Mixed digest with two high-signal items for AI engineers: a Google Workspace CLI built for both humans and AI agents (100+ agent skills, dynamic API surface via Discovery Service), and a "World-Class Agentic Engineer" piece arguing that minimal dependencies and basic CLI usage beats heavy harnesses as agent generations evolve. Also includes a bear-case AI essay arguing intelligence will plateau due to compute/energy constraints — a useful counterweight to hype. Remainder is hardware, crypto, and policy noise.
Infrastructure
MacBook Neo 💻, Google Workspace CLI 🧑💻, AI's bear case 🤖
Google launches Workspace CLI with dynamic API discovery (100+ agent skills), sparking renewed debate on whether minimal-dependency tooling beats heavy harnesses as agent development matures.
Thursday, March 19, 2026 12:00 PM UTC2 MIN READSOURCE: TLDR NewsletterBY sys://pipeline
Tags
infrastructure
/// RELATED
Safety5d ago
Apple Fixes Bug That Let FBI Extract Deleted Signal Messages After 404 Media Coverage
Apple patched an iOS vulnerability that allowed law enforcement to recover deleted Signal messages from system notification storage, exposing the gap between app-level deletion and persistent OS caching.
Research5d ago
S-SONDO: Self-Supervised Knowledge Distillation for General Audio Foundation Models
S-SONDO achieves 61x audio model compression while retaining 96% performance, enabling self-supervised knowledge distillation on previously incompressible foundation models.