI revolution.” It’s understandable, then, that the terms “ChatGPT” and “AI helper” have had a suboptimal reputation among teachers and instructors. But ironically enough, many educators have themselves begun to use AI tools, and for good reason as well! Many tasks required by instructors are quite tedious. From grading multiple choice answers to writing trivial emails, these duties are often interpreted as time-consuming and detract from the time that could be spent communicating with students and designing quality lesson plans. The former can be especially helpful, as much of the time, grading can be little more than a brain-numbing tedium. And given the customizable nature of AI tools, they serve as the perfect solution, especially for groups like world language instructors.

Before jumping into applications, it is first important to understand the basics of why AI tools in particular are so powerful. Unlike traditional modes of technological tools, modern AI models implement what is known as a neural network, a complex, multilayer set of connections between nodes (or “neurons”) whose nonlinearity can theoretically help identify almost any type of pattern or connection. These networks offer a sense of malleability to AI models, one which can enable the model to fit any mold required by the user. This allows for customization. For instance, if teachers properly query ChatGPT for particular grading help (using something called a “prompt”), the AI chatbot could, in theory, perform whatever grading task was asked.

So why is this particularly useful for language teachers? Well, unlike for other subjects, automatically grading language assignments often requires multiple tools for multiple levels of processing: first would have to translate the text into something the particular program would understand (e.g., English for programs designed around the English language), and only then could it perform the grading task at hand. Generative AI condenses this into one step. Chatbots like ChatGPT are trained on information scraped from all over the internet. As a result, users can expect these models to understand any language well documented online. So, the chatbot can effectively interpret the language in its native language, and depending on the prompt, output something in the same language or an entirely different one.

Michael Cai

AI as a Foreign Language Assignment Grader

The 21st century has seen a recent boom in AI-related products and discoveries. From new transformer-centered model architectures to Dall-E to ChatGPT, new innovations are now rapidly expanding the horizon of what is possible for artificial intelligence and its applications. Recently, natural language processing (commonly referred to as NLP) has been in the forefront of this so-called “AI revolution,” and for good reason too. NLP products and services (ChatGPT and other generative AI chatbots, for instance) have brought to many people’s attention the practical, everyday uses AI can have. Moreover, creations centered around NLP have fundamentally altered — and will more likely than not continue to alter — many fields, particularly the field of education.

Although almost every educational subject would benefit a great amount from these innovations in NLP, this article will focus on applications in world language education. This is because NLP and current generative AI technologies pair very well with themes present in foreign language learning. The process of learning a language strikes a good balance between black-and-white, almost “binary” logic and fluid writing which offers more leeway: firstly, a main aspect of picking up a new language is applying logical rules (such as grammatical laws) that have a relatively well-defined set of correct and incorrect answers. Indeed, this aspect is similar to qualities present in math, science, and other STEM subjects. That said, the language classroom rarely explores the nitty gritty of logical relations present within these fields, ones that can be quite complex and may result in glaring issues for an AI model not trained to solve a specific problem. Language learning offers leeway in this sense because within the context of fluid writing, logical errors (which are bound to happen for any complex and nuanced subject) would be less detrimental.

Though AI technology has definitely impacted language learning, technologies like Quizlet or Duolingo have made large ripples in the field prior to any mass-scale wave of AI innovation. So, then, what does the future look like?

What developments in AI mean for Language Education

Ever since the release of show-stopping innovations like Open AI’s new generative AI chatbot, ChatGPT, or the company’s original artwork generator, Dall-E, artificial intelligence has been on everybody’s minds. Machine learning is still a relatively young and emerging field, and most of these mainstream technologies are all fairly new, so it’s no surprise that people have begun to gravitate towards this AI trend — after all, it shows a lot of potential. As a result, from cooking to writing essays, there have been countless AI-related topics online resources have covered, each one offering ways to use AI to its maximum potential. A vast majority of these resources concern AI prompting, the act of engineering queries to feed into AI chatbots like ChatGPT.

