Daily Reading List – April 3, 2025 (#526)

It’s time for a mini-vacation before next week’s festivities. See you back here on Monday!

[blog] No elephants: Breakthroughs in image generation. All of us have access now to tools that help us do creative imagery. Where does that lead us?

[article] Gartner forecasts gen AI spending to hit $644B in 2025: What it means for enterprise IT leaders. It’s interesting to see where the spending comes from. Most in hardware?

[article] What barriers prevent engineers from achieving “flow”? If you’re a tech leader or architect, you should be obsessed with ways to keep your developers in a flow state.

[article] AI training lags despite increased use at work, survey says. Lots of AI getting deployed, but not enough training of staff. Let’s fix that!

[blog] In defense of ruthless managers. Competent, frank-talking, and outcome-oriented managers are valuable, even if misunderstood.

[blog] The Strategy Behind MCP. Thought-provoking! Fintan looks at competing models of LLM as a layer, or LLMs at the core. And which approach Anthropic is betting on.

[blog] New in NotebookLM: Discover sources from around the web. Such a great idea! Now tap into Google Search smarts to find additional sources of insight for your research projects in NotebookLM.

[blog] Meta Open Source: 2024 by the numbers. There’s a lot going on at Meta, and they still seem to prioritize open source investments.

[blog] The Augmented Architect: Real-Time Enterprise Architecture In The Age Of AI. Here’s an extended writeup from folks at Forrester Research. Good read for architects!

[blog] GKE at 65,000 nodes: Evaluating performance for simulated mixed AI workloads. That’s a lot of nodes. Nobody can build a bigger cluster than us, and here’s some great details on the tooling, test cases, and metrics used for a benchmark.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below: