Generative AI in Language Classes: Practical Examples with Gemini

By Wenjing Huang, Chinese Teacher, George School

DOI: https://www.doi.org/10.69732/XXIL7439

Introduction

Since the launch of ChatGPT by OpenAI in 2022 and its subsequent variations, generative AI powered by large language models has achieved a new level of competence. These chatbots can perform research, support reasonably good writing, and generate realistic videos, audio, and images. This article focuses on using Gemini (free version) – formerly known as Bard – Google’s AI chatbot launched in 2023. This article presents practical examples of its application in language classes in three areas: webpage analysis, video analysis, and image analysis. All free version features and limitations mentioned in this article are current as of May 2025.

Webpage Analysis

Gemini excels at analyzing structured, text-heavy, or Google-indexed web content by leveraging its integration with Google Search, Google Scholar, Google Shopping, and other Google tools. It summarizes key points efficiently by extracting headlines, dates, and author intent. One major advantage of using existing web content is that it grounds Gemini’s responses in real, verifiable sources. This reduces the risk of generating such as bias towards English-language content, or hallucinations.

Below are the types of web pages it handles well: 

• News articles
• Academic paper and research articles
• Product pages and reviews
• How-to guides and tutorials (e.g., recipes, cooking tutorials)

Example 1: Analyzing a News Article with Gemini

For my first example, I used a news article from World Journal, the largest Chinese language newspaper in the U.S. The article discusses current AAPI events and cultural experiences within the Chinese community. Here is how I applied Gemini:

1. Input Prompt: Summarize this article’s content in English, Chinese, and Chinese Pinyin. 
2. Results: Gemini provided a reasonably accurate and coherent summary in all three requested formats.

Gemini’s Response:

Here is a summary of the linked content in English, Chinese, and Pinyin:

English Summary: The article details the journey of Chinese chef Zhang Pengliang, who previously cooked for former Chinese President Hu Jintao and now owns 18 restaurants in the Washington, D.C. area. Zhang and his family immigrated to the U.S. in 2003 for a better life. Despite initial challenges and working under assumed names, he established his culinary career, opening his first restaurant in 2011 and expanding with partner Gen Lee. His culinary philosophy emphasizes customer feedback and staying true to his strengths. His daughter, Lydia, serves as the CEO, handling marketing and public relations. Zhang was a James Beard Foundation’s American best chef semifinalist in 2015 and 2020.

中文总结 (Chinese Summary): 文章详细介绍了中国厨师张鹏亮的经历，他曾为中国前国家主席胡锦涛烹饪，如今在华盛顿特区拥有18家餐厅。张鹏亮和家人于2003年移民美国，寻求更好的生活。尽管最初面临挑战并曾使用假名工作，他仍建立了自己的烹饪事业，于2011年开设了第一家餐厅，并与合伙人李根一起扩展业务。他的烹饪理念强调客户反馈并坚持自己的优势。他的女儿Lydia担任首席执行官，负责市场营销和公共关系。张鹏亮在2015年和2020年曾是詹姆斯·比尔德基金会美国最佳厨师半决赛入围者。

拼音总结 (Pinyin Summary): Wénzhāng xiángxì jièshào le Zhōngguó chúshī Zhāng Péngliàng de jīnglì, tā céng wéi Zhōngguó qián guójiā zhǔxí Hú Jǐntāo pēngrèn, rújiāng zài Huáshèngdùn Tèqū yǒngyǒu 18 jiā cāntīng. Zhāng Péngliàng hé jiārén yú 2003 nián yímín Měiguó, xúnqiú gèng hǎo de shēnghuó. Jǐnguǎn zuìchū miànlín tiǎozhàn bìng céng shǐyòng jiǎmíng gōngzuò, tā réng jiànlì le zìjǐ de pēngrèn shìyè, yú 2011 nián kāishè le dì yī jiā cāntīng, bìng yǔ héhuǒrén Lǐ Gēn yīqǐ kuòzhǎn yèwù. Tā de pēngrèn lǐniàn qiángdiào kèhù fǎnkuì bìng jiānchí zìjǐ de yōushì. Tā de nǚ’ér Lydia dānrèn shǒuxí zhíxíngguān, fùzé shìchǎng yíngxiāo hé gōnggòng guānxì. Zhāng Péngliàng zài 2015 nián hé 2020 nián céng shì Zhānmǔsī Bǐ’ěrdé Jījīnhuì Měiguó zuì jiā chúshī bàn juésài rùwéi zhě.

1. Input Follow-Up Prompt: Modify the Chinese summary to align with ACTFL’s Intermediate Low proficiency level, or adjust it to match a second-grade native Chinese speaker’s reading level. 
2. Results: The revised summary was successfully adapted to the target proficiency level, using simpler sentence structures and high-frequency vocabulary. While a few new words (see underlined ones in the below response) were introduced, they remained contextually manageable, making the text accessible yet slightly challenging, and ideal for learners at this level.

Gemini’s Response:

张鹏亮是一位很厉害的厨师。他以前在中国，给胡爷爷做过好吃的饭菜。后来，张鹏亮和家人一起去了美国。现在，他在美国开了好多饭店，一共有18家呢！

他们是在2003年去的美国。刚开始，他们遇到了一些困难，但是张鹏亮非常努力地工作。到了2011年，他开办了自己的第一家饭店。渐渐地，他和朋友李根一起，让饭店变得越来越多。

张鹏亮很喜欢听顾客说饭菜的味道怎么样，他总是努力把饭菜做得更好。他的女儿Lydia也很棒，她在饭店里帮助爸爸管理工作。张鹏亮的厨艺非常好，还因为做饭好，在2015年和2020年得了很重要的奖。

(Translation: Zhang Pengliang is a very skilled chef. He used to be in China, where he cooked delicious meals for Grandpa Hu. Later, Zhang Pengliang and his family moved to the United States. Now, he has opened many restaurants there—18 in total! 

They moved to the U.S. in 2003. At first, they faced some difficulties, but Zhang Pengliang worked very hard. By 2011, he opened his first restaurant. Gradually, together with his friend Li Gen, he expanded the business to more and more locations. 

Zhang Pengliang loves hearing feedback from customers about how the food tastes, and he always strives to make the dishes even better. His daughter, Lydia, is also amazing—she helps her father manage the restaurant operations. Zhang Pengliang ‘s culinary skills are outstanding, and his excellence in cooking earned him prestigious awards in 2015 and 2020.)

Student Application: After sharing the adapted summary, students engage in scaffolded interpretive reading activities (e.g., gist identification, true/false questions, and sequencing tasks) to reinforce comprehension and key vocabulary. Next, interpersonal discussions such as opinion polls, role-play interviews, and cultural comparisons encourage students to use the text as a springboard for conversation and critical thinking. 

Educators can also ask Gemini to adapt the same text for different proficiency levels (e.g., ACTFL’s Novice High or Advanced Low) by adjusting their prompts. Parallel interpretive and interpersonal tasks can then reinforce each leveled text, scaffolding learning to meet diverse learner needs. Another way is to reintroduce the original, more complex version of the text after learners have engaged with the simplified one. This allows them to compare the two versions and identify parallel sections.

Example 2: Creating Multimedia Slides with Gemini

For this example, I used Gemini to generate a slide deck based on a webpage about Hong Kong snacks. Below is the process and outcome:

1. Input Prompt in Chinese: Start a new chat session to ensure Gemini does not reference prior conversations. Use the contents of the following webpage to select 6 snacks and create 6 slides. Each slide should feature one snack, introducing a brief description and a related picture.
2. Results: Gemini successfully generated 6 structured slides, each featuring an accurately labeled snack title and a concise description that matched the original webpage’s content upon verification. Although the free version does not automatically generate images (requiring additional prompts later), it provides a practical “Share and Export” function to seamlessly transfer the text to Google Docs. This output demonstrates Gemini’s utility in rapidly curating foundational slide content and reducing prep time.

Gemini’s Response: See this Google Docs that it exported.

If the initial results don’t meet expectations, refine your approach by revising the prompt—for instance, translating a target-language prompt into English for better clarity. Persistence and iterative adjustments often lead to improved outcomes. Additionally, consistently use Gemini’s feedback tools (thumbs up/down) to train the model, helping it better align with your specific instructional goals over time. Small refinements and active feedback can create a more tailored and effective AI collaboration.

