Today’s reads are fairly Google-centric. On Fridays, I prepare 1-2 different newsletters, so I’m browsing through lots of that sort of content. Adjust your expectations accordingly!
[blog] Gemini image generation got it wrong. We’ll do better. I was embarrassed by these results, and I’m glad other folks at Google were too. Not ok. Accuracy is paramount to earning trust.
[blog] go run. This simple command lets Go developers run their app without any fuss. Chris writes a post celebrating its sneaky power.
[article] 3 Ways to Help Struggling Open Source Communities. It’s not someone else’s job to keep the open source we depend on healthy. It’s all of us, right? This article looks at how to keep showing support to your favorite community.
[blog] Orchestrate Vertex AI’s PaLM and Gemini APIs with Workflows. Chatbots won’t be the only way we interact with AI models. I can imagine a lot of back-office processes that can take advantage of AI in their workflows.
[blog] Cut Container Startup Time for Better Performance and Costs — Part2. Here’s a good deep dive into the various configuration and cluster changes you’d make to improve container startup time for a workload.
[blog] Summarizing meeting notes and amplifying visibility (prompt engineering). Can AI make meeting notes better? Probably. Tom has example prompts and workflows you can use too.
[blog] Test-Driven Development with Java, Spring Boot and Duet AI in GCP. This example uses one particular AI-assisted dev tool, but you can follow these terrific instructions in most other tools too.
[blog] What I like about Google’s Managed Prometheus. Some thing are inherently complex. Good products can still make it feel relatively simple to use. This is an example.
##
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 
One thought