It’s Saturday. Let’s talk AI. An Oscar-winning actor just secured eight trademarks on himself, including video clips and audio, to combat unauthorized AI-generated voice clones.
Plus, self-driving cars that explain themselves, AI’s latest viral singer, and the PTCF prompting method (forwarded this email? Join 523k readers).
TOP STORY TODAY
McConaughey Trademarks Himself
Matthew McConaughey secured eight trademarks from the U.S. Patent and Trademark Office to protect his likeness and voice from unauthorized AI use. The trademarks include video clips of him and audio of his famous "Alright, alright, alright" line from Dazed and Confused.
The actor's legal team plans to use federal trademark law to combat AI-generated fakes, believing it provides stronger protection than existing state publicity rights. McConaughey stated he wants "a clear perimeter around ownership with consent and attribution."
Hollywood unions and companies support federal legislation banning AI-generated replicas without permission. A bill introduced in 2024 hasn't been voted on yet. McConaughey recently partnered with AI voice company ElevenLabs, where he's an investor.
Physical AI Arrives
Nvidia CEO Jensen Huang declared "the ChatGPT moment for physical AI is here" at CES 2026, unveiling Alpamayo, an AI reasoning model specifically designed for autonomous vehicles. The system thinks through difficult driving situations step-by-step, explaining its decisions like a human driver would.
Huang also introduced the Rubin computing platform, Nvidia's first extreme-codesigned six-chip AI system now in full production. The new architecture cuts token generation costs to one-tenth of previous platforms while reducing GPU requirements by 75 percent for training advanced models.
The announcements emphasize AI moving from software into the physical world through robotics, autonomous vehicles, and smart manufacturing. Nvidia is positioning its infrastructure as the "Android for generalist robots," providing the computational backbone for machines that understand and act in three-dimensional space rather than just processing text. Everything NVIDIA announced at CES 2026.
ChatGPT Health
OpenAI launched ChatGPT Health, allowing users to connect medical records, Apple Health, and wellness apps to the AI chatbot. The company revealed 230 million people globally ask health questions on ChatGPT weekly, with 40 million daily queries. Conversations are encrypted and stored separately from other chats.
The feature operates in a dedicated sandbox with enhanced privacy controls. OpenAI partnered with b.well for health data connectivity and worked with 260+ physicians across 60 countries during development. Medical records and app data won't be used to train AI models.
ChatGPT Health is not HIPAA-compliant and isn't intended for diagnosis or treatment. Early access launched to Free, Go, Plus, and Pro users outside Europe this week, with broader rollout planned for coming weeks. Data remains accessible via subpoena despite encryption (more).
TOGETHER WITH BUZZTOWN
Better AI Videos
Everyone’s using AI to make videos. The hard part isn’t making them. It’s knowing if they actually work.
You can generate something impressive in seconds, and still have no idea whether it connects or what to develop.
Buzztown is a $5 community built around the 1 Minute Movie: one short, finished moment made with AI, phones, animation (or whatever tools you have), not to chase perfection, but to test what actually lands.
No theory. No prompt hacks.
Just small, finished experiments that rebuild trust in your work. If you’re using AI to create, this is a low-stakes way to find out what’s working.
Join Buzztown for five bucks and see how your work ACTUALLY lands.
Thank you for supporting our sponsors!
TODAY’S LIFE ADVICE
The PTCF Framework for Better AI Results
Stop getting generic AI responses. Use the PTCF method to structure every prompt:
Persona: Tell the AI what role to play ("Act as an experienced data analyst")
Task: Specify the exact action ("Summarize this quarterly report")
Context: Add relevant background ("I'm presenting to executives who care about cost reduction")
Format: Define the output ("Give me 3 bullet points, each under 20 words")
Example: Instead of "Help me with marketing," try: "Act as a B2B marketing strategist. Create 3 LinkedIn post ideas for our SaaS product targeting startup founders. Each post should highlight a different pain point. Format as headlines with 2-sentence explanations."
The more specific your structure, the better the output. This framework works across ChatGPT, Claude, and Gemini.
Chain Your Prompts for Complex Work
Don't ask AI to "write a business plan" in one shot. Chain your prompts instead:
Step 1: "Create an outline for a business plan"
Step 2: "Expand section 3 (market analysis) with specific data points"
Step 3: "What weaknesses exist in this market analysis?"
Each prompt builds on the last. This breaks complex tasks into manageable steps and produces higher-quality outputs than single mega-prompts.
TRENDING
5 Stories
▲ Stranger Things fans accuse Duffer brothers of using ChatGPT after documentary shows blurred computer tabs
▲ Microsoft stock drops 10% in three months wiping out six-month gains as AI infrastructure costs balloon
▲ Bush Legend AI account with 180,000 followers impersonates Aboriginal presenter sparking "AI blackface" criticism from New Zealand
▲ Symbolic.ai signs News Corp deal to deploy AI journalism platform at Dow Jones Newswires financial hub
▲ Higgsfield AI reaches $1.3 billion valuation with $80 million extension hitting $200 million run rate in months
Fun Links
AI can help us talk to animals [Video]
Turkey’s un-manned fighter jet [Video]
Save on Kling AI Video this month [Gear*]
40 jobs most/least affected by AI [List]
Workflow for AI Legend of Zelda [Blog]
Fake singer with 2.8M followers [Blog]