Although I hadn’t updated them much recently, I have deleted my accounts on X (Twitter), Zenn, note, and Bluesky. Moving forward, text-based updates will be exclusively on this personal blog. My main activities will shift to YouTube (for marketing) and platforms like PlayStore and Roblox (for product releases). Why leave the “town square” that is so important for marketing? The reason lies in the paradigm shift happening in software development and the limitations of human cognitive structure I’ve felt in response to it.
1. The Titans’ Pivot and Paradigm Shift
Recently, iconic hackers like Linus Torvalds (creator of Linux), Antirez (creator of Redis), and DHH (creator of Rails) have successively affirmed development using tools like Claude Code or Google’s Antigravity—so-called “Vibe Coding”. When veterans with over 30 years of experience, who know the depths of computer science, conclude that “it is better to use AI than to write by hand,” it is not just an evolution of tools; it is a historical branching point in development methodology.
2. Humans are Ruled by “System Prompts”
However, there is a barrier before logic to accepting this change. Humans decide conclusions based on “emotion” before logic. This is very similar to the “System Prompt” (pre-instruction) in LLMs (Large Language Models). If an engineer’s system prompt is etched with definitions like “I don’t like some unknown person suddenly achieving results with Vibe Coding” or “Knowledge acquired through years of hard work is what is truly noble,” then no matter how many logical facts (User Prompts) like “productivity is higher” or “bugs are reduced” you input, the output answer will not change. Logic is mobilized retroactively to justify the conclusion already decided by emotion. This is not about good or bad; it is the specification of human beings.
3. Structural Flaws of Platforms
The algorithms of X and technical sharing services hack this nature. By alternately displaying “content positive about AI” and “content negative about AI,” they bring people with different system prompts into contact, generating revenue through emotional friction (engagement). The more Linux developers say “AI is good,” the stronger the rejection reaction from the system prompts of those whose existing authority or skill sets are threatened. I have decided to stop because it is exhausting to spend my resources (MP) on such structural conflict incitement and endless debates.
4. Future Stance
I believe the coming era will be a red ocean where only the result of “what you created” using AI will be questioned. I don’t think there is a need to transmit opinions in short text at the risk of rubbing someone’s emotions the wrong way. I will place my products in the code, games, and videos that will be released from now on.