On June 7th Woodbine hosted the third in our series of roundtables, On AI Melancholy, which tried to critically and experimentally dissect, reconstruct, and redeploy some important tropes in American thinking about AI. What follows is an updated and expanded transcript of my presentation, which focused on the problem of ‘brainrot’ as a proof of concept for AI criticism. It sought to situate brainrot in the broader political economy of platform capitalism, and examined some possibilities for using the phenomenon for critical purposes.
In my introductory Substack piece, I laid out a series of contemporary social problems: attention regulation, political reflection, loneliness, anxiety, deterioration in the quality of cultural production, and more. In this text, I want to take some steps towards a unified treatment of these issues. The key thread here for me is the idea of intellectual labor. Both terms are essential, and I want to capture positive valences of both.
The important struggle of the present moment is to recover the characteristic virtues of intellectual production – of reflective thinking directed to the fulfilment of real social need. These problems are problems for the possibility of real thinking. It is harder to think when you can’t regulate your attention, when you are suffering from social pathologies, when the cultural raw material of thought feels produced and reproduced by a swamp of digital content, or when one’s proposed audience demands ‘content’ under conditions inimical to real cultural or artistic value.
It is already conventional wisdom that ‘smartphones rot the brain’, that AI destroys our humanity, that the imperative to produce content somehow degrades our collective store of life-giving culture and so on. But the abstract and general terms in which these intuitions are usually formulated also undermine their force. They leave two important questions unanswered.
First, what specifically is bad about contemporary technologies? A booster might respond that the faults laid at the door of contemporary digital technology might equally be attributed to any modernizing forces, from the 17th century public sphere to the Gutenberg Bible. It is a common trend to engage in romantic criticism of any advance in general communication, in any mediation in the symbolic realm, and of any automation of existing labor practices.
Another aspect of this problem is that our concepts of the technologies themselves are flawed. Simon Lindgren calls AI an ‘empty signifier’, noting that it “can mean many different things, not only in technological terms”. So, in using general terms for such technologies, we risk mystifying our objects. To respond to this criticism, we need to explain why this time is different. What is distinctively bad about the contemporary nexus of AI, smartphones and social media platforms?
Second, what sort of ideal do we envision? The other side of the booster’s rejoinder might be that tech critics often have vague and generalized standards in mind. For all of its intuitive power, what analytical force does it have to claim that ‘AI prevents us from thinking’, or that ‘the realm of culture is polluted by the tsunami of content’? Our lay criticisms of tech sometimes assumes that traditional humanistic vocabulary – ‘human’, ‘reflective’, ‘intellectual’, ‘cultivated’ – has enough coherence and purchase here. But these terms have famously been degraded and polluted by centuries of misuse. We need a more specific value vocabulary to orient us here.
Brainrot
I want to begin with the concept of ‘brainrot’. The notion has significant cultural cache. Beyond its popularity on TikTok and amongst Gen Z, it is useful to note that it was named the Oxford Word of the Year in 2024. Like many contemporary keywords, it has a set of meanings at once contradictory, contested, and interlocking.
‘Brainrot’ in the contemporary sense names both a type of content, as well as an experience of felt mental degradation, perhaps generated by consumption of the same content. The link is not essential, as it is possible to be ‘brainrotted’ without consuming brainrot. In both cases, there is an association of nonsense. Brainrot content involves images and memes that somehow strain against our intuitive sense of reality – an ‘uncanny valley’ picture of Putin shaking hands with the fictional Ronald McDonald might be a useful paradigm case here.
The brainrotted subject is cognitively disabled in some way – they have consumed so much content that they have lost touch with reality, found themselves unable to make meaning or memory. The vernacular is suggestive here – those suffering from brainrot are ‘cooked’ in a way that undermines their very ability to function.
A variety of psychological ills follow. This summary in Brain Science Review is representative:
“brain rot leads to emotional desensitization, cognitive overload, and a negative self-concept… impairs executive functioning skills, including memory, planning, and decision-making. The pervasive nature of digital media, driven by dopamine-driven feedback loops, exacerbates these effects.”
