#AI

When I asked AI about a synthetic fuel technology, it kept giving evasive answers hiding behind structural explanations. A record of the dialogue process of changing my questions until I got concrete answers, and the question of 'honesty' that emerged.

As AI accelerates convergent thinking, time for divergent mode shrinks. I'm not afraid of stopping, yet I'm not choosing when to stop. A reflection through Guilford's theory and the Pirahã people.

AI training data is the work product of journalists, writers, and authors who invested real costs. Using it to build products is structurally identical to procurement. That's just how market economics works.

Exploring the discomfort I feel with the proliferation of AI books through a cooking analogy. When the implicit message is 'you don't need to think,' what do we lose?

Working with Claude Code and Agent Teams daily has left me with a persistent discomfort: when AI becomes highly sophisticated, can humans continue to handle 'correctness'? I'm putting my current thoughts into words as a hypothesis.

The future of AI rewrite competition is automated rewrites. Instead of joining the machine vs. machine war of attrition, I chose a different game: learning-driven rewrites.

Now that writing with AI has become commonplace, what constitutes 'valuable information'? I explore the difference between content that merely compiles information and truly conveying one's experiences and thoughts.

From the power consumption problem of AI companies to the weight of 8,000 years, and what it means to document our decision-making process.

AI increases productivity, yet leisure time doesn't grow. Through the Pirahã way of life, I reflect on what we've lost and consider alternative paths.