{"id":3042,"date":"2024-02-06T16:00:08","date_gmt":"2024-02-06T16:00:08","guid":{"rendered":"https:\/\/www.languagetesting.com\/blog\/?p=3042"},"modified":"2024-03-18T16:52:29","modified_gmt":"2024-03-18T16:52:29","slug":"language-at-work-leveraging-the-power-of-ai-in-language-education-part-2","status":"publish","type":"post","link":"https:\/\/www.languagetesting.com\/blog\/language-at-work-leveraging-the-power-of-ai-in-language-education-part-2\/","title":{"rendered":"Language at Work: Leveraging the Power of AI in Language Education, Part 2"},"content":{"rendered":"

By Dan Edwards, Ed.D. and Kevin Gaugler, Ph.D.<\/strong><\/em><\/p>\n

From the learner\u2019s perspective, AI might turn out to be an ideal complement to Steven Krashen\u2019s Input Hypothesis<\/a>. According to Krashen, language learners advance by understanding input that is slightly beyond their current level of comprehension, often denoted as \u201d”i+1,\u201d where ‘i’ represents the learner’s current level of language proficiency.<\/p>\n

<\/p>\n

The Vesuvian Challenge and AI in Language Education<\/strong><\/h3>\n

A vivid illustration of AI’s potential in supporting comprehensible input can be drawn from the Vesuvian Challenge<\/a>, where technology played a pivotal role in deciphering ancient scrolls. In this challenge, machine learning and computer vision were harnessed to read the charred Herculaneum Papyri, which were rendered illegible due to deterioration from the eruption of Mount Vesuvius. A text that was incomprehensible to any human was made legible and comprehensible through the use of artificial intelligence recognizing patterns, making it understandable to the human reader. This scenario parallels the role AI might play in language education, decoding texts to meet a reader\u2019s proficiency, rendering previously incomprehensible material accessible, and thus aiding in differentiating content to each learner\u2019s specific needs. Think of this as a progressive scaffold, rather than a permanent solution. The intention is that readers will develop the skills to read higher level texts over time, eventually eliminating the need for AI text leveling.<\/p>\n

AI Tools for the Language Classroom<\/strong><\/h3>\n

Platforms like ChatGPT have shown potential in bridging the gap between a complex text and learner\u2019s comprehension. They can be programmed to align with a proficiency framework. For example, in our experimentation with ChatGPT, we asked the chatbot to produce text aligned with an A2 level on the Common European Framework of Reference (CEFR, a scale that correlates with the ACTFL proficiency scale). This approximates output that a Novice High learner can understand. Moreover, a learner might ask the chatbot to use specific vocabulary in context or to produce a text written mostly in the past, present, or future tense. In this way, classroom materials can be specifically customized to reach the optimal level of input+1 for any given learner in seconds. (See here\u00a0<\/a>for alignment of the CEFR scale to the ACTFL Proficiency Guidelines).<\/p>\n

Just like instructor-oriented tools like Magicschool.ai<\/a>, other platforms are using OpenAI\u2019s underlying technology to create services oriented toward language learners. The Reverso Dictionary<\/a>, for example, employs AI to furnish a grammar checker and sentence rephraser that is currently available in English, Spanish, French, and Italian for a monthly fee. Productivity suites like Microsoft 365 are now also incorporating AI. For example, Microsoft\u2019s Edge browser<\/a> offers a Copilot feature powered by AI that provides a chat function as well as a writing assistant, and is now building this Copilot technology into their entire suite of tools, which includes Word. The tool recently added a \u201cGenerate page summary\u201d button, allowing a language learner to access authentic readings in the target language. An article in a technical journal, for instance, could be adapted to be appropriate for use in introductory language courses. It is likely that Microsoft Word may soon to be able to coach learners on the best way to express themselves in the target language of their choice. We do not know yet how this may impact the development of writing proficiency. Assessment like ACTFL\u2019s Writing Proficiency Test (WPT)<\/a> may become even more relevant in programs to assess the actual proficiency of a given learner without the support of any AI-enhanced editing or revision tools.<\/p>\n

