Daily Reading List – December 11, 2024 (#458)

I’ll admit that today was super fun for me. Our big Gemini 2.0 set of announcements landed well and many people were taking this new technology for a spin. Much of today’s list relates to that, but I think you’ll enjoy it!

[blog] Introducing Gemini 2.0: our new AI model for the agentic era. Super post that has lots of details and links for those exploring the new world of Gemini 2.0.

[blog] The next chapter of the Gemini era for developers. This is a great post that explains all the interesting Gemini 2.0 features for developers including output modalities (audio, images), native tool use, the Live API, and even our new code agent.

[youtube-video] Behind the Scenes of Gemini 2.0. I liked this video between two of our product leaders bringing AI to everyone.

[blog] Announcing the general availability of Trillium, our sixth-generation TPU. One reason we can do so much, so quickly, and at such high quality is because of our TPU chips. You can use them too! Some news coverage here.

[article] Google Previews Jules AI Agent to Automate Software Engineering Tasks. Get some multi-step lans and processes from this experimental tool.

[article] Google’s Gemini Deep Research tool is here to answer your most complicated questions. Killer feature. This, along with things like NotebookLM, are going to change how we do research.

[blog] Gemini 2.0 Flash: An outstanding multi-modal LLM with a sci-fi streaming mode. I like our demos, but I enjoy your’s more. Simon tries out the new Live API and comes away impressed.

[site] Year in Search. What did you folks search for in 2024? This fun site shows you by country and by category.

[blog] How to Fine-tune PaliGemma 2. What does it take to fine-tune a model? And one that’s good at generating images? This is a terrific walkthrough.

[blog] Using fake deadlines without driving your engineers crazy. Good post. I’ve learned how important it is to set a clear milestone, and not settle for “it’s ready when it’s ready.”

[blog] How I reduced an API call from >5 seconds to under 100ms. Sentry is publishing some hot content lately. This is a fun read on tracing performance issues. They also published a nice piece on finding and fixing missing DB indexes.

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