The types of webpages Gemini may struggle with are:

• Literature Texts (especially complex, subjective, or highly symbolic ones): due to figurative language, symbolism, multiple interpretations, subjectivity or emotions, etc.
• Social Media Posts (especially short, nuanced, or highly contextual ones): due to lack of context, inside jokes, sarcasm, irony, slang, bias, rapidly evolving linguistic trends, private accounts and login requirements, etc.
• Image-Heavy or Visually Driven Webpages: due to the main ideas being conveyed through images, infographics, or videos with minimal text.

Video Analysis

Gemini can effectively analyze and summarize YouTube videos with subtitles (a Google-owned platform), extracting fact-oriented content and key information. However, its capabilities come with some limitations: 

• Dependence on Subtitles: It cannot summarize videos lacking subtitles or accurate transcripts. 
• Nuanced Content Challenges: Subtler creative elements (e.g., humor, metaphors, tone) may be overlooked or misrepresented.
• Style Consistency: Generated summaries might not fully align with the original video’s stylistic or rhetorical approach.
• Complex Information Processing: Multi-layered arguments or abstract viewpoints may not be captured with precision.

The process for analyzing a YouTube video that has subtitles with Gemini mirrors that of analyzing a webpage: simply provide the video’s link, then ask targeted questions like “What are the key points in this video?” or “Which terms/theories/events are mentioned?” However, since Gemini’s responses rely solely on the provided text, always fact-check its answers against the original video—especially for nuanced claims or creative content that may be lost in transcription. This verification step ensures accuracy while leveraging Gemini’s efficiency in distilling structured information.

Beyond saving educators and learners time while boosting efficiency, Gemini’s video analysis feature offers a powerful “third eye” for improving teaching practices through Reflective Teaching. Reflective Teaching involves educators systematically reviewing teaching videos to evaluate instructional strategies, student engagement, and learning outcomes. While traditionally time-intensive, this reflective process can now generate data-driven insights in seconds through Gemini’s analysis, enabling teachers to rapidly refine their classroom approach.

Example 3: Analyzing a Teaching Video with Gemini

Input Prompt: (Start a new chat session to ensure Gemini does not reference prior conversations.) 

• What content is the teacher teaching in this YouTube video? 
• Evaluate this teacher’s teaching methods.

Results: Gemini delivered a comprehensive analysis of the 11-minute sample French lesson (sourced from YouTube) based on the video transcript: 

• Content Identification: Clearly stated the lesson’s subject and learning objectives.
• Method Evaluation:

1. Strengths: Listed 7 effective practices (e.g.fostering a positive classroom environment, flexible use of teaching tools, timely feedback) with brief explanations.
2. Areas for Growth: Provided 2 transcript-based suggestions (e.g., individual student attention, time management for games)

• Summary: Concluded with a concise overview of the teacher’s instructional approach.

Key Observations:

• Timestamp Variability: Gemini may or may not include specific timestamps when citing strengths/weaknesses, depending on transcript formatting. 
• Language Limitations: While Gemini can analyze non-English subtitles (e.g., Spanish, French, Chinese), its accuracy may vary for less common languages due to limited training data or poor auto-generated subtitles. Therefore, always verify critical points against the original video.

Here is Gemini’s Response: 

This video captures Mary Beth Fischer’s French class, focusing on student games and competitions. From the clips, it’s evident that the teacher is using interactive games to review and teach French vocabulary, particularly words related to hotels (l’hôtel) and ordinal numbers (les nombres ordinaux).

