✨ Welcome back, Entrepreneurs

OpenAI hit “code red,” Google dropped a new reasoning engine, Anthropic pushed deeper into how humans use AI, and the NYT took Perplexity to court. If you’re building anything in this space, the ground just shifted again.

Here’s everything that actually matters - fast.

BIG AI MOVES

🔹 Anthropic Interviewer Launches to Map How Workers Use AI

Anthropic introduced Interviewer, an AI-driven research tool that ran 1,250 interviews with professionals, creatives, and scientists to track how they use AI, surface optimism and anxieties about automation, and feed anonymized insights back into Claude’s design and societal-impact research.

🔹 NotebookLM Boosts Chat With 10,000-Character Personas

NotebookLM expanded its chat customization limit from 500 to 10,000 characters, letting users define rich “roles” like product managers, teachers, or researchers so the model can answer in tightly structured, task-specific formats directly grounded in their source documents.

🔹 NYT Sues Perplexity for Copying and Republishing Its Articles

The New York Times filed a federal lawsuit accusing Perplexity of scraping, reproducing, and monetizing near-verbatim copies of its reporting, alleging the AI startup bypassed paywalls and robots.txt protections while generating summaries and outputs that substitute for the newspaper’s original work.

⚡ OpenAI Preps GPT-5.2 After “Code Red” Push From Sam Altman

OpenAI is accelerating the release of GPT-5.2, moving it up to early next week after Sam Altman declared a “code red” in response to Google’s Gemini 3 surge and Anthropic’s rapid gains.

Highlights

• GPT-5.2 is reportedly ready to ship, targeting a December 9 release

• Designed to close the gap created by Gemini 3’s leaderboard-topping performance

• Internal tests suggest the next reasoning model is ahead of Gemini 3

• Launch was originally planned for later in December but was pulled forward under competitive pressure

Why it matters:
GPT-5.2 marks OpenAI’s first major counter to Google’s momentum. With reasoning improvements and faster iteration cycles, OpenAI aims to reassert leadership in the model race and kick off a new phase of ChatGPT upgrades shaped by this internal “code red.”

⚡ Google Releases Gemini 3 Deep Think: A Big Leap in Reasoning

Google has rolled out Gemini 3 Deep Think to AI Ultra subscribers inside the Gemini app, delivering a major upgrade in long-form reasoning for math, science, and complex logic tasks.

Highlights

• New Deep Think mode uses parallel reasoning to test multiple hypotheses at once

• Scores 41.0% on Humanity’s Last Exam (no tools) — best in class

• Hits 45.1% on ARC-AGI-2 with code execution — an industry first

• Builds on Gemini 2.5 Deep Think, which earned gold-medal results at IMO and ICPC

• Available now: choose Deep Think in the prompt bar with Gemini 3 Pro selected

Why it matters:
Gemini 3 Deep Think excels in human-like problem solving by combining logic, hypothesis exploration, and code execution, making it ideal for advanced STEM tasks, competitive exams, and systems requiring rigorous thinking.

🧠🫁💩AI just reinvented parts of your body

Lung-cancer LLMs hit 98% accuracy. Cardiac MRI gets supercharged. Toilets become health scanners. A device teaches the brain to “feel” smells. And researchers found anti-aging molecules hiding in your blood.
This week’s AIHealthTech Insider feels like sci-fi.

See everything inside →  AIHealthTech Insider: Issue #77 

🏖️ Create Tiny Worlds with Midjourney’s Omni Reference

Midjourney V7’s Omni Reference lets you turn simple fruit photos into hyper-realistic tropical scenes - mini beaches for eye-catching social posts.

How It Works:

• Upload a clear fruit photo and lock texture with –sref + –ow 50

• Prompt ideas: “tiny tropical beach carved inside a pineapple half, ultra-detailed, macro style”

• Add micro elements: palm trees, umbrellas, swimmers, boats, wet-sand reflections

• Shape the vibe with lighting cues like golden hour or dreamy depth of field

• For motion pieces, include sway, ripple, glide for smooth loops

Pro Tip: Start with a clean, well-lit fruit photo as your –sref.

🚀 Sunday AI: The Upgrades That Change Everything

MIT’s new “Iceberg Index” shows 11.7% of jobs are now technically replaceable. The shift is bigger than anyone realizes.

👉 Catch the full breakdown, the career moves, and the tools you should be using right now inside this issue.