Whether you’re a fan of artificial intelligence or tired of hearing about it, there’s no denying one thing. AI-powered tools can be fun and useful when you know how to use them well. Millions of people are finding this out, as Google’s AI chatbot, Gemini, is preloaded on some of the best Android phones.
Gemini is my AI chatbot of choice. It’s the first one I use to solve a problem or find the answer to a tricky question. Sometimes, I throw requests at it to find its capabilities and limitations. That’s how I came across the six surprising uses for Gemini I didn’t know about. Let me share them with you.
Related
6
Read handwriting and blurry text
From messy notes to fancy cursive
Gemini read the blurred text above with 90% accuracy. The original text is below.
Someone recently sent me a blurry image of an email. Although it was somewhat readable, it put a strain on my eyes. I thought I’d see if AI could help. I uploaded the image to Gemini, and it responded with a flawless copy of the email text.
This made me want to test Gemini’s text recognition abilities further. I copied a paragraph from an article and blurred it in Photoshop so that I couldn’t read it. However, Gemini somehow could. It wasn’t perfect this time. It got about 10% of the words wrong, which was enough for some sentences not to make sense. Still, that’s better than what a human would have been able to decipher.
Next, I threw images of handwritten notes at Gemini, expecting them to pose a greater challenge to Google’s AI. Moments later, I had an almost perfect transcription. Gemini can be a powerful companion for organizing and making sense of the notes you take during lectures. However, I wonder if it could be used for nefarious reasons if it’s better at reading blurry text than a human.
5
Flag potential mistakes in text and documents
It’s like having a personal editor
Google Gemini can process huge volumes of data faster than a human. It also provides editing suggestions and flags mistakes in text and documents. In other words, Gemini can be your editor, whether you’re writing a college paper or learning a new language. You don’t need Gemini Advanced for this to work, although a paid subscription lets you upload larger amounts of text.
Editing tools like Grammarly come with a monthly subscription, but the free version of Gemini can flag typos, grammatical errors, or vague or unclear sentences. Unlike built-in proofreaders that underline potential mistakes, Gemini explains the reason for its editing suggestions.
For privacy reasons, do not share sensitive data or upload documents containing sensitive information to Gemini.
4
Detect AI-generated images
Nope, Venice didn’t freeze this winter
Google Gemini can generate images and detect photoshopped and AI-generated images. Perhaps “detect” is too strong of a word, but Google’s AI can analyze images and tell you if they are genuine or doctored. Its analysis is comprehensive. Gemini looks at a long list of factors, from lighting, perspective, composition, and reflections to digital noise, facial expressions, and the likeliness of the object in the image to exist.
I tested this use for Gemini with a couple of dozen pictures, including AI-generated images, viral photoshopped posts, and real photos I took. The AI’s observations were 100% accurate in my case. I can’t guarantee it will work as reliably every time, but it could be a good starting point in a broader analysis.
3
Summarize cooking recipes
Because nobody has time to read the history of chicken adobo
Cooking recipes are some of the most unnecessarily bloated pages online. On one hand, I understand the authors’ need to monetize their content and optimize it for SEO. On the other hand, this “optimization” often buries the actual recipe under ads, newsletter popups, and four paragraphs dedicated to the dish’s history.
Google Gemini can summarize pages with recipes for me. I give it the link and ask for a list of the ingredients and steps in the recipe. In the same prompt, I can ask Gemini to convert units from cups and ounces to grams or milliliters. I can also ask the AI to generate a shopping list in my Google Keep notes.
2
Suggest recipes based on ingredients
Works for mixed drinks, too
Cooking can be lots of fun. Coming up with dinner ideas isn’t much fun. When I’m out of inspiration, I ask Gemini to suggest dishes I can cook with the ingredients I have. Uploading a picture of my fridge to Gemini also works when I don’t feel like listing every ingredient. This works for drinks, too. If I give Gemini a picture of the spirits and sodas I have in my home bar, it recognizes them and suggests a cocktail.
1
Identify LEGO pieces
Useful for LEGO lovers of all ages
My kid loves playing with LEGO, and so do I. He’s amassed a 6-gallon (22-liter) storage box full to the brim with blocks of all sorts. Sometimes, it’s hard to tell which set a piece belongs to. As it turns out, Gemini is pretty good at identifying LEGO pieces and their unique numbers. With rarer pieces, it tells me which set it came from. It’s too bad Google’s AI can’t tell me what new things I can build with the pieces we have.
What will AI be able to do tomorrow?
I can’t say for sure where AI is headed, but I’m confident it’s here to stay. While generative AI results are not always perfect or usable, they will improve as top tech companies invest billions in AI development. I still wouldn’t trust a chatbot to troubleshoot my plumbing or diagnose a skin rash. Still, for other uses, I’m certain Google Gemini could be a valuable tool for researching topics, summarizing documents, or generating a fun, custom AI wallpaper.