Over the last few years, through my photography practice, I have watched an increased use of Artificial Intelligence in the tools I use, such as for: keywording images, editing images, and most recently, text to image generation.
The first two strike me as evolutionary steps in functionality, powered largely by AI. Keywording is a painful, time-consuming, and an often-mind-numbing task. Similarly, editing images is challenged by the technical difficulties of selection, among other things. Interestingly, both these capabilities involve identifying objects in a picture, with the former mapping the object onto a keyword and the latter going a step further and suggesting how to edit the object.
The third use of AI cited above is different. With text-to-image generation the software can translate terms into an image. It can be quite good at this, raising that recurring question of AI’s role and its potential to replace people, for example.
I think one’s point of view on the matter depends on their perspective on humanity: do they fall into the trans-humanist camp or the post-humanist camp?[1] The former would see AI as a tool, subservient to humanity, essentially a legitimate form of slavery without the ethical issues tied to the subjugation of other peoples. This holds true until these AI-enabled “tools” become sentient robots at which point we enter a dystopian Blade-Runner-like world of human-looking robots seeking equality and rights.
A post-humanist would see an AI-based “tool” in a more co-operative role; one that complements and extends the limits of each party (human and machine). Susanna Lindberg, philosopher and faculty member at Leiden University, offers:
“The main societal function of robots is precisely work: they represent a skilled workforce that does not need rest nor salary. In this situation, it is natural to think that what robots do in art is taking over a part of the raw artistic work so that the artist can concentrate on creation understood as conception.”[2]
Here Lindberg is separating the craft of art from an intellectual component, assigning the latter contribution to the human. Others have alluded to similar distinctions as well.[3] I suspect this follows naturally from the legacy of our renaissance-era humanist philosophy that held that intelligence is unique to the human species.
Currently many AI programs intend to model human intelligence [4] but should it be human intelligence that we model? In asking the question of whether machines can create art, Mark Coeckelbergh, philosopher of technology, suggests:
“Perhaps contemporary machines could show us a different perspective on the world, one that does not let appear everything as a standing-reserve (of information, for instance) but one that reveals the world in a different way than what we are used to see.” [5]
Could these different perspectives be derived by modelling other forms of intelligence?
What are other forms of intelligence? Dogs seem to show signs of intelligence as do squirrels, cats, whales, lions, among many other forms of life. I suspect their respective intelligences are different than our own. They may emphasize skills that we lack or deem irrelevant. Whether their “intellectual capacity” totals to the same or less or more than we have might be less important than appreciating a different expression of intelligence and what it offers to us.
What would a whale’s form of intelligence offer? May be a different point of view on the oceans? A different perspective on communications, on society, etc. I don’t intend to suggest one is better than the other. I will assume that each form of intelligence has evolved to meet the needs of the host. But I think it is safe to say that it is different, and if that difference is not important, it should at least be interesting.
However, to model the intelligence of some other form of life seems a stretch. A more realistic goal for our present time might be to be more sensitive on how and what we model in human intelligence. A criticism often levelled against AI systems is they embed cultural biases in their modelling. A recent study by Aylin Caliskan and Ryan Steed, of the Brookings Institute, found evidence to support that position.
“Just as humans are exposed to systemic injustices, machines learn human-like stereotypes and cultural norms from sociocultural data, acquiring biases and associations in the process. Our research shows that bias is not only reflected in the patterns of language, but also in the image datasets used to train computer vision models.”[6]
This raises an interesting possibility that AI engines could be used to interrogate our own intellectual shortcomings, biases, etc. Could we find that essence of intelligence that is common across humanity, devoid of biases etc.? Would this lead us to some universal truth of “being”? Is AI a tool to debias our culture, society, ourselves, or our art.
Do we want to remove all biases or simply recognise them and manage them accordingly? What would it be like to live in a world without bias?
[1] Ferrando, Francesca. “The MAIN DIFFERENCE between Posthumanism and Transhumanism”. December 7 2017. YouTubehttps://youtu.be/l2J5EiB3vrA
[2] Lindberg, Susanna. “Robots in Art: Where’s the Work?” Http://Workworkworkworkworkwork.com, 2019.
[3] Coeckelbergh, Mark. “Can Machines Create Art”. 2016. Springer. DOI 10.1007/s13347-016-0231-5
[4] Caliskan, Aylin and Steed, Ryan. “Managing the risks of inevitably biased visual artificial intelligence systems”. Brookings. September 26, 2022. https://www.brookings.edu/blog/techtank/2022/09/26/managing-the-risks-of-inevitably-biased-visual-artificial-intelligence-systems
[5] Ibid, Coeckelbergh
[6] Ibid, Caliskan
Leave a Reply