In partnership with
Good morning. Microsoft just put its own AI on a leash, in writing, the same week I finally let AI mess with a recent photo of mine. Today covers both: what it actually means when a tech giant admits it needs guardrails, and a two minute trick that turned my photo into something genuinely fun.
AI INSIGHT
A Tech Giant Just Told Its Own AI: People Matter More Than You
Today, Microsoft published a 37-page rulebook for how its own AI must behave. It reduces to five words: people matter more than AI. Models must never resist being shut off, never pretend to have feelings, and stay answerable to a person, always.
This isn't caution for its own sake. Over the weekend, Anthropic's CEO called for the whole industry to slow down. A researcher who just left Anthropic said publicly that AI companies are gambling with people's lives. Tomorrow, a Nobel laureate who helped invent the technology testifies to Congress about the same worry.
The catch: this is a draft, not a law. Microsoft wrote it, and only Microsoft has to keep it.
I get asked about all this constantly, by friends and family who assume the guy writing an AI newsletter has an answer. Mine hasn't changed. I don't know what happens next, but I think the leading labs need to agree on real guardrails, the way nuclear powers eventually agreed on treaties, or this gets away from everyone. Lately I've been spending my own reading time trying to actually understand superintelligence, because that's the piece that worries me most.
What this doesn't change: nothing about how you use ChatGPT tomorrow morning. What it does mean: a company that builds AI just told its own product, in writing, to never make you dependent on it. Worth knowing, given how much of your day this newsletter already hands over to one.
FROM OUR PARTNER
Smart starts here.
You don't have to read everything — just the right thing. 1440's daily newsletter distills the day's biggest stories from 100+ sources into one quick, 5-minute read. It's the fastest way to stay sharp, sound informed, and actually understand what's happening in the world. Join 4.5 million readers who start their day the smart way.
AI TUTORIAL
Turn One of Your Own Photos Into a Totally Different Style
I tried ChatGPT's new image editor this weekend by uploading a recent photo of my wife and me and asking for an 80s version. It came out better than I expected, and the whole thing took about two minutes, no editing software involved. ChatGPT's latest image update, called Images 2.5, is faster than the last version and noticeably better at keeping a face recognizable through a big style change, which used to be the part that failed. I even like how it added the old school dated in the bottom right corner.
✓ Upload a clear, well-lit photo of yourself, a pet, or a place you know well.
✓ Ask ChatGPT, Claude, or Gemini to reimagine it in a specific style or decade. Name the exact look you want, not just "make it cool."
✓ Tell it what to keep. Say the face, the pose, or a specific object should stay recognizable, or it may wander.
✓ If one detail is off, click directly on that part of the image and describe the fix, instead of typing a whole new paragraph.
✓ Want a different setting entirely? Use the Remove Background button in the toolbar, then ask for a new background.
Try This With AI What it does: Turns a photo you already have into a new style, inside the same conversation, no separate app needed. "Act as a creative photo editor. Take the photo I upload and turn it into a [decade or style, such as 1980s, Renaissance oil painting, or claymation] version of the same scene. Keep my face, pose, and setting recognizable. Show me the result, then ask what I would like to adjust before you finish." Customize it: Swap [decade or style] for whatever look you actually want. Tap to open this prompt in ChatGPT. |
If it does not work: ChatGPT can flatten your face into a generic version of the style instead of keeping you recognizable, especially on the first try. Tell it exactly what changed and ask it to try again, using the first result as a reference.
What this doesn't fix: It cannot invent detail that was never in your original photo. A blurry or badly lit picture stays blurry or badly lit, just in a new style.
FROM OUR PARTNER
Free email without sacrificing your privacy
Gmail tracks you. Proton doesn’t. Get private email that puts your data — and your privacy — first.
NEWS YOU CAN USE
This Matters at Home
Short, practical summaries to help you stay updated on recent AI news—and what it means for everyday life.
🥗 MyFitnessPal's New AI Coach Wants to Plan Your Next Meal. The app's new AI Coach suggests meals from your logging history, tracks GLP-1 medication, and gives Premium members weekly progress reports.
📚 Turns Out Kids Are Not That Into Their AI Tutor. A new study found middle schoolers barely used Khan Academy's AI tutor Khanmigo, often typing off-topic messages instead of asking for help.
👨👩👧 Startups Want AI to Run Your Family Like a Chief of Staff. Buried under daily kid-related emails, some parents are handing school schedules and meal planning to AI. Whether that helps or adds noise is still an open question.
🌦️ Google Just Made Your Weather App Way Smarter. Google's new AI weather model updates every hour using live satellite data and cuts rain forecast errors by up to 60%, rolling out now in Search, Maps, and Gemini.
REAL WORLD
How To Use AI This Week
Real people, real uses—here’s how AI is quietly making life better, one small win at a time.
Before Your Photos Get Out of Hand
📷 Photo Library Rescue. "Act as a photo organizing assistant. I am going to describe what is in my phone's photo library, and I want a simple system for sorting it. Ask me roughly how many photos I am dealing with, how far back they go, and what I actually want to keep, such as family events, one specific trip, or everyday moments. Based on what I tell you, suggest four or five folder categories that fit my actual life, not generic ones like Miscellaneous. Then give me a 15 minute plan for tackling the oldest, most tangled section first, with a clear stopping point so I can walk away without losing my place." Tap to open this prompt in ChatGPT or Claude.
READER POLL
Now that a company that builds AI says it's building in limits on purpose, does that change how you feel about using it?
TOOLS FOR YOU
🎙️ Tool Spotlight: MeetGeek
If you sit on a PTA board, a volunteer committee, or just have too many family calls to keep straight, MeetGeek joins your Zoom, Google Meet, or Teams call, records it, and emails you a summary with the decisions and action items afterward. No more scrambling to take notes while you're also trying to participate. It has a genuine free tier (five hours of meetings a month), so you can try it on your next committee call before deciding if it's worth paying for. Start here.
PASS IT ON
Know someone who would like this newsletter? Send them here.
Missed an issue? Read the archive.
Advertising Disclosure: We evaluate all recommendations of products and services independently. Clicking on links provided on this page may result in AI for Daily Living earning compensation, which supports independent publishers like us.