Brainrot in this second sense is a trope of contemporary tech skepticism. The criticism in its most familiar form comes from the discourse around smartphone use in children and young adults. The moral panic is that smartphone use comes with a whole host of mental pathologies: loss of capacity to attend, loss of capacity to reflect – in other words, brainrot by another name.
The received explanation for smartphone brainrot, promoted most famously by Jonathan Haidt’s Anxious Generation, is deterministic, both in the technological and the psychological senses. The thought is that algorithmically driven content primes our dopamine receptors in such a way that makes addiction inescapable.
The solutions mix Luddism and Christian Morality. They offer the following theory of the case: Brainrot is built into the technology-individual interaction. Thus, we can only make progress by banning the technology for impressionable youth, as seen in the smartphone bans increasingly popular in the advanced world, while older sufferers are encouraged with abstentionist moral education, which we see in the ‘digital detox’ mania.
This rigid ‘ban-based’ approach is inevitable, given the psychological assumptions built into the culture. If smartphone addiction is the product of certain basic human needs – for comfort, for meaning, for stability – then the regulatory solutions must thus strike at the heart of the problem, either by removing the technology completely or restricting the basic desires that push us to use it.
Those of us influenced by critical social theory are rightly skeptical of a purely psychological appeal. Christian appeals to desire can never be the end of the story, as ‘desire’ itself is a fraught category. Consider Deleuze and Guattari on desire:
“Desire has nothing to do with a natural or spontaneous determination: there is no desire but assembling, assembled desire. The rationality, the efficiency of the assemblage does not exist without the passions the assemblage brings into play, without the desires that constitute it as much it constitutes them.”
Desire itself is the beginning, not the end of explanation. Without thinking about all desires as ‘implemented’ or ‘imposed’, we might think of desires as an element brought into play by an assemblage: a network of techniques, effects, and mechanisms perhaps themselves driven by a complex of desires.
We thus need to think more deeply about the source of the desire for content, for addictive smartphones. What are the background conditions that promote this desire?
The specific feature of platforms that produces brainrot stems from the way they turn attention into a commodity; into something to buy and sell. I will follow the sociological literature in calling this the attention economy.
Consider the way that platforms measure your presence on them. They collect clicks, track time spent on the page, track the ads you click on, track where your mouse moves, charge you to avoid ads in the form of subscription fees, and then package all the resulting data into a product to sell to advertisers.
Why do platforms valorize attention in the first place? What is the principled reason for doing so?
I follow a proposal by the sociologist Hera Lee in thinking of attention as a kind of collateral, an asset whose value is underwritten by the trust of the borrower. Attention is a signal to investors that the company might make profits in the future.
‘Data mining’ in this context is not literal mining, but acquiring the means of signaling potential profitability. But this begs a further question: Why does attention even function as a signal?
Part of the answer is the financialization of the platform economy: the way that the majority of capital coming into platform firms is venture capital, driven by investor sentiment and expectations about a structurally uncertain future. Lee summarizes the situation thus:
“Contemporary tech monopolies operate not by selling data commodities, but by transforming user attention into rent-yielding assets kept within the platform’s enclosure to signal future value to investors.”
The use of rent is intentional. Attention is not an autonomous source of profit ex ante. It functions as a way for platforms to gain market share by showing that they have a captive user base and thus a large potential demand for any product they might produce. Thus, attention economy techniques are best seen as another kind of monopolization, where securing more rent-yielding assets generates a barrier to entry.
Thus, platforms can often have massive operating losses. This has often led to dire predictions about the radical destruction of platforms. Consider Nick Srnicek’s claim that:
“the sharing economy will be a short-lived phenomenon. Most of these firms will go bankrupt, or turn into luxury services for the rich, or transform themselves into a different type of business model altogether. (Incidentally, the latter is Uber’s strategy, with their efforts to develop and own a fleet of self-driving automobiles.) In any case, sharing economy platforms are not long for this world.”