Assessment of the Teacher’s Teaching Methods:

This teacher’s methods have several commendable aspects:

• Highly Interactive and Fun Learning: The teacher designed multiple game segments, such as having students take turns saying ordinal numbers, expressing vocabulary through actions and sounds, and using mini whiteboards for quick-fire answers. These games effectively boosted student engagement, allowing them to learn and review French vocabulary in a relaxed and enjoyable atmosphere.
• Emphasis on Oral Expression and Participation: Throughout the class, the teacher encouraged students to actively speak French, even offering encouragement for inaccurate pronunciation. This approach helps improve students’ speaking skills and builds their confidence in using French.
• Timely Feedback and Reinforcement: The teacher provided immediate positive feedback and praise after students answered questions, saying “Bravo!” and “Excellent!”, which helps reinforce student learning. Students who answered correctly also received rewards.
• Utilizes Visual and Auditory Stimulation: By asking questions verbally, having students express actions, and writing vocabulary on mini whiteboards, the teacher simultaneously stimulated students’ auditory and visual senses, helping them better understand and memorize French vocabulary.
• Fosters a Positive Classroom Environment: The overall classroom atmosphere was lively with high student participation, indicating that the teacher created an environment that encouraged learning and active interaction. With Christmas approaching, students even wore Santa hats, adding to the relaxed and festive mood.
• Clear Learning Objectives: The teacher clearly listed the lesson’s learning objectives on the whiteboard, including review, vocabulary acquisition, and listening practice. This helps students understand the key learning points.
• Flexible Use of Teaching Tools: The teacher utilized various teaching tools, such as whiteboards and mini whiteboards, making the teaching format more diverse.

Aspects to Consider:

• Individual Student Attention: In some group activities, it might be challenging for the teacher to ensure every student’s participation and comprehension.
• Time Management for Games: While games are fun, it’s important to manage the time effectively to ensure that the teaching content can be completed as planned.

Overall, this French teacher effectively increased students’ learning interest and engagement through lively and engaging games and active interaction. This is a commendable teaching approach, especially suitable for vocabulary acquisition and speaking practice, helping students grasp French knowledge in a relaxed environment.

Image Analysis

Gemini is often used to either find existing pictures or create original images. When finding images, Gemini searches the web to source authentic visuals like historical photos or cultural references, though educators must verify copyright compliance. For creating images, Gemini generates custom visuals (e.g., vocabulary cartoons or dialogue scenes) tailored to lesson needs. 

Typically Gemini responds faster when finding an image (searching existing ones) compared to creating a new one.

• Finding Images: Simply retrieves pre-existing images from the Internet. Response time is near-instant (a few seconds), assuming the image exists online.
• Creating Images: Involves generating visuals from scratch using AI, which takes longer due to computational processing. Processing time varies, depending on image complexity and server load. The free version of Gemini also imposes daily limits on image generation.

These functions pair well with interpretive tasks and visual literacy development, helping learners analyze authentic materials with AI support.

Example 4: Analyzing a Newspaper Title Image with Gemini

To enhance learners’ interpretive skills, in my example 4, I used Gemini to analyze two authentic images: both newspaper titles written in traditional Chinese characters. The first image features a one-line title that includes a special name in English, appropriate for intermediate-level learners. The second image contains a two-line title written entirely in Chinese, with a question mark that hints at the author’s perspective, and the language content is more suitable for advanced level learners. Here is the process using Gemini.

1. Input Prompt: (Start a new chat session to ensure Gemini does not reference prior conversations.) List the words or phrases from the attached image (see Picture 2 or 3), and provide their meanings in both English and in Chinese.
2. Results: Gemini provides a breakdown of the words and phrases found in each image, with meanings presented in both English and Chinese. 

• In the first example, Gemini correctly identified the English part as a brand name or company name. 
• For both headlines, Gemini accurately explains the overall meaning in both languages, including inferred viewpoints suggested by the wording.

Gemini’s Response:

Here’s a breakdown of the words and phrases from the image, with their meanings in English and Chinese:

1. iTaiwan Foods
   • English: iTaiwan Foods (This appears to be a brand name or company name, possibly related to Taiwanese food products.)
   • Chinese: iTaiwan Foods (艾台湾食品 – 可能是某家食品公司的名称，或指与台湾食品相关的某平台/组织。)
2. 迎 (yíng)
   • English: To welcome; To greet
   • Chinese: 迎 (yíng) – 迎接，欢迎。
3. 端午 (duān wǔ)
   • English: Dragon Boat Festival (a traditional Chinese holiday)
   • Chinese: 端午 (duān wǔ) – 中国传统节日，农历五月初五。
4. 试吃 (shì chī)
   • English: To sample food; Food tasting; Try eating
   • Chinese: 试吃 (shì chī) – 尝试品尝食物。
5. 台湾 (Tái wān)
   • English: Taiwan
   • Chinese: 台湾 (Tái wān) – 中国东南沿海的岛屿，一个地区。
6. 粽 (zòng)
   • English: Zongzi (a traditional Chinese sticky rice dumpling, typically eaten during the Dragon Boat Festival)
   • Chinese: 粽 (zòng) – 即粽子，一种用糯米和馅料包在竹叶或荷叶里煮熟的食物，端午节常吃。

