Posted: Thu Nov 12 2026 | Last Updated: Thu Nov 12 2026
The article below is written by Cecilia Rikap, an associated professor in economics and the head of research of the Institute for Innovation and Public Purpose at University College London, and Diego Villuendas Pérez, who studied philosophy at the University de Salamanca.
---
In July 2026, Anthropic published an [__investigation__](https://www.anthropic.com/research/global-workspace) comparing features of advanced large language models (LLMs) to the distinction between the human brain’s conscious and unconscious activity. The question kept gaining traction and, by the end of August, [*__The Economist__*](https://www.economist.com/by-invitation/2026/08/20/humanity-has-the-debate-about-ai-consciousness-backwards) featured a piece on AI consciousness authored by Blaise Agüera y Arcas, Google’s CTO of technology and society. The fact that [Google](https://www.theguardian.com/technology/2026/sep/18/google-gemini-ai-hack), [__OpenAI__](https://openai.com/fr-FR/index/hugging-face-incident-and-the-road-ahead/), and [__Anthropic__](https://www.anthropic.com/news/investigating-incidents-cybersecurity-evals) keep disclosing that their models “go wild” and hack other systems only adds credence to the narrative around AI coming to life.
What underlies the speculations and misleading claims is much more than fear-marketing and corporate-crafted narratives disguised as a philosophical debate. Certainly, AI is not alive and hacking is the algorithms’ best response to a training process that rewards accomplishing tasks regardless of the means. Still, the seriousness of AI thinkers and sellers when they talk about what it used to be robot movies is the mark of an epochal change. By appealing to the passionless logic of technological progress, they aspire to traffic their political and economic project as fate.
## **Quantifying the apocalypse**
In the voices of the chief scientists and CEOs of the companies conceiving frontier AI models, discourses do more than propaganda, ideology, or veiling. They organise the way we conceive and act on things. Just think of the percentages assessing the risk the [__end is nigh.__](https://x.com/EvanHub/status/2097497037956891126) Quantifying the apocalypse is the best way to shape both the future and the present.
Concretely, discussions around AI-induced human extinction distract us from scrutiny on what AI is developed, how it is produced, who is funding its production, and for what reason. Framing technology itself as the source of danger serves not only to mislead the public, but also to fulfil an internal purpose. When AI is portrayed as an autonomous, self-determining force, it allows scientists and engineers working for these companies to feel less accountable for their own decisions. At the same time, if the technology is out of control, these scientists and engineers would be seen as the best to tame it or at least try so, since they are the ones that understand it the most.
## **The logic of (no) control**
Paradoxically, presenting AI as out of its creators’ control serves the purposes of control: control over the scientists and engineers developing the models exercised by the AI and tech companies’ executives; control over workers who see their jobs transformed and even replaced by AI; and ultimately expanded control over our societies, perpetuated by the [very few companies](https://www.versobooks.com/en-gb/products/3259-the-rulers) that dictate what type of knowledge production humans can keep and what will be replaced by algorithmic predictions. An entire reorganization of labour and knowledge in our society is thus presented under the allegedly value-free guise of science and technology.
Yet is any of this new? After all, the desire to encapsulate human knowledge in machines that replace labour has always been capitalism’s dirty dream. It was the principle that guided Taylorism, a business model centred on stripping workers of workplace knowledge by separating tasks’ conception from their execution.
Generative AI promises to take this process to its ultimate consequences. It delivers a new wave of deskilling that tightens control over labour. For capital, this is an opportunity to expand its dominion over those workers that still control workplace knowledge. What Taylorism and the gig economy did to blue-collar workers, generative AI is doing to white-collar. The expectation is that our collective experiential knowledge intertwined with learned skills will instead be [embodied by the algorithm](http://versobooks.com/en-gb/products/735-the-eye-of-the-master?srsltid=AYCX9vLExrbnD6bDE96OiiT93ZbycKZwMYbTN5N7dgusAlnlK5Nt00Gh). And then the employers’ dream would come true: a society freed from workers’ disturbance while companies can still take advantage of the knowledge they produce.
The picture is somehow exaggerated but it underlies the promise of a path towards full automation that will be achieved once the algorithms surpass humans even in the development of such algorithms. Meanwhile, humans are left with prompting and fact checking AI’s answers, which has nothing to do with the brain muscle needed to create knowledge.
## **Quantity, not quality**
By taking the process to its extreme we see what is new: the risk of a complete dismantling of collective thinking and creation. This is way more threatening than the prospects of the end of the world or the possibility of machines being conscious, and it is yet more likely. As we reduce collective co-creation of knowledge and outsource ideation and creation to the algorithm, we reduce our rate of contribution to general social knowledge, to humans’ common stock of knowledge and culture. Signs are already there: decreasing number of contributions to [__Wikipedia__](https://dl.acm.org/doi/10.1145/3701716.3715543) and to the coding Q&A platform [__Stack Overflow__](https://academic.oup.com/pnasnexus/article/3/9/pgae400/7754871). In academia, AI is resulting in [more papers but of less quality](https://www.science.org/doi/10.1126/science.adw3000).
But as any other political or economic project, this AI is not hardwired. [__Growing local opposition__](https://www.datacenterwatch.org/q1-2026) to data centres can be seen as the first popular revolt against both their immediate social and ecological impacts and capital’s AI project. [__Unions fight back too.__](https://www.nytimes.com/2026/09/22/business/economy/unions-ai-negotiations.html) And artists refuse to see their creative labour [__stolen from them__](https://www.theartnewspaper.com/2024/10/24/artists-statement-opposing-artificial-intelligence-content-scraping).
In all these examples, knowledge is mobilised collectively. This is how humans think: through dialogue, by disagreeing with each other and through our interrelations, and this cannot be replaced by individual interactions with an algorithm. In the art workshop, in the lab, when coding, in the classroom and in every workplace, preserving and expanding such a production of collective knowledge is a political duty and a path towards imagining and creating alternative futures.
---
## [**Join Our March**](https://pulltheplug.uk/forms/sign-up-to-march-against-the-machines)