🔥 OpenAI Launches GPT-5.1

PLUS: LeCun to Leave Meta

Author

The Summary AI
November 14, 2025

Welcome back!

OpenAI has rolled out GPT-5.1, a model update focused on adaptability and personalization. It adds adjustable personas for tone and style, adaptive reasoning, and faster response modes. But it brings back the question: do users want a chat companion or a precise assistant? Let’s unpack…

Today’s Summary:

  • 🔥 GPT-5.1 adds custom personas and adaptive thinking

  • 🕹️ World Labs unveils Marble for 3D AI worlds

  • 🧠 Yann LeCun to exit Meta to build world-model startup

  • đź“° NYT seeks 20M ChatGPT chats in lawsuit push

  • 📸 Six new AI tricks in Google Photos

  • đź““ Google NotebookLM adds Deep Research

  • 🛠️ 2 new tools

TOP STORY

GPT-5.1 brings custom personalities and adaptive thinking

The Summary: OpenAI rolled out GPT-5.1 with two models, Instant and Thinking, built to answer faster and follow instructions more accurately. The update also adds eight preset personas and new sliders for tone, brevity, and style control.

Key details:

  • GPT-5.1 Instant aims for warmer talk by default, after earlier pushback about “too dry” GPT-5 replies compared to GPT-4o

  • Eight personas presets: Professional, Friendly, Efficient, Candid, Quirky, Cynical, Nerdy, plus Default

  • Adaptive thinking adjusts its compute for easy or hard tasks

  • Personalization sliders can adjust conciseness, warmth, and emoji use, with ChatGPT offering mid-chat prompts to update settings

Why it matters: There is a split between users who want a dependable work instrument and those who want a chat companion. The new controls try to bridge that gap, making the model more adaptable. Adaptive reasoning may also reflect optimizations to reduce compute costs.

FROM OUR PARTNERS

Create AI Agents That Work for You

The Simplest Way to Create and Launch AI Agents and Apps

You know that AI can help you automate your work, but you just don't know how to get started.

With Lindy, you can build AI agents and apps in minutes simply by describing what you want in plain English.

From inbound lead qualification to AI-powered customer support and full-blown apps, Lindy has hundreds of agents that are ready to work for you 24/7/365.

Stop doing repetitive tasks manually. Let Lindy automate workflows, save time, and grow your business.

3D AI

World Labs launches Marble for AI built 3D worlds

The Summary: World Labs has released Marble, a commercial world model that turns text, images, videos, and rough 3D layouts into persistent, editable 3D environments. The company is led by Fei-Fei Li, the Stanford professor known for creating ImageNet, the dataset that taught computers to see.

Key details:

  • Supports image prompts, video clips, panoramas, or 3D blockouts, and turns them into navigable scenes with free camera control

  • Chisel lets users sketch spatial layouts in 3D and then apply style

  • Produces worlds that can be edited, expanded, and exported as Gaussian splats, meshes or recorded videos

  • Four subscription tiers available including Free tier (4 generations)

Why it matters: Marble gives creators a shortcut to explorable 3D spaces for rapid idea testing. Robotics labs also gain a way to generate synthetic training environments at scale. Models like Marble might evolve into a more general interface for digital interaction with physical content.

FROM OUR PARTNERS

The State of Voice AI

Voice AI: Get the Proof. Avoid the Hype.

Deepgram interviewed 400 senior leaders on voice AI adoption: 97% already use it, 84% will increase budgets, yet only 21% are very satisfied with legacy agents. See where enterprises deploy human-like voice AI agents - customer service, task automation, order capture. Benchmark your roadmap against $100M peers for 2026 priorities.

META

Yann LeCun to leave Meta to build world-model startup

The Summary: Meta’s chief AI scientist Yann LeCun plans to leave the company to launch his own startup focused on “world models”, a different approach to machine intelligence that learns from video and spatial data instead of text. The move comes as Meta doubles down on LLM initiatives under new leadership. LeCun, who pioneered modern deep learning, aims to build AI that truly reasons like humans.

Key details:

  • LeCun has been Meta’s chief AI scientist since 2013 and co-won the 2018 Turing Award (the “Nobel Prize of computer science”)

  • He founded Meta’s FAIR lab, which produced the original Llama

  • His research centers on “world models” that learn through sensory data, with his new paper outlining a simpler way for AI to learn

  • Meta recently paid $14.3B for a 49% stake in Scale AI, appointing its 28-year-old CEO Alexandr Wang to lead its Superintelligence Labs

Why it matters: LeCun’s planned exit shows a philosophical split inside Meta about what artificial intelligence really means. While Zuckerberg bets billions on scaling LLMs, LeCun argues that they will not be the path forward for the next decade of AI.

TOOLS

🥇 New tools

  • Stream Ring - Finger-worn AI assistant for spoken input

  • Talo - Realtime AI translator for video calls

That’s all for today!

If you liked the newsletter, share it with your friends and colleagues by sending them this link: https://thesummary.ai/