Annealed Media: Extensions of Machine-kind
How the lines between hot and cold media converge in the digital age.
My previous newsletter, the “Simulcra Civilization”, explores the nature of AI as a powerful and dynamic medium. This letter explores conclusions from the piece, but from a media studies lens, grounded in the works of Marshall McLuhan. In the mid-20th century, he described how the medium, not the content, was the message:
We shape our tools. thereafter our tools shape us.
You can read the article here by clicking the link below, or read on to learn more about AI and AI-based generations as an emergent kind of media.
McLuhan characterized "hot" media, such as print, as media that demand sharp, intense focus due to their high-definition nature. Each word in a book must be processed in a linear, sequential manner. In contrast, "cool" media, like television, is more diffuse and requiring broader, less focused engagement. Television engages viewers in a tactile, immediate, passive, yet participatory manner, demanding interpretation and completion by the viewer.
McLuhan's book "The Gutenberg Galaxy: The Making of Typographic Man" (1962) covers the impact of the printing press, a “cold” medium:
The invention of the printing press brought about a profound shift in human consciousness, as the dominance of the written word and linear thinking gave rise to a new sense of individualism and detachment from the communal oral culture of the past.
For centuries, print media cultivated mass literacy and ignited critical faculties through books and pamphlets. These days, AI algorithms ingest the surviving texts, by absorbing a vast textual corpora, fragmenting and automating it into decentralized streams. When AI serves as an interface to data, it acts as a mediator between the user and these streams of content. In this way, the experience becomes participatory and two-way, more akin to the characteristics of "cold" media. This is reminiscent to the medium of television, with, especially in its early days, had low resolution, the channels of black and white tickling our need for tactility. Users interact, ask questions, and refine searches with AI. In turn, the AI filters, personalizes, and presents data to the user. Yet given the sheer volume of online information and possible generations an AI could make from it, it becomes a challenge and becomes more of a curation tool. It requires users to interpret, correlate, and actively seek understanding from a loose connection of information. It forces us to think less deeply, but makes us actively and continuously attempt to come up with a concrete viewpoint, which is impossible in the electric age. Then, they “heat” up again.
This landscape shifts in more unusual ways when users consume content generated by AI. In this way, it appears AI embodies elements of both "hot" and "cold" media. On one hand, AI-produced content, whether it be a sequence, some music or text, provides a narrative, positioning users as non-participatory recipients, echoing the completeness of "hot" media. But the knowledge that this content originates from a machine nudges users of AI into a more active role, questioning the content's authenticity, biases, and the correlation between what they ask for and what they get. Given a pre-trained mode, much of its value comes from high-level strategy or expert knowledge. The machine-generated content can render the information into a mosaic of patterns and discern meaning. Then, the result would be read and digested in the form of “cold” media. However, this depends on the user, as another user, or perhaps an unsuspecting person, will simply consume AI-generated content that was generated into a “hot” medium. To help me make sense of these contradictions or inconsistencies, I considered whether AI-generated media could be considered “annealed”, using the terms “hot” and “cold” rather loosely.
The medium of AI, is trained on initially cool or hot media. In this case, I will shift the focus to on the effect medium, rather than strictly following every established description of hot or cold media. As it happens, famous models like ChatGPT consume hot media yet the interaction with it is perhaps more reminiscent of cold media. At the same time, dicerning meaning and processing the text is consumption of hot media. In this way, it goes from cold to got, hence the term “annealing”.
"In the age of electric media, the content of any medium is always another medium. Our attention is drawn not only to the content itself but also to the medium through which it is conveyed, shaping our perception and understanding of the message."
- Marshall McLuhan
Beyond content, there are tangible shifts in geological forces when a powerful medium emerges. Data centers, driven by AI’s need for reference material, change our landscapes from a distance. McLuhan described how the railway recast geography through steam and iron - "Totally new cities and work appeared, accelerating previous functions." In order to supply advanced silicon brains, we continue to strain nature's bounty - lithium mining exhausting communal waters, while e-waste accumulates in Ghana's forests. The environmental message of this emergent medium echoes concealed consequences we must confront.
All the while surveillance tools becomes a risk to existence as we know it. As data increasingly defines us, direct experience risks having to take a back seat and let digitized approximations take the wheel. This echoes Baudrillard’s thought - where representation might overshadow the real, symbols preceding the territory. People dissolve into data pools, optimizing inscrutable functions. Tangible reality becomes inaccessible behind simulations.
Is this the realm McLuhan prophesized, where realities are mediated through networks into shared experiences? Perhaps not, for as participation is passive, vulnerability is increased. With cognizance, we can grasp AI's messages - how it maneuvers us into saturation by data. Literacy and critical thought remain vital. We must guide our collective trajectory, not be programmed unconsciously. We must fear the end of Narcissus, as alluded to by McLuhan.
"The extension of our senses and capabilities through technology has the power to both amplify and numb our perception. We become enamored with our gadgets, yet simultaneously lose touch with the deeper meanings and effects of our actions.”
We risk allowing ourselves to become consumers of a infinite, artificially generated pool of media. In some ways, turning all media to a hotter one. As touched on earlier, even our consumption of AI-generated content, if this feature of its quality remains true, could more resemble hot media until a new medium containing as its content the older medium emerges. Sequentiality is still an unsolved problem in AI modelling. Still, the strength of AI is that it can appear to adapt to each user while simultaneously conforming entire groups to becoming vectors to the transmission of the medium of AI. Even the projection of being or divinity onto AI spreads the medium as the message. Even while the printing press carried the Bible into the midst of many centuries, ultimately it was the printed word itself that became the way of thinking in various societies such as the USA, Germany and France.
We must be aware of that a human can endlessly consume content, and now, there is a way for this to become a reality. Using AI can engage the mind, offer new kinds of skills and strategic thinking, but it also misses the depth that comes with other kinds of research or knowledge creation. Literacy, creative thinking and logic are more essential than ever, and even while AI replaces many jobs that require these skills, they remain more essential than ever. Together, let’s traverse a path of active participation, not mere spectation, and make sure our tools don’t take us to a post-literate, automated, technocratic nightmare. Read on to learn more how we can embrace better, more inclusive and safer AI usage and governance.