D-Squared Musings Week 10

I stumbled across this idea of “Zero Touch Prod” a few times in the past, conveyed in different ways, but this actual implementation from Google helps clarify a good amount. If we’re able to remove all human interactions from production, we’re able to drastically reduce the number of incidents. Their approach is through heavy automation and a specific proxy pattern for admins. I’m starting to see more chatter around “everything-as-code” – Policy, infrastructure, etc., where all human interaction sits at the beginning of a Git pipeline, ensuring all changes are checked before deployment.

This is the second piece I’ve seen arguing for the transition back to on-premise infrastructure, basecamp was the first. I’m not convinced this trend will grow drastically, but it is interesting to see how Kubernetes and containers are allowing larger companies to port cloud-native practices to on-prem infrastructure.

I know next to nothing when it comes to Rust, but I’m sure that’ll change over time, as I dive deeper into the topic. This presentation by one of the creators of Rust gives good insight into the ethos behind Rust, the challenges their facing, and how the vision for Rust into 2024. The main takeaway currently – Rust is not meant for quick experimentation like Python, but instead, it’s intended for robust, reliable, and efficient applications.

The most interesting part of the conversation for me was around 21:03. At this point Chamath mentions that companies with unique datasets will likely be the next adjacent acquisitions from larger tech companies due to the importance of feeding their models for special use cases. That’s an interesting take I’ve not given much thought to before.