The problem here is too rationalistic a model of capitalist survival. Lee’s work helps us see how digital platforms evade a narrow deterministic logic of profitability. She understands techniques of attentional capture as a form of enclosure, paralleling older processes of primitive accumulation. The parallel is elaborated as follows:
“Platforms do not simply sell attention; they entrap it within an enclosure, thereby attracting investment capital. By treating attention as an asset that yields future rents, platforms decouple their valuation from traditional productivity, relying instead on the speculative expansion of the user asset through a logic of rentiership that demands the continuous extraction of cognitive surplus.”
This nexus between finance capital and the profit-making strategies of platform capitalist firms sets the background for brainrot. For Lee, Brainrot is a structural consequence of the platform as digital architecture:
“Platforms favor fragmented, high-frequency interaction because such patterns generate dense, rapidly updating behavioral signals capable of stabilizing speculative valuation, consolidating infrastructural dominance, and securing long-term monopoly rents.”
Let us come full circle. We began with a moral panic: the dopamine intensive pattern of smartphone consumption. We have now come to a structural diagnosis: the ‘addictive’/’dopaminergic’ quality of platform algorithms comes from a particular market dominance strategy specific to the financial conditions of contemporary capitalist production.
I will conclude by considering a question that Lee does not: what is the critical function of brainrot? How does brainrot fit into the traditional project of showing capitalism not just to be unjust but to be irrational and self-undermining?
Can brainrot be like wage theft, a concept that, if it does its job right, motivates people to overcome injustice?
Here’s a natural ‘hidden abode’ story: capitalism needs attention in the long run, but it eats attention in the short run. Brainrot is thus a sign that capitalism is starting to undermine itself. Attention is another ‘natural gift’ for capitalism, but which it treats as a free gift. Nancy Fraser puts the free gift concept thus:
“as an infinite reservoir from which we can take as much as we want and into which we can dump any amount of waste. In fact, neither nature nor social reproductive capacities are infinite; both of them can be stretched to the breaking point.”
I’m not sure. Does capitalist production really need quality attention the same way it needs labor or that it needs air? Couldn’t it get by fine with an army of brainrotted proletarians doing its bidding?
Furthermore, it is not clear that attentional capture is built into platform capitalism the way that, say, the wage relation is built into capitalist production writ large. Brainrot may just be a systematic externality, rather than a ‘central contradiction’ like labor exploitation.
Another reason to think attention is not central to capitalism in the traditional way is the strange coalitions that coalesce around attentional capture. At one level, there isn’t any determinate legislative or policy upshot to brainrot and attention capture, just an inchoate fist-shaking anger at ‘the way that big tech has taken over our lives’. This can be felt equally by reactionaries, liberals, and revolutionary socialists.
There are exceptions to this. Consider the social media bans recently proposed in Australia and Malaysia. Even there, the coalition can largely be seen as the familiar middle and upper-middle class parental coalition routinely mobilized around ‘moral panics’.
We are left in a troublingly ambiguous place. We have a phenomenon which is central to capitalist production, but due to its status as a ‘byproduct’ of exploitation, is not amenable to politicization along the wage relation itself. In this respect, brainrot is like pollution, which affects a vast and potentially universal constituency.
Any viable politics of brainrot must thus root its basic values – attention, reflection, ‘mental health’, ‘self-regulation’ – in more basic political materials. Perhaps the right connection is between attention and democracy. Or between attention and exploitation. Or even between attention and religious conversion. In any case, a pure anti-brainrot movement is thus probably too capacious to be strategically useful. Left ‘friends of attention’ must look elsewhere.
Amogha Sahu is Woodbine’s Research Fellow. He is a philosopher with interests spanning the foundations of ethics to the history and origins of work. He is an incoming lecturer at the Department of the Core Curriculum at Columbia University.