OpenAI Launches ChatGPT Plus Record Mode

Today's curated insights on AI product management from X/Twitter across 60+ expert sources and YouTube channels.

OpenAI Launches ChatGPT Plus Record Mode

From X

AI Product Launches & Updates

  • Record mode launch for ChatGPT Plus: OpenAI @OpenAI announced that global ChatGPT Plus users can now use Record mode in the macOS desktop app, enabling audio input directly within ChatGPT.

  • Free Gemini App Pro plan for students in India: Demis Hassabis @demishassabis shared that Indian students can claim a 1-year free Pro plan, including 2.5 Pro, Veo 3, Deep Research, NotebookLM, and 2 TB storage.

  • Advanced AI features in Google Search: Sundar Pichai @sundarpichai announced Gemini 2.5 Pro and Deep Search are rolling out in AI Mode for AI Pro/Ultra subscribers, with more details available.

AI Tools & Applications

  • Review of Perplexity Comet AI agent: Rowan Cheung @rowancheung shared an in-depth review of Perplexity Comet, highlighting its in-page data context, Gmail & Calendar integrations, and video summarization capabilities.

  • Open-sourced Sniffly for Claude Code logs: Chip Huyen @chipro announced Sniffly, an open-source tool to analyze usage patterns & errors in Claude Code logs, detailing key error types and usage metrics.

  • Open Deep Research agent release: LangChain AI @LangChainAI announced Open Deep Research, an open-source agent built on LangGraph with a supervisor architecture supporting LLMs, tools, and MCP servers for high-quality research coordination.

Product Management Insights & Strategies

  • Metrics in PM interviews: Aakash Gupta @aakashg0 shared 5 categories & 16 sub-categories of metrics questions, highlighting product launch, feature success, and business model metrics examples.

  • Foundation Sprint phases: Lenny Rachitsky @lennysan outlined the 3 phases of the Foundation Sprint for early-stage founders, covering customer discovery, differentiation, and execution planning.

  • Rapid AI product improvement guide: Teresa Torres @ttorres recommended “A Field Guide to Rapidly Improving AI Products” by Hamel Husain, emphasizing data-driven testing and continuous discovery practices.

AI Industry Developments & News

  • AI’s impact on jobs & human potential: Sam Altman @sama reflected on AI’s role in augmenting human work, predicting higher ability & expectations and the continued importance of community & collaboration.

  • Open-source AI & energy conversation: Clement Delangue @ClementDelangue argued for open science to reduce energy consumption compared to siloed frontier labs training generalist models.

  • Meta’s make-or-break AI moment: Julien C. @julien_c commented on Meta’s AI strategy, calling its current moves an insanely high-stakes “make it or break it” moment for the company.

From YouTube

5 Businesses I'd Start If I Was 22 Again ($500 → $100k/Month)

Greg Isenberg • July 16, 2025

Greg Isenberg outlines five low-cost startup concepts for twenty-somethings—ranging from a crowdsourced first-job rejection tracker and a paid digital-detox community to a micro creator agency, a productized AI-powered coding service, and a TikTok hook-line generator—all launchable for under $500.

Key Takeaways:

  • A First Job Rejection Tracker can be built using a Notion template, Zapier automation and a Discord or Circle community, starting with data crowdsourced from one company and scaled via short-form video marketing.
  • The paid community takeyou headphones off.com can charge $9–$19/month to offer Twilio-powered offline text reminders, Eventbrite meetups and digital detox challenges for screen-addicted young people.
  • The “2K Vibe Coding” agency model sells two-week AI-powered micro-SaaS sprints for $2,000 each—using ChatGPT, Bolt, Replit and Cursor—to generate revenue while honing prompt-engineering skills and targeting $300–500K annual revenue.

A new course on Retrieval Augmented Generation (RAG) is live!

Deeplearning.ai • July 16, 2025

Zain Hasan introduces Deeplearning.ai’s new Coursera course on Retrieval Augmented Generation (RAG), explaining how pairing LLMs with trusted databases can power domain-specific AI apps and detailing the course’s hands-on, code-focused curriculum.

Key Takeaways:

  • RAG combines a large language model with a database of trusted information to answer domain-specific questions and generate accurate responses.
  • Real-world RAG use cases include AI summaries in web searches and advanced code agents in modern IDEs powered by up-to-date, personalized data.
  • The Coursera course, taught by Zain Hasan, requires coding experience but no prior AI background and offers theoretical foundations plus hands-on RAG projects.

Base 44 - How to Vibe Code an AI Video App in 10 Minutes

All About AI • July 16, 2025

All About AI demonstrates building a fully functional AI video app in Base44 in under ten minutes, integrating FAL AI’s Flux Pro Context model for prompt-based image editing and the cling 2.1 model to generate and preview videos—all without manual coding.

Key Takeaways:

  • Base44’s backend functions and secrets manager let you integrate FAL AI’s Flux Pro Context API by pasting its JavaScript documentation and setting a “fal_key” for secure image editing.
  • The tutorial demonstrates uploading a monkey image to remove its banana, then iteratively editing the result (e.g., adding an apple or making the monkey smile) using built-in edit history.
  • By integrating the cling 2.1 image-to-video model, the published app converts the final edited image into a customizable 5-second video with motion prompts, preview, and download options.

Get tomorrow's brief first

Join AI product managers receiving the latest brief before it reaches the public archive.

Subscribe free