Latest perspectives on product development and organizational transformation in the AI era
We launched a web app that visualizes AI exposure across Japan's labor market at the occupation level. Forked from Andrej Karpathy's US version, rebuilt with Japanese data. Explore salary, employment, and AI scores via an interactive treemap.
Should you switch a GPT-4o pipeline to a Reasoning model? A look at the real decision-making process and why we chose not to switch.
How to embrace AI-assisted software product development at the individual, team, and organizational level — a practical approach.
Why the preparation methods of top-performing salespeople fail to spread across the team, and how to fix it.
When an Excel planning sheet becomes a personalized 'secret sauce,' here's what to do first to start optimizing.
"Will this data be used to train AI?" — The answer is usually "no." But the real danger lies in logs: prompt contents quietly accumulating in cloud storage.
AI agents submit 5 PRs a day. Reviews can't keep up. Individually fast, but the team isn't moving. Let's unpack why.
"Go fully automated? But what if something goes wrong? Full manual review won't scale." A framework for deciding where humans belong in AI workflows.
As AI tools dramatically reduce the cost and time of software development, the bottleneck has shifted from "how to build" to knowing "what to build." Lessons from Anthropic's Claude Cowork release on agile development and empiricism in the AI age.
Organizations try to treat AI adoption as a linear plan. It isn't. It is a complex adaptive challenge. This workshop helps teams embrace uncertainty with short feedback loops, safe-to-fail experiments, and rapid learning.
AWS concepts like Accounts, Users, Roles, and Policies can be confusing. Let's explain them through an RPG analogy for instant clarity.
Many AI leads have only used ChatGPT. Four short videos covering the relationship between ChatGPT and AI, all the way to understanding API usage.
"AI," "Machine Learning," "LLM," "ChatGPT" — these terms are often used interchangeably. A 4-minute video to get everyone on the same page.
ChatGPT for scripts, Codex for automated demos, Claude for visuals, Gemini for voice. How multiple AIs came together to produce a product video.
A client wanted AI-powered auto-replies. But mapping the actual workflow revealed the real bottleneck was something else entirely.
RAG is hard to grasp without hands-on experience. An interactive demo with 30 FAQs and 200 inquiries visualized on a single page.
OpenAI's Codex is now officially available as a VSCode extension. Plus/Pro/Team users can use it at no extra cost. Here's how to set it up.
Mir
