ChatGPT Chronicle is INSANE!
OpenAI has released a feature called Chronicle, part of the Codex app on Mac, which takes periodic screenshots of your screen to build AI memory and context without requiring manual input. The tool aims to eliminate the need for copy-pasting and re-explaining context to AI, functioning like a constantly observant assistant. While powerful for developers and knowledge workers, it raises significant privacy concerns similar to Microsoft's controversial Recall feature.
Summary
The video introduces ChatGPT Chronicle, a newly released OpenAI feature that represents a fundamentally different approach to AI interaction. Unlike traditional AI tools that wait for user prompts, Chronicle actively monitors the user's screen by taking periodic snapshots, extracting text and contextual data from those images, and building a local memory that the AI can reference when answering questions. This allows the AI to already understand what the user is working on without any manual explanation or copy-pasting.
The presenter explains that Chronicle is embedded within the Codex app on Mac and is currently available primarily to ChatGPT Pro users. All data processing and memory storage happens locally on the user's device, and raw screenshot images are deleted after a few hours. The AI retains only the extracted text and context as memory notes, not the full images.
The video walks through practical use cases across different professions. Developers benefit most immediately, as Chronicle can already understand a codebase, active files, and error messages, making debugging faster. Office workers and business users can get help with reports and data summaries without switching tabs to provide context. Content creators can get assistance mid-project without re-explaining their entire workflow.
The presenter also addresses the privacy concerns head-on, drawing a direct comparison to Microsoft Recall, which faced massive backlash and was pulled back after public outcry. The key difference cited is that Chronicle is opt-in by default, users can view, edit, and delete the memories it creates, and data stays on-device — suggesting OpenAI learned from Microsoft's mistakes. Despite these safeguards, the presenter urges viewers to be cautious about what is visible on-screen when the feature is active.
The broader significance of Chronicle is framed as a directional signal for where AI is heading — from reactive to observational and eventually anticipatory. The presenter argues that AI is moving toward always-on, background-running systems that proactively manage tasks, and that early adopters will gain a significant competitive advantage. The video closes with practical tips: developers should enable it immediately, document-heavy workers should close sensitive tabs before use, users should monitor privacy settings carefully, and everyone should watch for the feature's expansion beyond its current research preview stage.
Key Insights
- Sam Altman himself described coding with Chronicle as feeling like 'telepathy,' suggesting the experience of contextual AI assistance is qualitatively different from traditional prompt-response interaction.
- The presenter argues that Chronicle processes everything locally on-device and deletes raw screenshot data after a few hours, distinguishing it from cloud-based surveillance — though the AI-extracted text memories persist.
- The presenter draws a direct parallel to Microsoft Recall, noting that Microsoft faced massive backlash and had to redesign the feature when it attempted something similar, and argues OpenAI preemptively addressed this by making Chronicle opt-in with user-editable memory.
- The presenter frames Chronicle not just as a feature but as a directional signal, arguing AI is transitioning from systems that respond to prompts to systems that observe workflows and eventually anticipate needs and perform background tasks autonomously.
- Chronicle is currently in research preview and limited to ChatGPT Pro users on Mac via the Codex app, but the presenter asserts OpenAI will expand it broadly and that understanding it now provides an advantage while most people are unaware of it.
Topics
Transcript
[0:00] Chat GPT Chronicle is insane. What if I told you there is an AI that already knows what you are doing on your screen right now? No typing, no explaining, no copy-pasting. It just knows. And today I'm going to show you exactly how it works. This is brand new. This is huge. And honestly, it is a little bit scary. But if you use it right, this could change how you work forever. Let me break it all down. So right now OpenAI just dropped something called Chronicle. And look, I have covered a lot of AI updates on this channel. But this one is different. This one actually made me stop and think. Because Chronicle does not…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Julian Goldie SEO
NEW Nvidia Autonomous AI is WILD!! 🤯
Nvidia announced Nemo Clo, a new autonomous AI agent system that operates independently without continuous prompting. Powered by Nemotron 3 Ultra (a 550 billion parameter model), the system is five times faster and cheaper than previous versions, with OpenShell providing secure sandboxed execution.
Laguna XS 2.1: New FREE + Opensource Local AI!
Julian reviews Laguna XS 2.1, a new free open-source local AI coding model from Poolside that performs comparably to Qwen 3.6 and outperforms Claude Haiku on benchmarks. He demonstrates its practical capabilities by building landing pages and functional apps, highlighting its speed, offline functionality, and multiple deployment options through local setup, Claude Code, or OpenRouter's free API.
How to Run Hermes FREE Forever!
The video demonstrates how to run the Hermes AI agent for free using Gemma 4, a local open-source model from Google, with significant speed improvements through MLX optimization. The setup works on Apple Silicon Macs or via free APIs on Open Router, enabling autonomous agents to work offline and privately without subscription costs.
This NEW Chinese AI is INSANE! (FREE + Open Source!)
Long Cap 2.0 is a new open-source Chinese AI model from a food delivery app company that offers 1 million tokens of free context memory, beats GPT-4.5 on SWE bench pro benchmarks, and uses efficient parameter activation to reduce computational overhead while maintaining high performance.
Claude Code is now FREE: Here’s how…
Google's new Gemma 4 model running on Ollama is 90% faster on Apple Silicon, enabling free Claude Code usage locally without token costs. The setup requires three simple steps: downloading Ollama, Gemma 4, and installing into Claude Code, with alternatives available via OpenRouter API for non-Mac users.