Instead of describing “chairness” to a computer, we instead just plug it into the Internet and feed it millions of pictures of chairs
Deep Learning Is Going to Teach Us All the Lesson of Our Lives: Jobs Are for Machines
Scott Santens
7.5K192

So structuring the data into “chair” and “not chair” is the real work here. Images aren’t inherently tagged in this way, the nugget of truth that image recognition software is stacked on top of is an inherently human recognition + data entry.

Do you think there is a finite taxonomy of the world that we can work our way down? First you start with “chair/not chair”, work your way to “table/not table”, and in no time you’re at “beautiful/not beautiful” and “buzzword-filled-article/not-buzzword-filled-article”.

Like what you read? Give Ryan Iyengar a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.