The rapid advancement of artificial intelligence (AI) has transformed our understanding of language, meaning, and communication, with systems like ChatGPT demonstrating remarkable capabilities in simulating human-like dialogue. However, this paper argues that AI’s ability to process language remains fundamentally distinct from human understanding. While AI excels at formal syntax and linguistic representation, it lacks the intentional grounding and contextual awareness that characterize human semantics. Human meaning is inherently tied to intentional acts – mental states directed toward external objects or concepts – and is deeply rooted in lived experiences and social contexts. AI systems, operating within pre-defined computational frameworks, cannot replicate this intentionality or engage with the interpretive practices that give human language its depth. Drawing on key philosophical concepts, this paper highlights the limitations of AI, emphasizing that genuine semantic understanding arises from intentional and contextual dimensions inaccessible to machines. These insights have significant implications for the philosophical, practical, and ethical considerations of AI, urging caution against conflating probabilistic outputs with true understanding. Ultimately, the paper concludes that while AI systems excel in formal language processing, they remain incapable of replicating the intentional and contextual nature of human meaning, underscoring the unique qualities of human cognition.

The massive and rapid leaps forward in the development in large language models (LLMs) have been matched by an equally massive and rapid increase in their commercial deployment. The current architecture of such LLM-type AIs suffers from a principled inability to distinguish between things that are merely accidentally not said (perhaps because they are not pertinent to the topic) and things that interlocutors are or should be intentionally silent about. We diagnose this failure to be the source of a variety of (severe) harms caused by AI in the past and present, and anticipated for the future. The LLM industry reacts to any of these problems with a "more data" strategy, resulting in some superficial success. This is veneer over a rotten core. The real problem is that the logic the LLMs are built on is too coarse-grained. We draw together recent research in a variety of different areas in logic and semantics (three-valued free logics and their proof-theory, bounds semantics, discourse representation, question-under-discussion framework, semantics for monstrous contents) to show how to build a better logical foundation that can handle deliberate silences in a contextually sensitive manner.

In 2018, a video of Barack Obama calling Donald Trump a “dipshit” went viral. Such “deepfakes,” AI-generated audiovisual media, are becoming increasingly accessible, even to those with only basic computer skills. The potential spread of deepfakes raises various epistemological and ethical concerns: they might distribute misinformation, gaslight individuals, infringe on personal rights, and undermine the credibility of genuine media. Consequently, some actors, like political parties, have pledged to avoid using deepfakes. However, deepfakes could also have positive uses: law enforcement might use them to infiltrate criminal organizations without risking undercover agents, and the media could illustrate stories when no genuine recordings are available. This raises the question: Is it morally acceptable for well-meaning actors to use deepfakes for positive purposes?

To answer this question, I argue three points. First, some dangers of deepfakes will diminish as they become more common. Second, other dangers depend more on their accessibility than their prevalence. Third, once deepfakes are ubiquitous, we should compare the knowledge from audiovisual media to knowledge from testimony rather than perception. Based on these points, I will present two criteria for whether well-meaning actors should use deepfakes.

Computanionalism about cognition is widely accepted. Illusionists “eliminate” consciousness by reducing it to cognition. The illusionist framework is thereby often taken to be very permissive with respect to machine consciousness. However, the analogy between brains and machines depends on the underlying account of physical computation, i.e. what one takes it to mean for a physical system to compute and thereby realize cognition. According to David Chalmers’ account (2012) or the more recent robust mapping account (Anderson & Piccinini, 2024) physical computation in both brains and machines is entirely constituted by the system’s physical structure in order to avoid pancomputationalism in the form of observer-relativity. In this contribution, I will critically review the robust mapping account and argue that it cannot fully account for biological cognition. Instead, an account of biological computation – unlike digital computation – must involve a trade-off between observers’ explanatory interests and physical structure. This then corresponds to Dennett’s framework according to which mental states, i.e. computational states of the brain, are real patterns identified by an observer taking the intentional stance. It follows that illusionism is much less permissive with respect to machine consciousness than commonly conceived.