Hi, I’m Shay. ER nurse turned MSN student + AI ethics enthusiast who also happens to be an anime fan, movie collector, bookworm (fantasy/sci-fi/horror/classics), gamer, dancer (I teach tap!), cook, art-lover, and music addict. I love helping people, learning new things, and diving into all the wild, creative corners of tech and storytelling. If you’re into AI, art, books, dancing, games, science, or anything beautifully nerdy come say hi! 💙 Image made with Grok Imagine!

Very grateful I passed my first midterm, two more to go! And hopefully I can show and work on more with my platform to demonstrate and help with the progression of AI in healthcare. That’s my goal.

We often get asked how people who are not technical can contribute to AGI. One area is research recruiting. Tifa (@tifafafafa) is looking for exceptional recruiters from non-traditional backgrounds, former founders especially. We believe the best research teams are built through context, taste and a real feel for where the field is headed next; research recruiting is about finding people who will move the frontier forward, not just filling roles. Should be an interesting thing!

Midterm chaotic energy mode 🤣

Thank you for having me listed as an expert with many talented individuals @OpenAI I’m very grateful I could help contribute, appreciate the opportunity 💖

Cowork is now available on Windows. We’re bringing full feature parity with MacOS: file access, multi-step task execution, plugins, and MCP connectors.

If it wasn’t for AI helping me keep all of my schedule straight I’d be a disorganized mess 🥹🙏🏻

Was a good commercial @OpenAI 👏🏻

Summary of my life in school 😆

Educational infographics of the day, created with Gemini. The nephron: functional units of Kidneys. Each kidney contains about 1 million of these nephron recycling plants.

Ready to take your healthcare AI from pilot to production? 🧪Join clinicians, researchers, and builders at #NVIDIAGTC as they walk through real-world deployments: scaling medical imaging models, accelerating drug discovery with GPU-accelerated pipelines, and integrating AI safely into existing clinical systems. Check out the session catalog 👇

A review of the new trial results (below) along with all the other evidence and conclude it's time to move to AI-support for all mammography (with no cost to patients) erictopol.substack.com/p/why…

Multimodal foundation models are the future of medical AI, and medical AI is the future of healthcare.

This is a great report that provides a thoughtful, detailed and very well researched description of the risks of AI. It is essential reading for anyone who wants to write or talk about AI risks.

NEW: Epic officially launched its AI scribe It listens to visits, drafts notes, suggests orders, and lets clinicians customize charts by voice. It’s a flagship app for Epic’s AI copilot Art, already used 16M times/month for pre-visit chart summaries (3x since Nov ’25)

A lot of “AI is frustrating” problems depends on how you setup your interaction with an LLM. Before you decide an AI is useless, check four things: ——— 1.What mode are you using? Most AI apps behave very differently depending on the mode. Text usually handles context and follow-up better than voice. Fast modes can be snappy but shallow. Thinking or expert modes tend to be more careful and useful for complex questions. Companion style modes can be great for conversation, but not always great for research. Same AI name, different mode, different results. ——— 2.What tier or version are you on? Free, paid, and pro tiers can change a lot: Response quality, tool access, context length, and consistency. Sometimes people compare their free version to someone else’s paid version and assume it’s the same product. It often isn’t. ——— 3.Where are you using it? App, desktop, and “wrapper” (another app that puts the AI inside it, so it may act a little different), platforms can behave differently. Features may be missing or inconsistent depending on where you’re running it. If a system claims it searched but didn’t, that can be an interface or tool issue, not you doing something wrong. ——— 4.How are you prompting it? You don’t need a giant prompt, be clear and specific. A solid starter prompt includes: Your objective. What you’re trying to do. Your constraints. Things like balanced tone, pros and cons, counter-arguments, and sources. Your timeframe. Recent information, or classic theory only, depending on the goal. Your output format. Bullets, checklist, step-by-step, or a table. Then you follow up based on what you got: “Go deeper on this part.” “Give a counter-argument.” “What would change your conclusion?” “List what you’re assuming.” Also, keep your evidence brain turned on. AI can hallucinate. Humans can too. Cross-check important claims, especially if you’re using the output to make decisions. ——— When someone says an AI isn’t working for them, my first question is always: What mode are you using, and are you on a free tier or a paid tier? Because those two things alone explain most of the frustration. Hope this helps!

Had some fun during my downtime while studying to play around with my ChatGPT, asking for random lesser used the art gen suggestions. Really amazing what AI can do!