Second

Two months after coming back from Singapore, in January 2026, I registered a sole proprietorship and launched Second, an online programming school for elementary and junior-high students on remote islands like the one I grew up on. The name was meant to signal a second chance for learners; I had started programming late myself.

The classroom got attention. In March the Ehime Keizai Report and the Ehime Shimbun both covered it, the latter describing it as the first programming classroom in Kamijima. In April I pivoted Second toward youth-talent development for the SusHi Tech Teen Challenge 2026, the Tokyo Metropolitan Government’s startup program for teenagers: island children would bring local environmental problems to hackathons, KOSEN students would build proofs of concept, and municipalities and companies would be brought in as partners. With two fellow Yuge College students I pitched that version and reached the semifinal.

What went wrong

It is easy to list now. The islands were a blue ocean, but a blue ocean is not the same as demand. Some of the children I wanted to teach may not have owned a computer at all. I had almost no connections to parents, and the one thing that would have fixed that, going out and selling, I never worked up the courage to do. Then spring break ended, my time disappeared, and the anxiety set in. In May I closed the business. It never earned a yen.

Closing it was, at the time, a relief. That same month I told an interviewer that the experience had convinced me research suited me better than business. In June I was on stage pitching Beacon.

What it left behind

Second did not fail on technology; it failed because I never knocked on a door. Contests taught me to pitch, and to tell a slide deck from a pitch deck: explaining something versus making someone want it. Only customers can teach me to sell. The venture also burned off most of my fear of starting things.

It also had nothing to do with my research, which in hindsight is the point. The two ventures that came after, Beacon and Kyudo AI, sit on top of a year of work that already existed. In September 2026 I wrote Second up as a failure case in a preprint on entrepreneurship education for KOSEN students.