Some folks may work for the pure joy of creating something. Kudos. There are things I do just for the fun of it, but I like to see results from my actual work. Specifically, I want to see people use what we build. Today’s Google earnings confirmed that folks are using more cloud. Nice!
[blog] A return to hand-written notes by learning to read & write. How’s your penmanship? Do you still write much? Not me. But this look at a model for converting images of handwriting into a digital format is impressive.
[article] Why Offsites Work — and How to Get the Most Out of Them. Different things happen when you get together in person versus online. You can’t change my mind. Here’s a good article about offsites and making them more effective.
[blog] Gemini models are coming to GitHub Copilot. GitHub is diversifying its model options, and we’re glad that developers can get access to the terrific Gemini models too. More from TechCrunch.
[article] Java Best Practices. I came across this post from earlier this year. It’s a good list of tips and tricks for doing your best work with Java.
[blog] Learn how to write AI prompts in 5 easy steps with this new course from Google. This isn’t targeted at coders; it’s for anyone looking learn how to prompt AI systems to summarize docs, analyze data, or even tweak the content of an email.
[article] How To Define an AI Agent Persona by Tweaking LLM Prompts. Speaking of prompting, getting an understanding of system prompts (or system instructions) is a valuable exercise.
[article] Enterprise AI moves from ‘experiment’ to ‘essential,’ spending jumps 130%. Folks are less scared, more excited. This latest research shows that AI adoption is slowing down.
[article] Citi moves vital infrastructure to Google Cloud as part of broader AI push. You hear about those moving workloads out of cloud, but why does someone actively choose cloud computing in 2024? This article touches on areas that matter, including access to innovative tech, provide a forcing function for modernization, improve the operating model, use scalable systems, and gain competitive advantages.
[blog] Semantically Compress Text to Save On LLM Costs. If you have a lot of data that is saying the same thing, you can bunch it up and simplify your LLM interactions.
[article] DORA 2024: AI and Platform Engineering Fall Short. We can’t explain why we saw certain things in the latest State of DevOps report, but it’s good to ask questions and dig deeper.
Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:
Richard Seroter's Architecture Musings 
Leave a comment