My Blogs

April 30, 2026

AGI Is a Rising Tide, Not a Moment

AGI was never going to arrive on a Tuesday. It is a waterline that keeps creeping up — already past customer support, climbing through engineering and law, lapping at the feet of researchers. By most honest yardsticks, it is already here.

February 9, 2026

OpenClaw and the Shape of a Wider Circle

OpenClaw’s real contribution is not raw capability but packaging that broadens adoption, plus config-as-code that turns agent memory into an asset you can own, evolve, and eventually trade.

February 8, 2026

Why I Skip ChatGPT Memory

I want my context to be portable and debuggable, not shaped by a platform I can’t see or control.

February 7, 2026

Context First: Lessons from Mario Zechner and the Pi Agent

Mario Zechner builds the Pi agent framework with ruthless minimalism, and his writing convinced me that context discipline and externalized state beat feature-heavy agent shells.

January 29, 2026

Skills Are Runtime, Not Code

Skill-based development feels fast because it bakes in recovery. Instead of perfecting a brittle workflow, you define intent, add a few scripts, and let an agent adapt.

January 28, 2026

The Human Loop Behind the Wall

When “A” focuses on Great Wall tasks—automation, defenses, and autonomy—I want to pull it back to the human loop: the respectful pushback and clarification that make interactions actually good.

January 27, 2026

Chatbot Ads: Incremental Budgets First, Demand Later

Why chatbot ads likely add incremental budget in the short term and only expand the ad pie long-term if they create real new demand.

January 26, 2026

Optmize for pass@k

Early agents should chase the single wow moment (pass@k); mature products must optimize repeatable success (pass^k).

January 23, 2026

Design as Intermediate Representation

If design is the intermediate representation between intent and software, AI will make the artifacts thinner. The real value migrates to definition, constraints, and evaluation.

January 22, 2026

Why I Write Prompts in English

English is the control layer I use for prompting: faster input, fewer UI collisions, and clearer instructions for the model—even when the output isn’t English.