ChatGPT is now your personal stylist
People are using ChatGPT’s new image model to generate full hair, color, and makeup analyses from a single photo. The outputs look like real consultancy reports, with side by side comparisons and clean visuals.
Instead of guessing what suits you, the model creates structured visuals showing what works and what does not. It focuses on contrast, undertones, face shape, and styling variations, all in one image.
This turns ChatGPT into a visual decision tool, not just a text assistant. And it works in seconds with simple prompts.
This could replace how people choose their look.
Would you trust AI to style you?
Prompts:
🌈 Create a personal color analysis graphic using this portrait. Show side-by-side clothing color comparisons to highlight which colors suit the subject best. Make it visual-first, with short labels only and no paragraphs.
💇🏼♂️Create a hairstyle analysis graphic using this portrait. Show side-by-side hairstyles comparisons to highlight which hairstyles suit the subject best. Make it visual-first, with short labels only and no paragraphs.
💄Create a visual makeup analysis graphic using this portrait. Feature side-by-side comparisons to determine which makeup best suits the subject, and showcase their skin undertone. The graphic should be visual-first, using only short labels with no paragraphs.
“Made by humans” is now a flex.
Some films and TV shows are starting to add labels in their credits saying no generative AI was used. Others are blocking their content from being used to train AI models.
This isn’t a formal rule yet, but it’s spreading across indie projects, big studios, and streaming. What began as a small signal is turning into a clear creative stance.
The push is coming from creators, unions, and a need to stand out as AI tools become more common. In a world full of generated content, human work is being positioned as something premium.
Even when AI isn’t used, it’s still shaping how projects are marketed.
Do you think “no AI” will become a real standard or just a branding move?
Source: Forbes
Want better AI answers? Be rude.
A new Penn State study found that ChatGPT actually performs better when you drop the “please” and “thank you.”
Across 250 test prompts, “very rude” questions scored 84.8% accuracy, while “very polite” ones lagged at 80.8%.
Researchers think polite phrasing like “Could you please tell me…” adds extra words and confusion, while blunt commands like “Tell me the answer” make your intent crystal clear.
It’s a funny twist: the same tone that hurts feelings in humans might help clarity in machines.
Do you think it’s a good idea to be rude with AI?
Paper: Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)
A developer built a tiny hardware dashboard for Claude Code power users.
The gadget is called Clawdmeter, and it turns Claude Code usage into something you can literally keep on your desk: live token stats, weekly usage, quick controls, and pixel-art animations that get more chaotic as usage climbs.
The idea came from software developer Hermann Haraldsson, who reportedly used Claude itself to help build the embedded device in just a few days.
Claude is now helping people build hardware to monitor how much they are using Claude.
It is not an official Anthropic product, but it says a lot about where AI coding is heading: more ambient, more physical, and more deeply woven into the developer workspace.
What do you think?
Source: TechCrunch
OpenAI is bringing personal finance into ChatGPT.
The new experience lets U.S. Pro users connect financial accounts through Plaid, so ChatGPT can help analyze spending, subscriptions, upcoming payments, debt, and portfolio activity using real financial context.
Instead of asking generic money questions, users can ask things like where their money is going, what subscriptions they forgot about, or how their spending changed this month.
For now, it is read-focused. ChatGPT can help explain and organize your finances, but it cannot move money or make transactions.
AI assistants are slowly becoming the front door to everything: email, work files, health, and now your bank account.
What do you think?
Source: OpenAI
Like this post + comment “HIGGS” and I’ll DM you all of them
We tested 10,000+ AI photo edits on Higgsfield to find the best prompts that actually work.
Photo restoration, body reshaping, remove watermarks, professional portrait, famous selfie, 4K enchancer, product ad, anime effect.
Most people are just guessing.
These prompts are based on what actually works on Higgsfield.
Steal them, tweak them, and use them on your own photos.
Save this for later.
A man says he used Claude to wipe a huge part of his digital footprint in just 48 hours.
He reportedly used it to search what was visible online, organize data broker listings, draft opt-out requests, delete old accounts, suppress unwanted results, and prioritize leaked data from past breaches.
Claude did not magically erase the internet. It acted like a privacy assistant, turning a messy, exhausting process into a step-by-step cleanup plan.
AI is becoming useful outside the chat window. It can help people navigate forms, legal language, account deletion processes, and repetitive digital admin that most people never finish.
In the past, privacy cleanup felt impossible. Now it might just need the right prompt.
What do you think?
Source: digital ghost/X
Codex just went mobile
OpenAI is bringing Codex into the ChatGPT mobile app.
You can start coding work, review results, guide the task, and approve next steps from your phone.
The actual work still runs on your laptop, Mac mini, or devbox. Your phone acts like a remote control for Codex while it keeps working on the machine where your project lives.
This makes coding agents feel less tied to your desk.
Would you use this from your phone?
Source: OpenAI
China’s Nvidia deal is stuck
The U.S. has reportedly cleared around 10 Chinese firms to buy Nvidia’s H200 chips. H200 is an advanced AI chip used to train and run powerful AI systems.
The approved names include Alibaba, Tencent, ByteDance and JD.com. Lenovo and Foxconn have also reportedly been cleared as distributors, but no deliveries have happened yet.
The holdup is coming from Beijing, where officials are pushing companies to rely more on Chinese chips. That puts Nvidia in the middle of a political fight between U.S. export approvals and China’s tech independence goals.
This shows how messy the AI chip race has become, even when both sides technically open the door.
Would China rather slow down than depend on Nvidia?
Source: Reuters
An X user says he got banned from Claude after asking it to help make a hantavirus vaccine.
AI models are becoming good enough to discuss advanced biology, medicine, and research in a way that feels powerful. That makes anything involving viruses, vaccines, pathogens, or lab work much more sensitive.
Even if the intention sounds positive, AI companies have to draw hard lines around biosecurity. A model that can explain helpful science too well can also create dangerous shortcuts.
This is the strange new balance: smarter AI, stricter safety walls.
What do you think?
Source: birdabo/X
The internet judged a real Monet.
A viral X post asked people to critique what looked like an AI generated painting. The image even had a fake “Made with AI” tag, which made the setup feel believable.
People quickly started tearing it apart, saying it had no depth, no intention, weak composition, and looked nothing like real art. Then the twist landed, the painting was actually by Claude Monet, created around 150 years ago.
The thread became a perfect example of how fast people judge art when they think AI made it.
Would you have fallen for it?
Source: X / @SHL0MS
Nostalgia makes every photo feel like it came from another life.
These 12 prompts turn any portrait into a completely different era, from VHS home videos and Polaroids to PS1 graphics, MySpace profile pics, tabloid magazine scans, and 1980s pixel art.
The trick is to use the photo as the identity reference, then change the world around the person. Add a role, a scene, an action, an outfit, and a mood, and the result feels much more intentional than a simple filter.
Save this post for your next photo experiment.
Which one would you try first?
Comment “Nostalgia” and we’ll send you the link
