OpenAI Announces ChatGPT Projects for Free Users
Today's curated insights on AI product management from X/Twitter across 60+ expert sources and YouTube channels.
OpenAI Announces ChatGPT Projects for Free Users
From X
AI Product Launches & Updates
Voice-Driven Code Creation: Lovable Dev @lovable_dev introduced Lovable Voice Mode, enabling users to turn ideas into reality without touching the keyboard.
Expanded ChatGPT Projects: OpenAI Team @OpenAI announced that Projects in ChatGPT are now available to Free users, adding larger file uploads per project (up to 5 for Free, 25 for Plus, 40 for Pro/Business/Enterprise) and project-only memory controls.
Comet Browser for Students: Aravind Srinivas @AravSrinivas shared that Comet is now available for all students with features like Comet Assistant, Flash Cards, Ad Block, and Study Mode.
AI Tools & Applications
AI Tool Bundles: Lenny Rachitsky @lennysan revealed that paid newsletter subscribers get over $10,000 in tech tool credits including Lovable, Replit, n8n, Bolt, Linear, Descript, and more.
ChatGPT for PDF Conversion: Claire Vo @clairevo demonstrated using ChatGPT to convert poorly formatted calendar PDFs into .ics files and share via WhatsApp group.
Product Management Insights & Strategies
Weekly Newsletter Recap: Lenny Rachitsky @lennysan shared an executive summary linking to key insights from this week’s newsletter to help PMs stay informed.
Community Spotlight on SEO: Bolt Dot @boltdotnew announced a live session with Conor Martin to discuss SEO strategies and demo projects like an Irish Tax Calculator built in Bolt.
AI Industry Developments & News
Promote Your AI Tool: Theresa Forit @theresanaiforit invited AI tool creators to describe their tool in one sentence for promotion on TAAFT and in a newsletter reaching over 1.7 million readers.
Startup Acquisition Insights: Lenny Rachitsky @lennysan raised questions about Wiz’s re-acceleration versus Base44’s acquisition success in the tech industry.
From YouTube
I copied TikTok trends and made millions with AI
Greg Isenberg • September 03, 2025
In this episode, Greg Isenberg interviews Ben Benkhin about how he applied a “copy-and-spin” strategy—identifying viral TikTok trends and open-source AI models—to build simple, four-screen mobile apps like WOMBO and Dream that amassed over 250 million downloads and secured Nvidia backing.
Key Takeaways:
- WOMBO launched in March 2021 by wrapping the first-order motion model into a four-screen app (selfie upload, song selection, loading, output) and hit 50 million downloads in its first three months with zero marketing spend.
- Benkhin pre-selected 15 iconic, meme-friendly songs for WOMBO’s launch to engineer a viral content loop—users generated and shared AI-powered sing-alongs on Instagram and TikTok, driving organic growth.
- Dream, built on VQGAN+CLIP with fast inference and prompt templating, became the #1 art app and grew to over $500 K/month by offering subscriptions ($5/week to $5/month) alongside ads.
Nano Banana API Part 2: Adding New AMAZING Features to The App!
All About AI • September 03, 2025
In this video, All About AI extends the Nano Banana API app by adding a UI Designer for restyling React components, a Celeb Selfie feature to merge user and celebrity photos, and a Fake Holiday mode that generates realistic vacation images.
Key Takeaways:
- The UI Designer feature lets users upload a TSX component screenshot and code plus a style prompt (e.g., '90s retro matrix font'), then uses Nano Banana API for a prototype and GP5 to output updated TypeScript code.
- The Celeb Selfie mode processes uploads of a celebrity image and the user’s image by removing backgrounds, applies a hardcoded random scene prompt (like a concert or New York street), and returns a merged handheld-style selfie.
- The Fake Holiday feature generates four realistic images—two selfies and two location shots at landmarks like Tokyo’s famous Shibuya Crossing—by uploading a user photo and specifying a destination.
Claude Code Beginner's Tutorial: Build a Movie App in 15 Minutes (2025)
Peter Yang • September 03, 2025
Peter Yang walks through installing Claude Code in both terminal and Cursor IDE, cloning a React-TypeScript movie discovery app, configuring the TMDB API key, and then using Plan mode and a claw.md file to spec and implement a “watch list” feature in a single session.
Key Takeaways:
- Claude Code installs with one CLI command and can be launched inside Cursor by typing “claude” and trusting the project files, enabling AI-driven dev in your IDE.
- Plan mode (Shift+Tab) generates a phased feature spec—requirements, design, tech stack—that you must review and trim before implementation to prevent scope creep.
- Creating a claw.md file seeds persistent project context and best practices (env setup, testing steps, commit messages) for every Claude Code session.
Ranking 15 PM Skills: What Survives vs. Gets Disrupted by AI | Nan Yu (Linear)
Peter Yang • September 01, 2025
In this episode, Nan Yu and Peter Yang rank fifteen product management skills by AI resilience—separating emotional and strategic strengths from tasks AI will automate—and demonstrate Linear’s AI agents for issue triage, code inspection, and automated pull requests.
Key Takeaways:
- Linear scaled to over 15,000 companies, including OpenAI and Ramp, with only two PMs by focusing hires on strategic ownership and growth potential rather than mechanical tasks.
- Emotional skills—product taste, branding, stakeholder management, EQ, and risk appetite—will endure, while AI disrupts structured tasks like strategic prioritization, data analysis, market research, and documentation.
- New essential PM skills include context engineering, AI workflow and agent design, and evaluation & QA, enabling PMs to orchestrate AI tools effectively within their product processes.