November 2023

October was full with a couple of conferences (Obesity Week in Dallas and Concierge Medicine Forum in Atlanta) so I missed a month, but I’m back. Lots of interest in Calorify and doubly labeled water (DLW), lots of follow up work for me.

The amount of interest and questions being driven by the latest crop of GLP-1’s is pretty astounding. I suspect GLP-1’s will tie with AI as the biggest stories of the year in tech, if we expand “tech” to include pharma and medical technologies. The All In Podcast covered some analysts view of how these drugs are impacting the market and how it parallels AI. You can listen here and this picture highlights the story.

Andreesen Horowitz on the LLM stack - I’m doing some work on my own pet project in the AI agent space and came across this tech stack view of how AI based projects are getting built. It’s been helpful. This platform also came to my attention and I plan to give it a try. There are a bunch of leaks in my Twitter / X feed today about features from OpenAI that appear to promise many new ways to engage AI and chat solutions as well. You can’t sleep on this market it’s changing so fast. That said, as I dig into what’s really required to build useful applications, I’m coming to the conclusion that the main impact of LLM’s is in the UI layer of application building and not in the core intelligence / analytics layer. This is likely not a novel view and perhaps the real experts have know this all along. I’ll explain more next month when I’m not already a month behind on my newsletter.

On a related note I’m finally investing the time to learn Python (yes, my 17 yr old son occasionally gives me tips). First step has been getting a basic dev environment and using this library to parse FIT files and convert them to CSV, a common need if you spend a lot of time looking at fitness data from wearables or other devices and something Alan Couzens covered a while back.

I’ve really been enjoying early morning runs at St. Joseph’s in Los Gatos. Nothing like starting your run in the dark with a headlamp and slowly watching the world wake up.