In partnership with
Welcome Automaters, 👋
So while the robots were busy trying to take over the world (or at least our grocery lists), we’ve been tracking the data. And the results are in. It turns out, building a cool AI app is like dating a magician: the first date is mind-blowing, but eventually, you just want someone who can remember where you left your keys.
And word is: AI apps can’t even keep a date
According to a spicy new report from RevenueCat (the 2026 State of Subscription Apps), AI-powered apps are having a serious retention crisis. While these apps are seeing record-breaking download numbers, the long-term loyalty is a bit of a disaster.
Here is the cold, hard data from a study of 75,000+ developers:
The Good: AI apps are absolute beasts at conversion. They convert trials to paid at 8.5%, which is 52% higher than regular apps.
The Bad: The bucket is leaky. Annual subscribers cancel their plans for AI apps 30% faster than for non-AI apps.
The Ugly: At the one-year mark, AI apps only keep 21.1% of their users, while the "boring" regular apps hold onto 30.7%.
Why Is This Happening?
We’re suffering from a severe case of "Novelty Fatigue." Everyone has released an AI "assistant" that basically just puts a fancy hat on an LLM, and users are starting to notice. But there’s more:
The Photo & Video Glut: These apps are the most "AI-heavy" (over 60% use it), but because there are so many, users just hop to the next shiny tool as soon as a better filter drops.
The Novelty Gap: Once the "magic" wears off, users realize they don’t actually need a chatbot to help them choose a socks-and-sandals combo every morning.
Refund Regret: AI apps have a 4.2% refund rate, which is 20% higher than average. So yeah, people pay for the magic, realize it’s just math, and ask for their lunch money back.
The Prompt Fatigue: If an app makes you "prompt" it for ten minutes to do a two-minute task, humans (who are notoriously lazy) will just go back to the old way.
The Bottom Line:
AI adoption is happening at breakneck speed, but developers are learning the hard way: you can't build a kingdom on a "cool trick." To survive 2026, AI needs to move from being a "fun toy" to a "utility."
Think of it as the digital version of a toothbrush: it’s not flashy, but you’d never dream of going a day without it. The apps that embed themselves into actual daily workflows are the ones that will still be here when the "magic" wears off.
Here's what we have for you today
😎 ChatGPT’s New Interactive Feature And Google’s 3-Million-User Pentagon Deal
ChatGPT just made math and science actually fun (yes, really)
Remember staring at a math equation until your brain turned into mashed potatoes? Well, OpenAI just dropped a lifesaver. ChatGPT can now create interactive visuals!
On Tuesday, they introduced "Dynamic Visual Explanations," a brand new feature that lets you interact with math and science concepts in real time. Instead of just telling you how a concept works, it can actually show you a simulation you can play with. It’s like moving from a dusty old textbook to a video game. Whether you’re trying to visualize a physics/math problem or see how a cell breathes, you can now poke, prod, and slide elements around in real-time.
Here’s the magic:
When you ask about the Pythagorean theorem, you can literally grab the sides of a triangle and drag them, watching the numbers update live. It is basically a video game where the final boss is your calculus midterm.
The Stats:
It covers 70+ Topics, we’re talking everything from Ohm’s law, Charles’ law, linear equations and kinetic energy to compound interest.
140 Million users are already using ChatGPT for Math and Science help every single week.
The Goal: Moving from a "cheatsheet" to an "interactive tutor" that actually makes you engage with the logic.
Meanwhile, in considerably heavier news: Google is rolling out Gemini AI agents to the Department of Defense's 3 million employees.
This isn't just a chatbot; it’s a full-scale deployment of "Agent Designer," a no-code tool that allows military and civilian staff build their own digital assistants for the boring stuff (summarizing meeting notes, building budgets, and checking project plans against national defense strategy).
The "Receipts":
1.2 Million DOD employees have already been testing Google’s tools since December.
These users have already uploaded over 4 million documents into the system and each used up 40 million prompts
The Future: While it’s currently on unclassified networks, talks are already happening to move Gemini into the "Top Secret" cloud.
Also, timing-wise… this move is spicy.
Anthropic was recently labeled a “supply chain risk” by the United States Department of Defense after refusing to strip out its guardrails against things like mass surveillance and autonomous weapons.
While Anthropic is busy suing the administration, Google ( which quietly tweaked its AI principles back in February to be a bit more “military-friendly”) is stepping right into the gap. The signal couldn’t be louder: AI isn’t just headed to office workflows… it’s already embedded inside some of the world’s most powerful institutions.
But will this transition be smooth? Not exactly looking that way. Around 900 Google employees have already signed an open letter urging the company to keep the same guardrails Anthropic refused to drop. With this new move, Google might be heading for both internal and external turbulence, because when the military starts knocking, things tend to get complicated… fast.
So yeah, buckle up. This one might turn into a full-blown showdown.
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.
🧱 Around The AI Block
🤑 Meta is buying Moltbook, the ridiculous social network populated by AI bots.
👍 Google's Gemini AI hits state-of-the-art in sheets.