Overall meaning of the headline “iTaiwan Foods 迎端午 试吃台湾粽”:

• English: iTaiwan Foods welcomes the Dragon Boat Festival (with an event to) sample Taiwanese Zongzi.
• Chinese: iTaiwan Foods 迎接端午节，提供台湾粽试吃。 (暗示该公司在端午节期间推广台湾粽子，并提供品尝活动。)

Gemini’s Response:

Here’s a breakdown of the words and phrases from the image, with their meanings in English and Chinese:

1. 高中 (gāo zhōng)
   • English: Senior High School; High School
   • Chinese: 高中 (gāo zhōng) – 指初中之后的三年教育阶段。
2. 双休 (shuāng xiū)
   • English: Two days off (referring to weekend rest, typically Saturday and Sunday); Weekend break
   • Chinese: 双休 (shuāng xiū) – 每周有两天休息时间，通常指周六和周日。
3. 假的？ (jiǎ de?)
   • English: Fake? False? Is it true? (Implying doubt or disbelief)
   • Chinese: 假的？ (jiǎ de?) – 表示疑问，是不是真的，是不是虚假的。
4. 教改 (jiào gǎi)
   • English: Educational reform
   • Chinese: 教改 (jiào gǎi) – 教育改革的简称。
5. 未 (wèi)
   • English: Not yet; Haven’t
   • Chinese: 未 (wèi) – 还没有，没有。
6. 救赎 (jiù shú)
   • English: Redemption; To redeem; To save
   • Chinese: 救赎 (jiù shú) – 解救，挽回，使摆脱困境。
7. 反 (fǎn)
   • English: On the contrary; Instead; Reversely
   • Chinese: 反 (fǎn) – 相反地，反而。
8. 加剧 (jiā jù)
   • English: To intensify; To aggravate; To worsen
   • Chinese: 加剧 (jiā jù) – 使程度更深，使更严重。
9. 焦虑 (jiāo lǜ)
   • English: Anxiety; Anxious
   • Chinese: 焦虑 (jiāo lǜ) – 忧虑不安，担心。

Overall meaning of the phrases:

• 高中双休 假的？
  • English: Senior High School (having) two days off? Is it fake? (Implying skepticism about a policy allowing high school students two days off.)
  • Chinese: 高中生能有周末双休吗？这是假的吧？ (表达对高中生双休政策真实性的怀疑。)
• 教改未救赎 反加剧焦虑
  • English: Educational reform hasn’t brought redemption (or hasn’t saved the situation), instead it has intensified anxiety.
  • Chinese: 教育改革非但没有解决问题，反而加剧了人们的焦虑。 (批评教育改革没有达到预期效果，反而带来了更多负面影响。)

This activity helps develop learners’ autonomous learning ability by providing instant, scaffolded support for interpreting authentic materials. When learners use Gemini to analyze images like newspaper titles, they learn to independently break down unfamiliar vocabulary, interpret meaning in context, and explore cultural cues. 

Another important thing to note is that when taking a screenshot of a newspaper for Gemini to analyze, avoid including horizontal titles along with multiple columns of content in a single image. For accurate decoding, Gemini requires a clear, column-by-column screenshot. If the news content spans three or more columns in one image, Gemini may not process the text correctly.

Example 5: Identifying a Store Name with Gemini

When traveling to target-language countries or cities, learners often encounter store or shop names that include unfamiliar words, cultural references, or idioms. These names can offer rich insights into local values, humor, or marketing styles, but they may be challenging to interpret at first glance. In such cases, Gemini can serve as a valuable tool to help learners decode the meaning and cultural significance of these names. My following example explores how Gemini analyzes a store name that contains a Chinese idiom.

