AI Colorization Lab Week 4

Here is my Omeka item.

The black-and-white image I chose to colorize is Men’s Dormitory / West Hall interior: lobby, with eleven men from Carleton’s Digital Collections. On the left is the original photo, and on the right is the colorized version. I was surprised by how powerful AI is nowadays because I think it did a great job fitting the picture into what people assume the colors should be. For example, it’s normal and classic for suits to be blue. Despite the excellent colorizing job it did, I still have several concerns. As Sonja Drimmer mentioned in her article:

Adding color does not show things as they were but recreates what is already a recreation – a photograph – in our own image, now with computer science’s seal of approval.

Drimmer, Sonja. “How AI Is Hijacking Art History.” The Conversation, 3 Dec. 2024, theconversation.com/how-ai-is-hijacking-art-history-170691

She highlights the ethical concern of misrepresenting historical accuracy. Besides the color blue, it’s highly possible that the suits these gentlemen wore were actually black. In this case, the AI-generated colorization of old photographs distorts our understanding of the past by imposing modern interpretations, potentially erasing the authenticity of historical records.

Another quote I was pretty impressed by is:

The field’s sensitivity to historical particularity and cultural difference makes the application of the same code to widely diverse artifacts utterly illogical.

Drimmer, Sonja. “How AI Is Hijacking Art History.” The Conversation, 3 Dec. 2024, theconversation.com/how-ai-is-hijacking-art-history-170691

I totally agree with what Sonja said. A photograph reflects the creator’s intention, and every creator has different intentions. How can we interpret millions of different intentions with the same cold code? As a result, it’s reasonable to say that AI colorization reduces the meaning of the artwork and flattens the richness of human history.

To sum up, I believe AI colorization does a good job visually presenting the past by bringing old photos to life with color. However, we need to be careful when interpreting the content of these images because they might mislead us with a distorted perspective.

3 thoughts on “AI Colorization Lab Week 4

  1. Hi Peter! I definitely agree with your stance on the ethical concerns of AI re-colorization. I specifically resonated with Sonja’s second quote, as it is true that using the same code for every photograph will ultimately cause them to all look similar and strip away some of their originality. ‘Historical particularity’ as she puts it is something I feel AI does not take into account when re-colorizing photos.

  2. I completely agree that AI image manipulation can be extremely misleading and does not capture the creator’s true intentions. After all, a machine is not human and won’t be able to capture the true meaning of the piece of art. Your first quote sums up this idea quite nicely. When it comes to recreating art, AI is good at recognizing patterns and that’s about it, but in doing so, it removes the human dimension of art. Well done Peter!

  3. Hey Peter, I really agree with your point about the ethical concern of misrepresenting historical accuracy. I really like that you wrote: “the AI-generated colorization of old photographs distorts our understanding of the past by imposing modern interpretations, potentially erasing the authenticity of historical records.” I feel like this phrase greatly sums up the issue of fiction that AI is doing.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

css.php