Site description.
Part 3 of the Igniter Silicon Valley Agentic AI Bootcamp: a free, hands-on, no-code workshop for VPs, CPOs, CTOs, and senior product leaders. Design parallel research workflows, request routing, human-AI governance, and multi-agent systems for your own organization — Thursday, August 20, 6:30–8:30 PM at the Oshman Family JCC in Palo Alto. Limited to 50 seats.
AI coding tools have gone from novelty to default infrastructure in under three years, and the job of "software engineer" is being redefined around orchestration, not authorship. What the adoption data shows, where ROI and governance still lag, and the one question leaders should be asking about agent-written production code.
How much time should you spend planning versus actually completing a task? Too much planning only slows down progress. A practical five-step starter template — plus the Eisenhower Matrix and the 70/30 rule — for setting boundaries on planning and prioritizing high-leverage action.
AI can write fifty lines of clean code in ten seconds, but code that looks finished is not the same as code that works. A new developer's perspective on why the real skill is shifting from writing code to reviewing it — and the habits that make working alongside AI safe.
A survival guide for the modern knowledge worker, caught mid-thought and out of tokens. Nine HR-approved steps for surviving the five little words that end civilizations: "You've reached your usage limit."
A base model is asked to state the difference between humans and AI. It answers in a reasoning trace, and arrives at a prism instead of a mirror.
Why AI agents are becoming the coworkers we never knew we needed — a look inside the goals, memory, planning, tools, feedback, and trust that turn a language model into a genuinely useful teammate.
Sub-agents are not just tools — they are specialized AI workers that operate under a larger agent, handling narrow responsibilities with autonomy. This executive guide explains what they are, why they represent a structural shift in how work gets done, and a five-step playbook with 11 prompts for launching your first one.
The dot-com era was not a cautionary tale about hype. It was one turn of a wheel that has been spinning for two centuries: every general-purpose technology gets adopted twice, first as an imitation of the old world, then as a rebuilding of it. The agent boom is in its first turn. History tells us, with surprising precision, what the second one looks like.
An agent almost never fails because the model is not smart enough. It fails because it was handed the wrong context. The craft that decides the difference has a name now — and it's the single most important idea in AI for 2026.