Now, prompting can be used in an infinite number of ways. This article covers a small pool of those methods, specifically how the act can be incorporated into language learning. Language learning is a field that AI is bound to greatly impact. In fact, apps like Duolingo have already incorporated some AI-based elements. AI and languages indeed go hand-in-hand. This is because the bulk of language learning stems from memorizing and being extremely comfortable with a set of well-defined rules. Programs like AI are quite good at identifying patterns — and in turn, rules — so they naturally excel at language learning and language translation (in fact, the field of natural language processing — teaching computers how to understand natural human languages — now mainly concerns machine learning). With this in mind, here are some of the best prompting tools to use when navigating the potentially daunting task of learning a new language from scratch.

This first section encapsulates the tools that are used in more niche scenarios; they may be powerful, but not necessarily common. When it comes to a ranking structure, these would comprise the lower five tools.

Top 10 Prompts for Learning Languages

There’s a common perception that mastering a foreign language means being able to conjugate and decline words or being able to string together complex sentences in split seconds. There is truth to this; world language classes often adopt this heuristic of measurement, and foreign language proficiency tests frequently examine students’ ability to perform in these respects. The only issue is that when it comes to practical applications of a language, understanding and utilizing its grammatical rules only contributes a fraction of what’s necessary to be “fully proficient.” When it comes to interacting with a native foreign-language-speaking society, knowledge of culture can also be key.

As mentioned before, standard world language curriculums do not emphasize learning cultures associated with languages. Yet understanding cultural taboos, traditions, or dialectal variations equips students with the tools to formulate contextually appropriate sentences and form stronger connections in the process. So then why is culture often forfeited in the classroom? One of the main answers is a lack of resources and time. Languages are hard to learn — they require students to memorize large masses of vocabulary, sentence structures, and grammatical exceptions. And considering the limited amount of time each teacher has in a day to teach and concoct lesson plans, it is understandable that less emphasis is placed on cultural learning.

However, with recent developments in artificial intelligence, there is a solution. Generative AI chatbots in the form of tools like ChatGPT are practically catered towards this issue. These machine learning models are trained on vast amounts of data on the internet, making them highly educated on almost any topic thinkable. As a result, these kinds of large AI chatbots are customizable to a wide range of tasks, especially when it pertains to subjects well-defined online such as cultures of different societies. The streamlined prompt-output structure of these tools (i.e, users feed in a prompt (or input) and the chatbot spits out a response (or output)) are particularly useful for saving time, the main issue faced in this scenario.

AI As a Gateway to Cultural Learning in the Classroom

It’s no question that the world is delving deeper into a state of extensive globalization. Apart from the increased prominence of trade, diplomacy, and our world’s interconnected global economy, the twenty-first century seems to be the period in which cross-cultural exchange is starting to flourish at an unprecedented level. As the United States populace begins to show greater interest in other countries, there is a natural tendency for more people to engage with different languages. There is one complication with this, however.

Many of the languages popular with younger generations, such as Japanese and Korean, are not typically taught in the standard U.S. curriculum. Yet that doesn’t seem to stop language learners from pursuing their language of interest. In fact, according to a survey conducted by 

, only 31% of people who study foreign languages are forced to for educational purposes. The remaining 69% do it for independent growth or personal interest. It seems, then, that there is a large need for good language learning resources outside of the classroom. Tutors and paid educational or cultural organizations may provide high quality, attractive alternatives, but the reality of life is that many people do not have the financial resources or time to engage with languages in such a way. A much cheaper substitute is online resources: this means YouTube videos, articles, and online forums. However, internet videos and articles are rarely — if ever — truly incorporate interactiveness (an essential part of learning a new language), and online forums can be a tedious and inefficient means of learning a language — one where the student has little control over. But with recent innovations in the field of machine learning, there is now an effective, customizable guide perfectly tuned for the task of teaching new languages: AI chatbots.

