Some reading from this past week.
Energy
AI
In Brazil’s Amazon, AI is making healthcare safer – Rest of World
Ethan Mollick talk to Nicolai Tangen
Some reading from this past week.
In Brazil’s Amazon, AI is making healthcare safer – Rest of World
Ethan Mollick talk to Nicolai Tangen
Colour me surprised.
Meta and Yandex are de-anonymizing Android users’ web browsing identifiers
Some insightful pieces below.
Why I have slightly longer timelines than some of my guests
Tips on prompting ChatGPT for UK technology secretary Peter Kyle
It took me too long to understand this but YouTube is one of most important distribution channels today.
Why the best journalists on YouTube are all former Vox employees
Anyone involved in ecommerce in any way should read this. As we move to agentic ecommerce, who controls the interface between you and your customers and what does this mean for the future?
Remarks on AI from NZ – by Neal Stephenson – Graphomane
The first-ever UX Study of Google’s AI Overviews: The Data We’ve All Been Waiting For
Tesla (TSLA) Chinese deliveries dropped to a scary low level | Electrek
The World’s Largest Search Engine Doesn’t Want You to Search
It’s that dreaded time of year again when our health insurance is up for renewal. What do we need? What are the key differences between the hundreds of available plans. Is our current plan still the most suitable? This year, on a whim I decided to try something different.
It’s a really difficult and complex topic to research as the plans are not particularly clear, there are too many of them, they are too complex (excesses, procedures not covered, differences between hospital types etc. ) and the plan that best suits you will depend on your own particular circumstances at a moment in time.
Today, as an experiment, I gave Gemini Advanced Deep Research with 2.5 Pro a custom prompt, gave it some general information about our ages, health status, specific issues that may be important to us and our current plan. I then asked it to evaluate this against other plans from the providers in the market.
Less than 5 minutes later it produced a pretty detailed report (that would have taken me hours to do) comparing our current plan to others in the market, the advantages and disadvantages of each and points to consider when making a choice.
You always have to be extremely careful when using an Ai model for research like this and double check everything, but I was seriously impressed by the results. It was a very good first pass that at the very least will save some time.
I’m sure that this could be improved if a private (perhaps State regulated) model had deeper access to our health data, but I feel like I have seen a glimpse of the future. Try it yourself. You may be surprised by the results.
I’ve been trying out the various models for some time now. I swapped from a ChatGPT subscription to Claude. In the last week, Gemini Advanced 2.5 Pro strikes me as being very very good and now thinking about switching my sub to Gemini for a while. An interesting experiment is to ask both the same question, then provide the answers of each to the other and ask each to evaluate which is the better answer and why.
I started reading Weimar Germany, Promise and Tragedy, by Eric D. Weitz. If I had to summarise what I’ve read so far in one word, that word would be chaos.
Seem familiar?
Finished reading Working by Robert Caro. Once I started I couldn’t put it down. Fascinating read. Recommended.
It’s a crazy world we live in.