This is a wonderful article, thank you for writing it.


The strength of newer machine learning systems is that they can work with less explicit models, fewer assumptions. This can easily led to unwanted outcomes. But when that happens it also can expose things about our own assumptions, or our own values, that we hadn’t thought about.

Leaving discretionary power in the hands of us “I know it when I see it” humans isn’t a good solution. We need this push to improve our own thinking.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.