TL;DR
It is argued that both AI and bullshitters are untrustworthy informants, and for similar reasons, it is natural to describe AI’s informational outputs as bullshit, as it signals their distinctive kind of epistemic deficiencies, which they share with bullshit.
We explore and argue for the view that intelligence beyond a certain threshold is negatively correlated with consciousness, so that the more intelligent a system is, the less likely it is to be conscious. ‘Intelligence is the enemy of consciousness’, as we put it. This view entails that superintelligent AI models are likely to be zombies. After presenting an initial defence of the view that draws on the research programme of resource-rational analysis, we argue that recent developments in AI research provide additional evidence supporting it. We conclude by exploring its moral implications. First, we suggest that if our view is right, a certain type of moral risk can be averted since researchers are unlikely to stumble into accidentally creating AI models that are conscious. Second, we discuss a new form of AI risk that arises from our view.
The philosophical debate about bullshit has focused exclusively on what bullshitters say. In this paper, I suggest we should also pay attention to how they say what they say. In particular, we should look for what I propose to call highfalutin bullshit: the needless use of “big words” with manipulative intent, whether deceptive or otherwise—that is, an intent to influence the audience’s beliefs through means that bypass rational agency. First, I seek to explain its nature using Grice’s conversational maxims. As I show, all forms of bullshit philosophers have hitherto focused on can be analysed in terms of non-observance of Grice’s maxims under the categories of Quality, Quantity, and Relation. In contrast, highfalutin bullshit involves a manipulative non-observance, whether deceptive or otherwise, of Grice’s category of Manner and, specifically, the first maxim of Manner: “Avoid obscurity of expression.” After that, I respond to two arguments against my proposal. According to the first, “highfalutin bullshit” is just a new name for phenomena already described in the literature (e.g. obscurantism). According to the second, it is unclear why we should care about it. In response, I argue that highfalutin bullshit should be viewed as a form of bullshit in its own right, widespread and problematic. Finally, I contend that it is instructive to view highfalutin bullshit through the lens of argumentation theory, namely, as a means for masking or diverting attention away from argumentative bullshit.
Background: Readers of Don Quixote have long divided between two verdicts: the knight is mad, or the knight is a martyr to a disenchanted world. Both readings assume that the truth of his acts lies inside the acts themselves; this paper argues that it lies in the world’s answer. Methods: The paper develops pretension as an epistemic category and constructs a three-term reading grammar — frame, uptake, and aftermath — through Austin, Butler, Goffman, Schechner, and related theories of performance and recognition. It applies this grammar to two structurally opposed scenes in Cervantes’s novel: the inn where Quixote is dubbed a knight and the windmill plain where his wager is refused, and a third, where the verdict is neither granted nor refused but held in dispute. Analysis: At the inn, the keeper grants sanction without belief: the rite is completed, the designation acquires social currency, and Quixote leaves with a standing conferred by others. On the plain, Sancho refuses endorsement of the frame, and the same kind of wager returns to Quixote as error and madness. Conclusion: Madness and genius are not simply differences of content or psychology; they are also verdicts rendered by audiences before whom claims are staged. The paper extends this structure to scholarship itself, where knowledge is never ratified privately but waits on uptake. Novelty: The article reclaims pretension from its moral weight and proposes it as a technical term for performance with consciousness of performance, tested by social uptake.
Our reliance on the new wave of generative AI has increased exponentially in the last few years, with little sign of abating. This raises the natural question of whether we should trust this technology as a source of information. Drawing on the epistemology of trust, it is argued that generative AI does not meet the conditions for being a trustworthy source of information. In particular, it is argued that generative AI is untrustworthy because it is an unsafe source of information. This, in turn, entails that trusting generative AI as a source of information is not a route to knowledge.
David Eliot trained as a card magician before turning to AI research. His Artificially Intelligent sets out to democratise AI for general readers, and on its own terms it succeeds: the history is rich, the writing is clear, and the central argument that AI is socially constructed and democratically redirectable is held with conviction. This review reads the book alongside the regulatory architecture being built on the same theory of agency: the EU AI Act, the harmonised standards beneath it, the UNESCO and OECD ethics frameworks above, and the November 2025 Digital Omnibus provisionally agreed upon in May 2026. The argument here is additive, not corrective. Knowing the trick is not the same as breaking it. The real subject of magic is attention, not secrets. Alongside AI ethics and governance literature, the magician’s eye points to what visibility cannot reach: the case for a code of practice from inside a profession that understands attention management.
Is the human predictable by AI, and if so, should we allow it? This paper examines the relationship between these empirical and normative questions from the perspective of the right to free development of personality enshrined in Art. 2 of the German Constitution. Based on an analysis of 50 decisions of the German Constitutional Court, I advocate for the recognition of a right to unpredictability derived from Art. 2(1) of the Constitution. I argue that, even if humans statistically follow patterns, the freedom to act against all probabilities is worth protecting.
Generative AI
Google DeepMind partners with game studios to prototype breakthrough AI gameplay.
A Blog post by Liquid AI on Hugging Face
A new method for surgically removing training examples from a model reveals that as datasets grow, the link between what a model learns and what it produces dissolves.
