Figuring the Human in AI and Robotics
Pathology as Critical Theory
In this article, Lucy Suchman posits the critical theorist examining objects of Artificial Intelligence as an illicit pathologist. For Suchman, AI’s are materialised figurations*. They contain within them ideas of what is essential to humanness which are brought into meaty reality. When the illicit pathologist dissects these objects they open up a gateway to understand the technoscientific AI creator’s human figuration.**
Technologies can be understood as materialised figurations because they assume and imply certain interactions and associations between humans and humans, humans and nature, humans and… [fill in with your own ideas of what is separate to humans] Production of spades assumes human work the land. A ship assumes humans sail across water, implying trade and travel. A ship changes the possible lives of humans. A phone implies humans should interact with those beyond their immediate physical vicinity.
Suchman splits the understanding of the Technoscientist figurations into three domains; embodiment, emotion, and sociability. AI mimics specific functionalities of each of these domains. So, AI, and in particular robotic AI, can be a gateway to gain access to technoscientific understandings of the role of bodies in figuring the human, and the role of emotions and sociability. Suchman finds that these domains sutured together by a figurative thread. To briefly paraphrase, Suchman finds in each domain AI Technoscientists figure the human as a rational, autonomous, individualistic and functionalist being.
I would like to try my own dissection of an AI, or at least the beginnings of one. The AI I’m examining is from OpenAI’s* Multi-Agent Hide and Seek. (https://www.youtube.com/watch?v=kopoLzvh5jY)
This example is a simulation within which two opposing groups of ‘agents’ play hide and seek. OpenAI ran this simulation 481 million times. After approximately 22 million runs OpenAi claim to see humanlike behaviour which becomes more developed as the simulation gets to 481 million runs. Following Suchman, I will consider what human figurations are [digitally] materialised in this AI, and what this could tell us about the cultural collective imagination of this specific branch of technoscience. The example I am examining is specific, not general. I take it to be one part of their method to achieve “AGI”. I think it is a useful nexus for revealing the imagination of corporate AI.
“Coevolution and competition on earth led to the only generally intelligent species known to date: humans”
This quote does much of the examination for me. This AI presents itself to me with a note which contains a map to the anatomy of its figuration. There is far too much for me to say from this one revealing line. There are dozens of books on this specific line of thought. So, I’ll instead point towards some ideas in this figuration, and some books and articles I like on this. I think this should be done in the context of this quote from Harraway which sees science and Technoscience’s role in society as serving emancipatory practices which is drawn from the idea that human agency is relational and so is in some part constituted by technology in a net of interactions.
with the hope that the technologies for establishing what may count as the case about the world may be rebuilt to bring the technical and the political back into realignment so that questions about possible livable worlds lie visibly at the heart of our best science.
– DONNA HARAWAY
This figuration is deeply neoliberal. It hangs on the idea that competition and the free market are essential to intelligence.
Materialisation of the homo economicus
This AI figures humans as homogenous. It is the antithesis of difference [intersectionality]
In this Figuration intelligence = success - anti social?
*Figurations: From Harraway. Ideas of what is the essence of humanness. Similar to representations.
**TechnoScience: Also from Harraway. An active, practical outlook responding to the interconnectedness of technology and science.
Non-binary of tech and nature
***OpenAI is, in my opinion, in the corporate facing world of AI. It is based in Silicon Valley, was co-founded by Elon Musk, and has funding from Microsoft.In youtube lecture series and research papers, OpenAI presents itself as scientific. Its mission is to “ensure that artificial general intelligence (AGI)—by which we mean highly autonomous systems that outperform humans at most economically valuable work—benefits all of humanity.” (openai.com/about/) For Suchman and Harraway, I’d expect OpenAI epitomises the technoscientific. (TechnoScience link)