このプロジェクトは人の偏った認知バイアスを機械学習等で学ばせ、逆張り、もしくは別のバイアスを道具に実装し「公平な社会を目指す」、その是非と方法について問います。私たちは一体どのようにそれをデザイン、実装してゆくべきなのでしょうか?叩き台としてアメリカの「Black lives matter」活動を反映した「非武装で警察に銃殺された黒人」の過去数年のデータから、殺されやすい人を学習、判別をし条件にあった場合は銃の引き金を数秒止める「Anti Bias Gun」を提案します。

What if we install an algorithm that works as a counter to human false cognitive biases, by using tools that learn the current biases using deep learning and similar technologies? This project questions how we should design “the fair society” using technologies, and whether this is the right or wrong direction to go. As a speculative draft, I exhibit an “Anti-Bias Gun” that reflects the “Black Lives Matter” activist movement. This gun locks the trigger for a few seconds depends on the target image. The gun has learned the concept of unarmed black people who were killed by police, based on the last few years of data.

I would like to portrait the bias of this case, in the other word facial image of victim tendency. Collect the victims face image and use deep learning to extract the attribute of the who were tend to be killed by police without arms.