Here is the process using Gemini.

1. Input Prompt: (Start a new chat session to ensure Gemini does not reference prior conversations.) Identify the store name in Chinese and provide details in both Chinese and English.
2. Results: Gemini provides precise English and Chinese meanings for words and phrases, along with cultural context and broader meaning behind the idiom.

Gemini’s Response:

好的，我们来分析图片中的店名及其中文意思，并提供详细信息。

图片中的店名是： 衣见钟情 (Yī Jiàn Zhōng Qíng) 下方小字：精品服饰 (Jīng Pǐn Fú Shì)

中文意义分析：

1. 衣见钟情 (Yī Jiàn Zhōng Qíng)：
   • 这个店名是谐音梗，巧妙地改动了中文成语“一见钟情 (yī jiàn zhōng qíng)”。
   • 一见钟情 (yī jiàn zhōng qíng)：指男女之间一见面就产生爱情，相互爱慕。其含义是“初次见面就爱上”。
   • 衣 (yī)：在这里代替了“一”，字面意思是“衣服”或“服装”。
   • 因此，“衣见钟情” 的谐音含义是：顾客一看到这里的衣服，就立刻喜欢上，甚至爱上它们。它传达了一种强烈的情感吸引，暗示店里的服饰具有很高的颜值、款式或品质，能够瞬间俘获顾客的心。这是一个非常聪明且吸引人的店名，既有文化底蕴（借用成语），又直接点明了行业特色（与衣服相关），并表达了对产品魅力的自信。
2. 精品服饰 (Jīng Pǐn Fú Shì)：
   • 精品 (jīng pǐn)：指精心制作的、质量上乘的、有独特价值的物品，即“精品”、“优质产品”。
   • 服饰 (fú shì)：指服装和装饰品，是服装行业的统称。
   • 意义：高品质的服装和服饰。这进一步强调了店铺的定位，即销售的不是普通大众服饰，而是注重设计、材质和品质的精选产品。

The image displays a signboard with a store name and a subtitle. Let’s analyze its meaning in both Chinese and English.

English Meaning Analysis:

1. 衣见钟情 (Yī Jiàn Zhōng Qíng):
   • This store name is a clever pun on the Chinese idiom “一见钟情 (yī jiàn zhōng qíng),” which means “love at first sight” or “to fall in love at first glance.”
   • The character 衣 (yī), meaning “clothing” or “garment,” replaces the homophonous character “一 (yī),” meaning “one” or “first.”
   • Therefore, the pun’s meaning for “衣见钟情” is: Customers will fall in love with the clothing at first sight. It conveys a strong emotional appeal, suggesting that the garments in the store are highly attractive in appearance, style, or quality, capable of instantly captivating customers’ hearts. This is a very creative and appealing name, combining cultural depth (by referencing an idiom) with a direct indication of the industry (related to clothing), and expressing confidence in the products’ charm.
2. 精品服饰 (Jīng Pǐn Fú Shì):
   • 精品 (jīng pǐn): Refers to high-quality, meticulously crafted, or valuable goods; premium products; boutique items.
   • 服饰 (fú shì): Refers to clothing and accessories; apparel.
   • Meaning: High-quality apparel and accessories; boutique clothing. This further clarifies the store’s positioning, indicating that it sells not just ordinary mass-market clothing, but carefully selected products emphasizing design, material, and quality.

This capability offers language learners a dynamic way to understand authentic materials, bridging the gap between raw visual input and comprehensive linguistic comprehension. By uncovering the literal and cultural meanings embedded in everyday visuals like store names, Gemini empowers learners to engage more deeply with the target language environment.

Example 6: Analyzing A Restaurant Menu with Gemini

Gemini can also assist language learners in reading and interpreting restaurant menus when traveling to a target-language region. This is particularly helpful when menus contain unfamiliar dish names, cultural references, or handwritten text. Similar to the previous examples, learners can upload an image of the menu and prompt Gemini to analyze the content. In this case, 

1. Input Prompt: (Start a new chat session to ensure Gemini does not reference prior conversations.) Identify the dish names in the attached image in Chinese, provide Chinese Pinyin, translate them into English, and extract the prices.
2. Results: Gemini responds by organizing this information into a clear table format, making it easier for learners to compare and understand the items. Additionally, Gemini offers an option (at the end of its response) to export the table directly into Google Sheets, allowing learners or teachers to further annotate, sort, or integrate the data into classroom activities or travel preparations.