😳 White House preps Executive Order targeting Anthropic.
🤝 Thinking Machines Lab inks massive compute deal with Nvidia.
👨💻 Adobe is debuting an AI assistant for Photoshop.
👨 Zoom launches AI office suite with avatar stand-ins.
🧑⚖️ Family of child injured in Canada school shooting sues OpenAI.
🤩 Google rolls out new Gemini capabilities to Docs, Sheets, Slides, and Drive.
🕵️♂️ YouTube expands AI deepfake detection to politicians, government officials, and journalists.
🛠️ Trending Tools
For the Visual Polishers: PicFixer is your one-click "studio upgrade" for professional branding. It transforms basic selfies into high-end LinkedIn portraits while fixing lighting and resolution issues, and it can even extract signatures or turn text into stylized movie posters.
For the Instant Creators: Pexo acts as a chat-based creative partner for social ads and UGC content. Just describe your vision in plain English, and it autonomously handles the script, storyboard, voiceover, and transitions to hand you a finished video in minutes.
For the Digital Lifesavers: Wondershare Dr.Fone is the ultimate "all-in-one" toolkit for when your tech hits a wall. It specializes in recovering lost WhatsApp chats, repairing iOS/Android system glitches, and safely transferring your entire digital life between phones without losing a single message.
For the Global Educators:Voila Voice transforms any static PDF, slide deck, or webpage into an interactive, multi-language video lesson. It adds AI avatars, professional voice-overs, and quizzes automatically, making it the fastest way to turn "content" into a "course."
Hope these tools help you reclaim some of your time today!
🤖 AI Workout Of The Day: How to Get Gemini Singing Your Tune
AI Generated
In case you missed it in the heaps of daily AI updates, Google Gemini just turned into a full-blown recording studio. And no, we’re not talking about those bleep-bloop MIDI sounds from 1995. We’re talking high-fidelity tracks with realistic vocals, instruments, and even AI-generated album art to match.
The "February 2026 Drop" officially integrated Lyria 3, Google DeepMind’s most advanced music model, directly into the Gemini app. So whether you need an Afrobeat track about your dog’s obsession with socks or a lo-fi beat for your "Deep Work" sessions, Gemini handles the lyrics, melody, and production while you just sit back and play "Producer."
How To Use It?
Here is the dead-simple way to get Gemini singing:
Step 1: Open the Gemini app (Android, iOS, or Web) and look for the "Music" tool in the Tools menu.
Step 2: Type your prompt. Be specific! Instead of "sad song," try: "A melancholic indie folk song about a long road trip, female vocals, fingerpicked guitar."
Step 3: Use an image! You can actually upload a photo of your vacation or your lunch and ask Gemini to "Create a 30-second tropical house track that matches the vibe of this photo."
Step 4: Hit Generate. In about 30-60 seconds, you’ll get a 30-second track complete with Nano Banana 2 generated cover art with vocals and lyrics.
Step 5: Download as an MP3 or MP4 and send it straight to your friends.
Also: While it won't let you mimic Drake or Taylor Swift (it has strict filters against artist impersonation), it will take "broad inspiration" from genres to help you find your sound. And yes, all audio includes SynthID watermarking for verification plus, it’s free for users 18+ in 8+ languages (English, Spanish, French, etc.), with higher limits for Google AI Plus, Pro and Ultra subscribers.
Pro-Tips for Perfection:
The "Mood" Trick: Use adjectives like crunchy, airy, bright, muddy, or crispy. Gemini’s Lyria model translates these "texture" words into specific EQ settings.
Structure Your Song: If you want a specific flow, tell it: [Intro] -> [Heavy Drop] -> [Outro].
Use The Language Swap: Gemini is multilingual! Try asking for a "Bossa Nova track with Portuguese vocals" or "K-Pop with a mix of English and Korean." It handles the linguistic flow better than almost any other tool.
💡 Prompts To Try:
For "Viral TikTok" Vibe (High Energy):
Prompt: "Create a 30-second [Phonk/Hyperpop] track. Tempo: 140 BPM. High-energy, distorted bass, and fast-paced percussion. Vocals: Pitched-up male lead with a slight robotic filter. Vibe: Competitive, dark, and energetic. Lyric Theme: Winning the final level of a video game."
The "Nostalgic Indie" Soul (Vocal Focus):
Prompt: "Generate a soulful [Indie Folk] song. Instruments: Solo acoustic guitar with soft reverb. Vocals: Breathy, intimate female lead—think 'whispered' singing style. Pacing: Slow and deliberate. Atmosphere: A rainy Sunday morning in a coffee shop. Lyrics: Finding a forgotten polaroid in an old book."
The "Cyberpunk Synth" (Instrumental & Mood):
Prompt: "A cinematic [Synthwave] track inspired by 80s sci-fi. Instruments: Heavy analog synths, gated reverb drums, and a driving bassline. Vocals: Distant, ethereal harmonies (no lead lyrics). Tempo: 110 BPM. Structure: Start with a slow build-up and end with a high-intensity pulse."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.
