in Uncategorized

How Bots See Art

Back in 2015, several months before releasing Expand I returned back to working a regular day job. As I’d previously been teaching at university this return meant that I was now programming full time for someone else, something I hadn’t done before. One of the downsides with programming jobs is that after work you feel less inclined to work on your own personal programming projects. This is a good thing. After sitting inside all day, it’s probably a good idea to get out and do something else.

Even so I feel inclined to still make stuff so I started playing around with Twitter bots. I started by making the Who’s That Poke’mon bot based on the Poke’mon TV show. Then I made three bots that would take a daily quote and display it in ASL, Auslan and Braille. One night I made a small bot called Powerball Bot that would randomly pick lottery numbers before the US lotteries. It’s kind of amusing to think that it’ll eventually pick a winning set of numbers.

Recently I made a new bot called How Bots See Art. It takes images of pieces from the Metropolitan Museum of Art in New York and describes them using Microsoft’s Computer Vision services. It’s produced some amusing results.

I’ll probably continue to make a few more bots in the future. I really love how for such a small amount of work they can produce rather neat results.

Write a Comment