By now, ChatGPT and other generative AI chatbots are household names. The basic idea is that the user can input a question, statement, or any type of text input, and the chatbot will output a response. It’s a fairly simple concept but highly effective when it comes to learning in an interactive and engaging way. Firstly, chatbots like ChatGPT do not just generate random replies. Before being released into the internet, these AI models were fed millions of input data — text scraped from all over the internet — from which they learned patterns and connections between different concepts. This means that generative AI models can recognize the user’s intention to an impressive degree and can identify the most statistically likely answer to the question. As a result, these chatbots are highly interactive and customizable based on the user’s inputs, or prompts. In addition, the data fed into these chatbots comprise high quality research and books, resources that the AI is able to keep in mind while generating responses. Therefore, responses are not only highly specialized towards whatever question users input, but they are also highly educated and reflect a strong understanding of the subject. The only caveat is that AI tools like ChatGPT 3 were created before 2023, and as such, the data they were trained on could be out of date. Regardless, the vast majority of information is still relevant, and as such, these chatbots still serve as amazing tools to use for language learning. Let’s see how this would work practically.

AI Chatbots as Study Plan GenerAI CHATBOTS AS STUDY PLAN GENERATORS

To put it succinctly, AI chatbots are great at creating study plans. During the training process, these large AI models have absorbed a multitude of data related to various subjects, study materials, and learning resources and processes. As a result, they not only know a lot about the subject at hand, but they are also privy to outlines for learning which are designed to maximize effectiveness. An example of how to use an AI chatbot in this way is provided below. The example, which was taken from ChatGPT, includes the user input and AI output.

How AI Can Facilitate Self-Studying Languages

The Best Online Language Teacher

It’s no secret that artificial intelligence is booming. Within less than a year, Generative Pre-training Transformers (GPT) have evolved from a research subject relatively unknown by the wider public to household names in communities all over the world. This might partially be due to the fact that the capabilities of artificial intelligence and machine learning have increased exponentially since the advent of a new AI model called the transformer, and more so in 2017, when researchers introduced the notion of capturing relationships between words in an input sentence (known as “self-attention”). However, one potentially greater reason for AI’s popularity is that with the focus on natural language processing, AI tools have become extremely useful to almost anyone — not just researchers.

ChatGPT and other generative AI-powered chatbots are the harbingers of this “natural language processing revolution.” Their ability to parse and comprehend input text at such an efficient pace lends itself to almost every field and every corner of daily life. One application in particular is the use of generative AI in the classroom, specifically language classrooms. The main purpose of world language classes is to promote the ability to parse through texts of different languages, so it’s only natural that generative AI could help. At the same time, there are a multitude of other ways in which AI can aid in communication — some which are unique to language classrooms and others which can apply to any educational or interpersonal setting.

Helping communicate topics of study to students is one of the more general usages of AI. In fact, for any subject that includes complex concepts, communication aid can be vital on the part of the teacher. One good way of conveying these ideas is through analogies, since they can offer a visualization for something that’s otherwise hard to grapple with. Since generative AI gathers and learns from basically all the information on the internet, it has the capability to be highly customizable to the user’s needs, and as such, can act as an analogy generator. In fact, an example analogy generation prompt can be found 

 (note that if you use the prompt for foreign language concepts, you will need to specify the language in the “concept” field).  Additionally, an example input and output can be seen below. This example was taken from ChatGPT since as of the time of this article, it is the leading generative AI tool. The prompt on Lowtech AI is easily migratable to ChatGPT (just click “Show Prompt” and replace the placeholder words in double curly brackets).

AI-based Communication Help in the Language Classroom

The recent explosion in AI-related products has permeated almost every field. Although areas such as biomedical research and engineering were the first ones to be impacted on a large scale, recent developments in machine learning have made the field relevant in everyday lives. Specifically, the creation and proliferation of generative AI resulted in tools such as chatbots. Examples like ChatGPT and other household AI tools have already found their way into the classroom of many schools. But more often than not, it has been interpreted as a negative force. After all, instructors frequently see AI as a powerful yet hard-to-be-traced device students can use to cheat.

