Author Ted Chiang said in a June 2026 [1] piece for The Atlantic that artificial intelligence is not conscious.
The debate centers on whether large language models (LLMs) are genuinely experiencing the world or simply simulating human-like responses. As AI becomes more sophisticated in its mimicry, the distinction between simulated intelligence and actual sentience becomes a critical point of philosophical and scientific contention.
Chiang said that AI behavior is a simulation rather than a genuine experience [1]. This perspective is supported by neuroscientist Anil Seth, who said at TED2026 that artificial intelligence is unlikely to become conscious [2].
Other observers have echoed this sentiment. An author for The Verge described the Atlantic piece as a well-deserved smack in the face to the clowns who suggest LLMs might be conscious [3]. The argument posits that the ability to process and generate text does not equate to the internal life associated with consciousness.
Despite these assertions, the topic remains a point of public and media debate. Some outlets, such as The News (Pakistan), have questioned if AI is secretly conscious, describing the possibility as a million-dollar question [4].
The discussion has also migrated to technical communities. On Hacker News, the article garnered 30 [5] points as users debated the intersection of machine learning and philosophy. The core of the disagreement lies in whether consciousness is a byproduct of complex information processing, or a biological phenomenon that cannot be replicated in silicon.
Chiang and Seth said that the appearance of consciousness in AI is a reflection of the human data it was trained on—not an emergent property of the software itself [1], [2].
“Artificial intelligence is not conscious.”
This debate highlights a growing tension between the functional capabilities of AI and the philosophical definition of sentience. By decoupling intelligence from consciousness, experts are attempting to prevent the anthropomorphizing of software, which could otherwise lead to misguided ethical frameworks or unrealistic expectations regarding the nature of machine learning.
