OpenAI just launched ChatGPT Health, a dedicated space within the app that connects your medical records and wellness apps. It's an attempt to turn the chatbot into a centralized, personal health analyst.
This gives ChatGPT some extra spice, adding on to their general-purpose tool a specialized personal assistant for a critical area of life.
Topics of the day:
ChatGPT as your personal health analyst
Utah greenlights AI to renew prescriptions
An AI copilot running a particle accelerator
A look back at how AI changed work in 2025
The Shortlist: xAI raises a massive $20B Series E, Lenovo launches a system-level AI assistant across PCs and phones, Universal Music partners with Nvidia on AI music tools, and researchers uncover a new “ZombieAgent” prompt injection attack targeting ChatGPT.
OpenAI unveils ChatGPT Health
What’s happening: OpenAI launched ChatGPT Health, a dedicated and private space within the app that connects with your medical records and wellness apps for personalized guidance.
I've been waiting for something like this to centralize all my scattered health data, which I have been chatting with ChatGPT about already.
In practice:
You can connect data from apps like Apple Health and MyFitnessPal to turn ChatGPT into a personal health dashboard that helps you spot trends in your sleep, activity, and nutrition.
Before your next check-up, upload recent lab results and ask it to summarize the key takeaways and generate questions, helping you get more out of your time with the doctor.
It helps automate tedious health admin by analyzing different plans and navigating insurance options based on your actual medical history and needs.
Bottom line: This turns ChatGPT into a personal health analyst, helping you connect the dots between all your data.
Utah greenlights AI to renew prescriptions
What’s happening: Utah just became the first U.S. state to let an AI system autonomously renew certain medical prescriptions. The pilot program is a major test case for using AI in highly regulated industries.
In practice:
This provides a blueprint for automating credentialed work, freeing up doctors from routine administrative tasks to focus on complex patient care.
The startup behind it is proving out a new, low-cost model that could dramatically save patients money and improve access, especially in underserved areas.
With a 99% accuracy rate compared to human doctors and other states watching closely, this is a scalable model for other compliance-heavy tasks.
Bottom line: Regulators are officially opening the door for AI to move from being a copilot to an autonomous decision maker. This creates a massive opportunity to automate core operational tasks previously reserved only for human experts.
AI Copilot Steers Berkeley's Particle Accelerator
What’s happening: An AI assistant is now helping run Berkeley's particle accelerator, a system with 230,000 variables where downtime is costly. The copilot autonomously sets up experiments and troubleshoots issues, reducing human effort by 100x according to a new research paper.
In practice:
This is a blueprint for building an internal "expert system" that consolidates scattered documentation, turning tribal knowledge into an interactive problem-solver.
For technical teams, an AI like this can translate a plain-English goal into executable code, automating complex operational tasks that once required manual scripting.
By drastically cutting setup time and minimizing downtime, this model allows you to increase the operational tempo of any complex system, from manufacturing to software deployment.
Bottom line: AI is moving from answering questions to running critical physical systems. This is another take on how specialized copilots can become standard for operating any complex business infrastructure.
A look back: How AI changed work in 2025
What’s happening: Looking back, 2025 was the year AI graduated from a fun chatbot to a reliable work platform. The big theme wasn't one viral launch, but a quiet shift toward tools you could trust to handle real, multi-step tasks without constant babysitting.
In practice:
Assistants became platforms, letting you hand off entire projects with memory and context instead of just firing off one-off prompts.
Creative tools for video and audio got consistent, making it practical to produce on-brand assets for marketing or training without endless rerolls.
The model market got competitive, giving you more choices for specialized tasks and powerful open-source options for more control.
Bottom line: The conversation is no longer if AI can do something, but how reliably it can be integrated into a real workflow. This shift from impressive demos to dependable tools is what unlocks true automation.
The Shortlist
xAI announced a $20 billion Series E funding round, reinforcing the intense capital race among top AI labs.
Lenovo introduced Qira, a system-level AI assistant for its PCs and Motorola phones designed to provide context and continuity between devices.
Universal Music Group teamed up with Nvidia to use its Music Flamingo model on its catalog and co-develop new AI tools with artists.
Radware detailed ZombieAgent, a new prompt injection attack that bypasses ChatGPT's recent security fixes, showing how difficult it is to secure LLMs from data exfiltration.
This newsletter is where I (Kwadwo) share products, articles, and links that I find useful and interesting, mostly around AI. I focus on tools and solutions that bring real value to people in everyday jobs, not just tech insiders.
Please share any feedback you have either in an answer or through the poll below 🙏🏽