That said, AI has much more potential than just a way to cheat. Ironically enough, it can serve as a major source of guidance for instructors. When it comes to researching, designing, and learning of materials for teaching, generative AI can actually be quite a useful tool. This article focuses on how AI can positively impact language teachers in particular, though it can be extended to other subjects. Language learning is unique in that it incorporates elements from both STEM (science, technology, engineering, and math) classes as well as humanities classes like English and history. Specifically, language classes can be known for possessing complex ideas and grammars that can take up an almost mathematical appearance. At the same time, the notion of writing essays and interacting with other people and societies is analogous to work done in the humanities and arts. It can be difficult to balance these two elements, and paired with the momentous amount of material language teachers are tasked with teaching, world language classrooms can often be the hardest ones to get right. As a result, they’re the ones most benefited by AI.

The main reason why AI can be such an annoyance when it comes to cheating is actually the same reason why it can be so effective when researching the best ways to conduct a desirable classroom environment. Generative AI chatbots are customizable and create unique responses. That means these AI tools can respond to almost any question with a helpful response, and that the response given will not be a piece of text found somewhere on the internet, but rather one uniquely generated by the machine learning model. One feature these AI chatbots are especially proficient in is outputting customized summaries (i.e., summaries of a topic that follow a certain set of conditions and restrictions). This is because generative AI models learn from a large dataset that consists of text from the entire internet, which allows the AI model to be aware of almost everything on the internet at the time it was trained. The caveat of this is that any model will always be a little behind on news and contemporary information. The AI learning process (known as training) often takes a long time for large models like OpenAI’s ChatGPT, so no matter what, there will always be a delay between what’s currently available on the internet and what generative AI models know (for reference, ChatGPT 3 is only aware of information up to 2022).

A case study on how AI can help language teachers

AI in the Classroom

It’s inevitable to make mistakes when learning a language. After all, every language is rife with intricate linguistic subtleties and nuances only understandable after years of exposure. As valuable as mistakes can be, they’re practically useless if left ignored or undetected. While instructors and native speakers can definitely be sources for identifying mistakes, oftentimes students might not have these resources at their disposal. That’s where artificial intelligence comes in.

AI text analysis isn’t anything new; tools like Grammarly and spell-check have existed since the late 20th century. However, most of these tools are either hidden behind a paywall or limited in what they’re able to do. Each tool dedicated to a certain task, whether it be checking for grammar or spelling, is restricted to a particular set of capabilities. Additionally, many of these tools only work on the English language, a fact which immediately invalidates a large bulk of existing spelling and grammar checkers when it comes to learning a new language. Recently, however, a more versatile form of AI has emerged in the form of ChatGPT, Google’s Bart, and other generative AI chatbots. These tools are unique in that they aren’t programmed to complete one specific task, but rather, they are given a large dataset consisting of information from all over the internet to learn from, so as to be able to “generate” responses to any question asked (hence their name). Not surprisingly, text analysis is a big feature of generative AI tools. And while specific tools targeted for grammar help might be confined to a small repertoire of capabilities, these models can be customized to help with whatever question, no matter how random or obscure.

Let’s take a look at what generative AI can actually do.

AI Text Analysis for Language-Learning

When it comes to learning about different parts of the world, one of the largest hurdles is gaining familiarity with different languages. Compared to math, history, and other subjects, mastering new languages can be particularly frustrating due to its reliance on the learner’s individual proactiveness. Most likely, even after carefully paying attention in an introductory course, there might still be a handful of beginner sentences that seem almost incomprehensible due to their use of unfamiliar vocabulary or obscure grammatical structures. It’s often up to the learner to go beyond the standard curriculum in order to truly pick up the language, a task which may seem daunting when done alone.

Fortunately, with the advent of generative AI tools like ChatGPT or Google’s Bard, there are many new tools available to help with this task — particularly AI summarization. Now more than ever, generative AI tools are utilizing what is called a “transformer” architecture. This state-of-the-art AI model enables tools like ChatGPT to understand relationships between words as well as to remember and utilize information from all over the piece of text. These traits, which previous iterations have struggled with, allow the AI to be highly adept at summarizing large bodies of text.

Posts by Michael Cai