~/talk

Developer keynote deep dive

on YouTube 02h 58m

Talkat Google Cloud Next '26 · 2026-04-22..04-24· [ai-assisted-development, large-language-models]

speakers

name:Stephanie Wong

links:none on file

name:Michele Catasta

website:https://pirroh.fyi

name:Sarah Kennedy

links:none on file

name:Ricky Robinett

bluesky:rickyrobinett.bsky.social

name:Karl Weinmeister

github:kweinmeister

name:Jeff Whelpley

website:https://g.dev/jeff

name:Ankur Kotwal

links:none on file

name:Salman Ladha

links:none on file

name:Harrison Chase

links:none on file

name:Jason Davenport

links:none on file

name:Kevin Moore

website:https://kevmoo.com

name:Rody Davis

website:https://rodydavis.com

name:Ines Envid

links:none on file

resources

video:https://youtu.be/JemyjTlOvy0

slides:none on file

summary

Live coding shows which announcements actually work: hands-on demos of Gemini Enterprise and the new Agent Engine, Replit’s agent already builds >50% of its internal tools and publishes apps on Cloud Run that scale from day‑one, Wiz’s red agent auto‑discovers an exploit that exposes API keys and PII and the green agent writes a GitHub PR fix that a coding agent completes, LangChain/DeepAgents proves harness tuning moves TerminalBench ranks from ~30th to 5th, and Dart functions + Firebase produce tiny native or Wasm binaries and full deploys in ~20–30s — good pick if you want runnable demos, infra, and concrete remediation flows with red/green/blue agents.

AI-written summary · about

chapters (8)

related talks