Not only is AI reimagining the role of the dictionary, the browser, and the word processor, it is leading to a rethinking of the traditional flashcard. Quizlet has introduced a feature known as Q-chat<\/a>. Q-Chat, powered by OpenAI\u2019s ChatGPT<\/a>, is designed to help the learner practice. The synergy between Q-Chat and Quizlet’s digital flashcards demonstrates a significant shift in how even the most traditional of study techniques is becoming a potentially more dynamic learning experience.<\/p>\n

These examples showcase just a small subsection of what AI can allow teachers and learners to do. The innovations we have shared in this and our prior blog post undoubtedly herald a new phase in second language teaching and learning. AI will certainly impact assessment as well. ACTFL even recently announced the development of machine scoring<\/a> as an enhancement to the human-rating process for the AAPPL. Rather than sidelining actual human-to-human interaction, it is likely that the ACTFL gold-standard Oral Proficiency Interview<\/a> (OPI) will become even more significant as a means to measure true oral proficiency involving negotiation of meaning.<\/p>\n

Conclusion<\/strong><\/h3>\n

Beyond the time-saving advantages for educators, these tools can enhance the quality of education for learners through customization of content and by making authentic texts more accessible. Although AI cannot change the amount of time it takes to learn a language, it can make the process more engaging and efficient for educators and learners alike. As artificial intelligence becomes more ubiquitous, world language educators and learners will likely develop vastly different expectations for educational outcomes and assessments in the coming years. This is only the beginning!<\/p>\n

Watch a PBS special on AI in education featuring the authors of this article: https:\/\/www.pbs.org\/video\/how-ai-is-transforming-education-g06h9h\/<\/a>.<\/strong><\/p>\n

For More Information<\/strong><\/p>\n

Dive deeper into the topics and tools discussed in this post to better understand how to leverage AI in language education.<\/p>\n

    \n
  1. Krashen, S. (1987). Principles and Practice in Second Language Acquisition. Retrieved from http:\/\/www.sdkrashen.com\/content\/books\/principles_and_practice.pdf<\/a><\/li>\n
  2. The National Endowment for the Humanities. Reading the Invisible Library: the Herculaneum Papyrus Scrolls. Retrieved from https:\/\/www.neh.gov\/project\/reading-invisible-library-herculaneum-papyrus-scrolls\u00a0<\/a><\/li>\n
  3. Reverso Dictionary. Retrieved from https:\/\/www.reverso.net<\/a><\/li>\n
  4. Microsoft 365 Team.\u00a0 Announcing Microsoft 365 Copilot general availability and Microsoft 365 Chat. Retrieved from https:\/\/www.microsoft.com\/en-us\/microsoft-365\/blog\/2023\/09\/21\/announcing-microsoft-365-copilot-general-availability-and-microsoft-365-chat\/<\/a><\/li>\n
  5. Quizlet. Q-Chat: Meet Your New AI Tutor. Retrieved from https:\/\/quizlet.com\/qchat-personal-ai-tutor<\/a><\/li>\n<\/ol>\n","protected":false},"excerpt":{"rendered":"

    By Dan Edwards, Ed.D. and Kevin Gaugler, Ph.D. From the learner\u2019s perspective, AI might turn out to be an ideal complement to Steven Krashen\u2019s Input Hypothesis. According to Krashen, language learners advance by understanding input that is slightly beyond their current level of comprehension, often denoted as \u201d”i+1,\u201d where ‘i’ represents the learner’s current level […]<\/p>\n","protected":false},"author":36,"featured_media":3051,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[183],"tags":[375,377,376,278],"class_list":["post-3042","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-academic","tag-ai","tag-ai-in-education","tag-artificial-intelligence","tag-higher-education"],"acf":[],"aioseo_notices":[],"jetpack_featured_media_url":"https:\/\/www.languagetesting.com\/blog\/wp-content\/uploads\/2024\/01\/AI_blog_2-scaled.jpg","_links":{"self":[{"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/posts\/3042","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/users\/36"}],"replies":[{"embeddable":true,"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/comments?post=3042"}],"version-history":[{"count":6,"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/posts\/3042\/revisions"}],"predecessor-version":[{"id":3344,"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/posts\/3042\/revisions\/3344"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/media\/3051"}],"wp:attachment":[{"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/media?parent=3042"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/categories?post=3042"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.languagetesting.com\/blog\/wp-json\/wp\/v2\/tags?post=3042"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}