- AI PlanetX
- Posts
- GPT-5 Details Leaked by GitHub
GPT-5 Details Leaked by GitHub
Google Rivals ChatGPT Study Mode

Welcome to another edition of AI PlanetX.
GitHub leaks GPT-5 coding upgrade; Google unveils Guided Learning; OpenAI drops local-use LLMs rivaling o4-mini.
Inside This Edition: đź’Ž
Hottest AI News
Top AI & SaaS Tools
AI Tutorial: Turn YouTube Videos into Audio or Video Podcasts
Top AI & Tech News
Resource of the Day: OpenAI’s Fine-Tuning Guide for Open Models
AI Video Tutorial
F-R-E-E AI Course of the Day: OpenAI Academy Courses
Hottest AI News
OpenAI
GitHub Accidentally Leaks Details on OpenAI GPT-5 with Advanced Coding Abilities

GitHub accidentally leaked OpenAI's upcoming GPT-5 announcement through a prematurely published blog post that was quickly deleted. However, Reddit users managed to capture and archive the details before it disappeared, revealing comprehensive information about the next-gen AI models.
Details:
GitHub’s leaked post confirms four GPT-5 variants: the flagship for logic-heavy tasks, gpt-5-mini for cost-efficiency, gpt-5-nano for speed, and gpt-5-chat for advanced multimodal conversations. All promise better reasoning, coding, and user experience
GPT-5 brings stronger agentic abilities, handling complex coding with minimal prompts. It marks a leap in autonomous problem-solving, helping developers code faster with less input
The leak is well-timed — OpenAI teased “a big week,” Altman hinted at GPT-5, and a “LIVE5TREAM” event is set for today. It follows two recent GPT-OSS model releases, including one that runs locally
The accidental leak confirms earlier reports of GPT-5 launching in early August, with mini and nano variants available through the API—setting the stage for one of the most anticipated AI releases this year.
Join 400,000+ executives and professionals who trust The AI Report for daily, practical AI updates.
Built for business—not engineers—this newsletter delivers expert prompts, real-world use cases, and decision-ready insights.
No hype. No jargon. Just results.
DeepMind
Google Unveils Guided Learning in Gemini to Rival ChatGPT Study Mode

Google is stepping up its AI education game with the launch of "Guided Learning" in Gemini, directly challenging OpenAI's recently released ChatGPT Study Mode. This move signals a major shift - from simple answer engines to comprehensive learning partners.
Details:
Google’s Guided Learning acts as an AI tutor that breaks down problems step-by-step and adjusts to each user’s needs. It uses visuals, videos, and quizzes to explain concepts, helping users learn the "why" and "how"
Gemini adds visual aids and YouTube videos to explain complex topics. It can also generate custom flashcards and study guides from quizzes or class materials, making it a more complete learning tool
Students in the U.S., Japan, Indonesia, Korea, and Brazil are being offered a complimentary one-year subscription to the AI Pro plan. This includes extended use of Gemini 2.5 Pro, NotebookLM, Veo 3, Deep Research, and other premium tools
This launch follows concerns that AI chatbots may weaken critical thinking by giving quick answers. Google and OpenAI present their tools as learning aids, aiming to reshape education in the AI era.
Top AI & SaaS Tools
Automaticall (Life-time Deal): Let AI answer missed calls, filter spam, send recaps and follow-ups to WhatsApp and email, and sync with your calendar
Higgs Audio: Generate expressive speech and audio without fine-tuning, trained on over 10 million hours of audio and text data [F-R-E-E]
Qwen-Image: Text-to-image model that generates accurate English and Chinese graphic posters and a variety of image styles [F-R-E-E]
Jules: Google’s autonomous coding agent has exited beta, integrating with code repositories to securely write tests, build features, fix bugs, and update dependencies [F-R-E-E]
Tasker: AI assistant that automates work and life tasks—like emails and meetings—by coordinating AI agents through simple language commands [F-R-E-E]
Seeking impartial news? Meet 1440.
Every day, 3.5 million readers turn to 1440 for their factual news. We sift through 100+ sources to bring you a complete summary of politics, global events, business, and culture, all in a brief 5-minute email. Enjoy an impartial news experience.
AI Tutorial
How to Turn YouTube Videos into Audio or Video Podcasts with NotebookLM

NotebookLM’s latest update introduces a powerful feature that lets you transform lengthy YouTube videos into AI-generated podcast-style summaries—either as audio or even a video podcast. It’s a quicker, more enjoyable way to absorb long-form content.
Open NotebookLM
Go to NotebookLM, sign in with your Google account, and launch a fresh notebook from the main dashboard.
Import a YouTube Video
In the bottom panel, look for the “Link” section and choose “YouTube” instead of “Website.”
Paste the URL of the video you want to summarize and select Insert to begin the import process.
Choose Format
On the right-hand side, you’ll see two options once the content is loaded:
Audio Overview – Generates a spoken summary, ideal for listening like a podcast
Video Overview – Creates an explainer-style video summary, narrated by AI (in English)
Processing time will vary depending on the video’s length, so give it a moment to complete.
Explore and Interact with the Content
Once ready, you can:
Listen to the AI-generated audio summary
Engage with the AI using the chat feature to ask questions or clarify details
Create content based on the summary—such as blog posts, notes, or social media clips
Try the Interactive Mode
If you select Interactive or Interactive Mode, the podcast will feature two AI-generated speakers discussing the topic—mimicking the natural flow of a real conversation, just like a traditional podcast episode.
Tip: For best results, choose videos with structured content—such as interviews, educational talks, or deep-dive documentaries. The clearer the material, the better the AI-generated podcast will be.
Top AI & Tech News
OpenAI released gpt-oss-120b and gpt-oss-20b, open-weight LLMs rivaling or surpassing o4-mini and o3-mini, available for local use
xAI’s Grok Imagine, a video and image generation tool, is currently being introduced in an early release on Android devices through Grok app
Anthropic added a /security-review command to Claude Code for automated terminal-based vulnerability checks
MidJourney offers HD videos with 4x more detail at about 3.2x the price, ideal for professional-quality footage
Resource of the Day
Fine-Tuning gpt-oss Models Using Hugging Face Transformers by OpenAI

This cookbook offers a step-by-step guide for fine-tuning OpenAI’s gpt-oss-20b model to reason in multiple languages with Hugging Face Transformers and the TRL library. It’s for AI practitioners and NLP engineers aiming to make large language models easier to use and more interpretable in under-represented languages.
Top AI Video Tutorial
The Ultimate AI SEO Playbook - Rank #1 in ChatGPT, Perplexity, Gemini & More
Complimentary AI Course of the Day
OpenAI Academy Courses

In this edition, instead of offering a single course, we are introducing the OpenAI Academy — a platform designed to equip individuals with the knowledge and skills needed to effectively harness artificial intelligence. The Academy features expert-led learning, community collaboration, and up-to-date information on AI products and trends.