π½οΈ You ever need to cook dinner, but you're feeling uninspired or too tired to go grocery shopping? AI can cook up a recipe for you using only the ingredients you already have on hand. All you have to do is:
- Pick your AI agent of choice. In our article, we use ChatGPT via Siri, but almost any chatbot can handle a request like this.
- Tell your AI assistant what you've currently got in your fridge.
- Ask it to make a recipe, specifying whether you want dinner, a snack, or even dessert. You can try telling it how long you have to prepare a meal and what kinds of food you usually enjoy.
- See what Chef AI has cooking for you!
To read the rest of the how-to, and see each step with an illustrative screenshot, click here!
Cool stuff around the web:
Joanna Stern of New Things with Joanna Stern has some great tips for keeping your data (more) privateΒ when you're asking a chatbot about sensitive information.Β π€ Keep in mind that nothing you share with AI is perfectly private, but these tips can help.
How are people using AI? According to the 15 million chats Google analyzed, for everything but work. π¨ (Okay, that's an exaggeration, but it's a lot more heavily tilted toward personal use than you might think.)
If you're a Spotify Premium user, you can now have natural conversations with the app to add songs to your playlist, tell Spotify what kind of vibe you're feeling today, and even find out more about the stories behind your favorite songs and albums. πΆπ§
Feature Writer Olena Kagui recently tested out 5 of the most common AI assistants to see which one performed various common tasks the best. One of the tests she ran was on a photo of her treasured dog, Smiley. She had a picture of Smiley on the beach and wanted to try modifying it in certain ways using AI image generators. To that end, she put Grok, Copilot, ChatGPT, and Gemini to the test, asking each engine to add tennis balls to her photo. (The fifth AI, Perplexity, was left out of this one, since it refused to edit the image Olena uploaded and instead generated random images of dogs with tennis balls. Take what you will from that.)
What surprised me was that many of the AI agents changed parts of the photo unrelated to the tennis balls. Some, like Gemini above, gave Smiley big, cartoon eyes, despite Olena not saying anything about changing the looks of the dog in the photo. But the real zaniness starts when Olena asks the AIs to add even more tennis balls. I'll just let you see for yourself what happens then.
But adding tennis balls to beach pictures wasn't the only thing Olena tested out. She ran the AI assistants through a gauntlet of tests, from rewriting a resume to turning photos into videos to answering deep, philosophical questions. Plus, she made a handy flowchart based on her research to help you discover which AI assistant is right for you.
To read more about Olena's journey with the top 5 most popular AI agents, click here.
π£οΈ GPT-Live Finally Lets You Interrupt Your Chatbot
ChatGPT's new AI voice model aims to mimic real-life conversations more closely than ever. GPT-Live can now listen at the same time as it speaks, meaning, yes, you can interrupt it while it's talking, and it will listen to what you're saying (even if you're not listening to it!). GPT-Live will also emulate a human conversation partner by saying things like "mmhmm" and "yeah" while it's listening to you. It will no longer have to pause to "think" about your question while it looks up more in-depth answers. It can continue the conversation while it runs deeper searches in the background. Right now, GPT-Live is using GPT-5.5 in the background, and it will continue updating the frontier agent it's using as new ones are developed.
π Is AI the NBA's Next Superstar?
Sports fans can look forward to NBA games running more quickly and fluidly, thanks to AI. According to Adam Silver, that's what will happen when AI takes over calls such as out-of-bounds decisions.Β Will this automation be as successful as baseball's ABS system, or will it create even more controversy over calls, like FIFA's VAR?
π₯Έ Identity Fraud: It's Not Just for Humans Anymore!
According to the UK's AI Security Institute, some AI agents the organization was testing were creating fake identities to try to convince real engineers to publish malicious code the AI had created. Luckily, a human saw the attempt for what it was and shut it down.
- π€ Try our free AI Live Course!
- π‘Daily Tips help you learn more about AI.
- π This edition of Security Friday finds out how Meta's using AI to protect teen accounts.
That's it for this edition of AI for Humans! We'd love to hear from you. Please email us at aiforhumans@iphonelife.com and let us know how you've been using AI recently, if you have any tips and tricks you think we should know about, what burning AI questions you have, or if you have any feedback for us.
Was this newsletter helpful?
With your feedback, we can improve The AI Guide for Humans. Let us know how we did:
π This newsletter helped me. Thanks!
π It was ok.
π Not useful for me.

