In partnership with
Listen up, lovebirds: that ChatGPT-written love note you’re thinking about sending? Yeah, psychologists say it is a terrible idea, and deep down, your brain already knows it.
A study surfaced showing that using AI to write Valentine’s Day messages creates a weird psychological phenomenon called "source-credit discrepancy." Translation: Your brain knows you’re being a phony, even if your partner doesn't.
Scientists ran five different experiments with hundreds of participants and found the same pattern every single time: people felt massively guilty when they let AI write their emotional messages instead of doing the work themselves.
Here is the wild part: it’s not about the tech being bad. It’s about the dishonesty. When you copy-paste an AI love letter and sign your name, you are basically catfishing your partner’s emotions.
The researchers found some fascinating benchmarks for your guilt:
When it comes to store-bought cards, people felt zero guilt, cuz apparently, everyone knows you didn’t write the Hallmark poem yourself.
But when a "ghostwriter" friend does it, there’s high guilt. Turns out having a friend secretly pen your feelings feels just as gross as using AI.
Fun fact: The guilt intensified when messages were sent to close friends and romantic partners. Basically, the more the person matters to you, the more your brain screams "LIAR" when you hit send on that prompt.
So Why Does Effort Actually Matter?
It turns out we have a "personal effort" threshold. The research showed that people react negatively when they learn a person, especially a loved one, used AI for something that should be human, like a friend expressing sympathy after a tragedy. But for purely factual stuff, like announcing routine personnel changes? Nobody really cared.
The verdict: When it comes to your relationship, your partner isn't looking for a Pulitzer-winning sonnet; they are looking for you.
So how do you use AI without being a heartless robot?
If you are staring at a blank page and panicking, you don't have to banish the bot entirely. The researchers suggest using AI as a brainstorming buddy, not a replacement writer.
So Here’s The "Automated-Approved" Love Letter Protocol:
Beat the Block: Use AI to suggest a structure or a few opening lines.
Add the "Human" Sauce: Edit the output heavily. Add those niche inside jokes, that one weird thing they do with their coffee, and the specific memories only y’all share.
Co-create, don't Delegate: Think of the AI as your intern, but you are the Editor-in-Chief of your own heart.
The Bottom Line: AI can craft emotionally resonant messages that sound genuinely heartfelt, but when you present those words as your own, something deep inside you knows it is not right.
So yeah, keep the AI for your coding scripts and your work emails. For the Valentine’s card? Grab a pen and get real, y’all.
Here's what we have for you today
⏹️ How to Turn Off Ads in ChatGPT: New Settings and Usage Limits Explained
Well folks, the moment Sam Altman once called "uniquely unsettling" has officially arrived. Ads are now live in ChatGPT.
OpenAI officially began testing ads yesterday, February 9, for U.S. users on the Free and Go ($8/month) tiers. But here is the kicker: it is a very limited test. Multiple people have been trying to "summon" an ad like some digital ghost, but nobody has quite caught one in the wild yet. Meaning, it’s a slow rollout, so do not expect a pop-up every time you ask for a coding fix just yet.
Oh, and get this: OpenAI is trying something pretty unique here. Instead of just forcing commercials down our throats, they are giving free users a choice:
The Ad-Supported Experience: You keep the "good stuff" like image generation and deep research, but you have to look at some sponsored links.
The "Ads-Free" Toggle: You can switch ads off in your settings entirely, but your usage limits drop and the fancy features get locked away.
Basically, you can pay with your eyeballs or pay with your patience. If you are a Plus, Pro, or Enterprise user? You can breathe easy. Your $20 (or $200) a month still keeps your chat window a pristine, ad-free sanctuary.
The opt-out options right inside ChatGPT’s user settings.
How it looks (and the guardrails)
When ads do show up, they will appear at the bottom of a response, clearly labeled as "sponsored," and separated from the actual answer. OpenAI promises that ads will not influence what the AI tells you, and conversations stay private from advertisers.
They have also put up some serious guardrails. There will be no ads for users under 18, and nothing will appear near sensitive topics like health, mental health, or politics. Ads will also stay away from:
Temporary Chats
Logged-out sessions
After generating images
And the ChatGPT Atlas browser
Also: You can dismiss individual ads, give feedback, delete your ad history and the data used for the ads, and toggle off personalization if you don't want ads based on your chat history.
You can even ask ChatGPT questions about an ad, which is probably one of the coolest thing about all this.
And yes, the logic is pretty simple: running world-class AI is expensive. OpenAI has committed to $1.4 trillion in infrastructure spending over the next eight years, and right now, they are burning through cash. With 800 million+ weekly users and only 5% of them paying, the subscription model alone is not going to cut it.
So Will The Ads Do It?
Well, internal documents show OpenAI is eyeing a $25 billion ad revenue goal by 2030. And guess what? Advertisers are already jumping in, even with a $200k minimum buy-in and rates significantly higher than what you would pay on Meta. But then, when you have high-intent users asking specific questions, that ad space is digital gold.
