今日は妻にAntigravity 導入ガイド。横で色々見てた娘が「AIはアタリの絵でも違う角度出せる？」と描いた絵を持ってきた。 全然違う仕上がった状態になるか…

今日は妻にAntigravity 導入ガイド。横で色々見てた娘が「AIはアタリの絵でも違う角度出せる？」と描いた絵を持ってきた。
全然違う仕上がった状態になるか、全然だめかのどっちかかも…と伝えてNano Banana Proで試したらこれ。

確かに目は描かれてしまったけど、ツインテール、フード、袖の形にスカートのプリーツにポーズまで、元の絵にあった娘の意図がちゃんと拾われてて、すげーすげー騒いでた。零れたのは胸元の紐がネクタイと判断されたとこぐらい。

ここまでラフでも回せる事もそりゃ凄いけど、拙い線としか見ていなかった私の認識の浅さと、きちんと伝わる表現をしていた娘の方に驚いた。

形は違えど、AIがこういう気付き与えることは無数にあるはずで、親や教育者の不明超える効果はこれから凄まじく広がってくと実感した。

Today I gave my wife a guide on how to use Antigravity. My daughter, who was watching various things beside me, came with a drawing she made, asking 'Can AI draw the same picture from a different angle?'. I told her it might turn out completely different or be a complete failure... but when I tried it with Nano Banana Pro, this came out. Sure enough, eyes were drawn, but from the twin tails, hood, sleeve shape, skirt pleats, to the pose, my daughter's intention from the original picture was properly captured, and I was incredibly amazed. The only thing that went wrong was that the string at the chest was interpreted as a tie. It's amazing that it can generate something this rough, but I was more surprised by my own shallow perception that only saw拙 lines, and by my daughter who was expressing things properly. Though the form is different, I realized that there must be countless instances where AI provides such insights, and that the effect that surpasses parents' and educators' understanding will spread tremendously from now on.