If a menu contains multiple columns or densely packed sections, Gemini may not decode all the content accurately in one attempt. For best results, it’s recommended to take separate screenshots of each column or section and upload them individually for analysis.

Example 7: Integrating Google Maps with Gemini

As part of the Google ecosystem, Google Maps integrates seamlessly with Gemini, allowing for location-based tasks connected to target-language contexts. By using specific prompts, educators can guide Gemini to generate maps or identify culturally relevant locations such as historical sites, local markets, or famous restaurants in the target-language region. This feature supports cultural exploration, travel preparation, or thematic units related to geography, daily life, or tourism. Learners can interact with real-world maps while using the target language to navigate directions, understand place names, or plan itineraries. This integration enhances both linguistic and cultural competence through authentic, map-based tasks.

This example can be used as a simulated field trip to explore cultural landmarks in a target-language community. In this case, multiple prompts were input to generate a customized Google Map of Chinatown in Manhattan, New York.

• First Prompt: Display Manhattan, New York’s Chinatown on the map.

Results: Gemini generated a Google Map image with zoom functionality, highlighting the correct location, along with a clickable link to the live map.

• Second Prompt in Chinese: On this Manhattan Chinatown map, mark several typical landmarks such as famous restaurants, temples, and shops with linguistic and cultural significance. Then provide the cultural backgrounds of these places.

Results: Gemini returned a detailed list of notable landmarks, including restaurants, temples, and shops, along with explanations of their cultural relevance. However, since this output was provided in text format without an updated map, it could be given a thumbs-down feedback.

• Third Prompt: Mark the above-mentioned landmarks and locations on the Manhattan Chinatown map.

Results: Gemini then generated an updated zoomable Google Map (see Picture 10) with the selected landmarks clearly marked.  Some typical locations were also accompanied by  clickable Google Map links.

This example lends itself well to collaborative group work, allowing students to engage in meaningful research and cultural exploration. Each group can be assigned one or two landmarks within a specific Chinatown to investigate using Gemini and other digital tools. Students gather information about the location’s history, cultural significance, and linguistic features, then prepare a presentation to share their findings with the class. After all groups have presented, the class can vote on the most engaging and informative presentation. As a culminating activity, the winning group’s selected landmark can become the destination for a real-world field trip, turning digital exploration into an authentic, immersive cultural experience.

Conclusion

Gemini’s free version offers educators powerful tools to enhance language instruction through webpage analysis (summarizing articles, extracting vocabulary), video analysis (processing subtitled content for comprehension tasks), and image analysis (decoding visuals or generating custom materials), connecting directly to YouTube and other Google products. Unlike ChatGPT’s free version (limited access to file uploads, advanced data analysis, and image generation), Gemini integrates these features seamlessly, enabling dynamic lesson planning with authentic resources. While limitations exist, it still stands out as a choice for educators aiming to merge AI efficiency with pedagogical creativity.

References

Foote, K. D. (2024, March 5). A brief history of generative AI. DATAVERSITY. https://www.dataversity.net/a-brief-history-of-generative-ai/

Team, S. (2019, June 13). Exploring the limitations of AI summarization in research. Enago Read – Literature Review and Analysis Tool for Researchers. https://www.read.enago.com/blog/exploring-the-limitations-of-ai-summarization-in-research/

AI summarization. (n.d.). Google Cloud. https://cloud.google.com/use-cases/ai-summarization

Chen, L. (2024, June 4). 從ChatGPT到Gemini: 生成式AI於中文教學的展望 [From ChatGPT to Gemini: Prospects for generative AI in Chinese language teaching] [Video]. Google Drive. 

https://drive.google.com/file/d/1cCkSFSK_pwrOO5gxwLz9cm6oh1S0zZUO/view?t=1