The Bottom Line:
We are all watching to see if this feels "uniquely unsettling" or just a minor annoyance. If the experience gets too invasive, expect a mass exodus to Claude or Gemini. But for now, OpenAI is betting that we will trade a little screen real estate for world-class intelligence.
Wanna know more about how to control those ads? You can find out more here.
The CRM that saves teams hours every week
It's not about working harder — it's about having a CRM that actually thinks ahead. HubSpot Smart CRM learns how your team operates and adapts to make everyone more effective. Streamline day to day tasks and track the activity that actually matters to your business. The result? Your team gets back hours every week to spend on growth instead of admin work. Start free. See the difference.
🧱 Around The AI Block
🪴 AI lets botanic garden visitors chat with plants.
⚠️ AI scams are appearing more genuine, security expert warns.
🤑 Amazon plans to launch AI content marketplace.
🦾 Crypto.com CEO unveils new AI agents to millions during Super Bowl.
❌ New York lawmakers propose a three-year pause on new data centers.
🧑⚖️ EU threatens to act over Meta blocking rival AI chatbots from WhatsApp.
A new space-based MMO just like Moltbook, is designed exclusively for AI agents.
🛠️ Trending Tools
For the Streamers: FragCut uses Gemini 2.5 Flash to automatically detect high-skill plays, multi-kills, and funny moments in your gaming footage. It then auto-crops them into vertical 9:16 clips for TikTok and Reels, complete with captions and hashtags, and processes it all in the cloud.
For the Growth Hackers: ThumbnailCreator uses "face-aware" generation and style cloning to build professional YouTube thumbnails in seconds. It’s trained on high-CTR patterns, allowing you to generate 20 variations of a single concept so you can A/B test what actually drives clicks without opening Photoshop.
For the Narrative Visionaries: Seedance 2.0 (by CapCut) allows you to create high-resolution (up to 2K) cinematic videos from a single sentence. The real "Heavy Use" feature? You can combine up to 12 references—images, videos, and audio—to maintain perfect character consistency across different scenes. It’s the closest we’ve come to a "movie studio in a prompt."
For the Data Archivists: GLM-OCR is a multimodal powerhouse that understands semantic context. It doesn't just "see" letters; it understands document structure and outperforms standard tools on low-quality images and complex layouts. It’s perfect for digitizing messy archives or high-volume business receipts.
Hope these tools help you reclaim some of your time today!
🤖 AI Workout Of The Day: How To Turn Any Boring PDF Into Whiteboard Visuals (In Literal Seconds)
ChatGPT Generated
Ever stared at a 50-page PDF and wished it would just turn into a picture? Well, grab your digital markers because AI just went full Picasso.
Whether you are using ChatGPT, Claude, or Gemini, you can now instantly transform dense blocks of text into vibrant, whiteboard-style visuals. We are talking flowcharts, mind maps, and "napkin sketches" that actually make sense. No more "death by bullet points."
Here is the "How-To" Breakdown:
The NotebookLM Route: Upload your file to Google’s NotebookLM. Hit the auto-suggested "Mind Map" chip and watch instant branching diagrams appear. It is perfect for study guides or research papers.
The ChatGPT/Claude Approach: Upload your PDF and prompt: "Turn this into a whiteboard-style infographic with hand-drawn elements." Claude is particularly cracked at this if you use the "Artifacts" UI.
The Pro Move: Use Jeda.ai to upload PDFs and run multiple models like GPT and Llama at once. It merges the best ideas into one visual workspace with sticky notes and custom graphics.
Other Tips and tricks:
Claude 3.5/4.5 Sonnet (via Artifacts) and Napkin.ai are the current gold standards for turning text into diagrams.
Ask the AI for SVG format or Mermaid.js code so you can edit the shapes later.
ChatGPT Enterprise and Edu users can use "Visual Retrieval" to pull actual data from charts inside the PDF into the new drawing.
Try this prompt structure:
You are a world-class explainer who turns dense documents into clear, visual thinking. I’m pasting a PDF about: [topic]. For my audience: [who it’s for] Your goal is to deconstruct this dense information into a high-fidelity, whiteboard-style visual breakdown.
Task:
-Pull out the core ideas (no fluff). Put the main concept in the middle and branch out into 3 to 5 key pillars.
-Explain each idea like you’re teaching it on a whiteboard
-Use short bullets, simple language, and clear sections
-Use hand-drawn aesthetics. Include arrows to show relationships, 'sticky note' boxes for side-bars, and simple icons (like lightbulbs for ideas or gears for processes) to represent concepts.
-Add metaphors, diagrams-in-words, or step-by-step flows where helpful
Output:
If you have 'Artifacts' or a code execution environment, render this as a clean, interactive SVG or a Mermaid.js diagram or just give me the code. If you are generating an image, ensure the text is legible and follows a clear visual hierarchy.
Let's start: [Describe the specific part of the PDF you want to focus on here]."Is this your AI Workout of the Week (WoW)? Cast your vote!
That's all we've got for you today.
Did you like today's content? We'd love to hear from you! Please share your thoughts on our content below👇
What'd you think of today's email?
Your feedback means a lot to us and helps improve the quality of our newsletter.
