Weekly 5 - Reading Response

It’s interesting the distinction being made in Hito Steyerl’s article between data and our perception of data. He makes a point that the modern forms that data take mean nothing to our human senses without a machine to play middle-man and translate that data into a form that we can understand. This is exactly how we can be drowning in data and not have enough at the same time. For most of the history of the internet, we were learning as a society how to think like a computer, but now thanks to AI we can teach the computer how to think like us for the first time. As AI models get better and better, humans have to do less interaction with middle-men to get data in a familiar form.

Instead of having to format systematic, optimally worded search queries to find the information we want, we can now ask the internet our questions in a human way- an imperfect, even run-on-sentence kind of way. Same with image generation. We can simply write out what we’d like to see instead of having to iteratively compile, transform, and distort. The better AI gets, the less data we have to interact with, but in order to make AI better, we need to get more data. Thats the catch 22.

For The Class:

Does the overwhelming amount of information available online paradoxically limit our knowledge rather than expand it? If so, how?