Image Net Roulette Response

Reading Response for NY Times Article

When introduced to Image Net Roulette I was unfamiliar with how artificial intelligence machines are actually able to learn, process and organize information. I simply saw the website as a satirical take on the uses of technology and the methods through which they “learn”. I was cautious of where the data provided to the site goes so I did not provide personal pictures.

However, I decided to take images of a cast of shows, particularly one with a remake to draw some comparisons. I settled on the cast of Full House, a show I loved watching as a child. The original cast resulted in some comical descriptions including “towhead” and “grinner”. The second image was the cast of the Full House remake “Fuller House”. This image seemed to have less comical, and in my opinion rather confusing and odd categorizations.

The NYTimes article “‘Nerd,’ ‘Nonsmoker,’ ‘Wrongdoer’: How Might A.I. Label You?” actually covers Image Net Roulette and highlights the advantages of as well as dangers of this “deep learning” for technology. It brings to light the ways in which AI technologies categorize and parse through large amounts of data. Image Net Roulette displays that biases and sweeping judgements are created through these labels that AI systems create. Some of them extremely offensive – some which I saw in my own search. While Image Net Roulette aims to uncover this risk in creating artificial intelligence technologies, without context it can reinforce the issue.