https://brianarfi.com/en/blog/ Writing by Brian Arfi Faridhi (Brobri) on AI, careers, and building products. en Mon, 13 Jul 2026 00:00:00 +0700 https://brianarfi.com/en/blog/use-ai-start-with-your-job/ https://brianarfi.com/en/blog/use-ai-start-with-your-job/ Mon, 13 Jul 2026 00:00:00 +0700 Stop asking which AI tool is the best. Map your daily tasks, find the repetitive ones, and let AI handle them. Here is how I do it. https://brianarfi.com/en/blog/one-person-eight-channels/ https://brianarfi.com/en/blog/one-person-eight-channels/ Sun, 12 Jul 2026 00:00:00 +0700 A year ago this needed an engineering team. Now one video feeds eight channels on autopilot, built by one product guy and AI. Here is what actually changed. https://brianarfi.com/en/blog/ai-native-workforce-guide-indonesia/ https://brianarfi.com/en/blog/ai-native-workforce-guide-indonesia/ Sun, 12 Jul 2026 00:00:00 +0700 Most corporate AI training fails. A practical framework from 20 years of building products: the AI-Native Ladder, how to run a pilot, and what to measure. https://brianarfi.com/en/blog/4m-saved-before-ai/ https://brianarfi.com/en/blog/4m-saved-before-ai/ Sun, 12 Jul 2026 00:00:00 +0700 The $4M+ on my homepage is yearly savings, achieved before the AI era. The principle has not changed since. Who gets to apply it has. https://brianarfi.com/en/blog/one-thing-that-matters-today/ https://brianarfi.com/en/blog/one-thing-that-matters-today/ Wed, 08 Jul 2026 00:00:00 +0700 Four teams, four roadmaps, everything feels urgent. Focus means turning down good work for the one thing that matters most today. https://brianarfi.com/en/blog/approvals-that-are-easy-to-edit/ https://brianarfi.com/en/blog/approvals-that-are-easy-to-edit/ Wed, 08 Jul 2026 00:00:00 +0700 I made a decision doc impossible to edit, on purpose. Once changing it became expensive, the approval finally meant something. https://brianarfi.com/en/blog/ai-roadmap-memory-test/ https://brianarfi.com/en/blog/ai-roadmap-memory-test/ Wed, 08 Jul 2026 00:00:00 +0700 I asked AI to draft a semester roadmap from memory. It looked finished. Then I checked it against the original vision doc and found the holes.