Even under general anesthesia, the human brain may continue processing sounds and language. A recent study found that certain hippocampal neurons remain responsive to auditory stimuli and can perform sophisticated language-related computations, even when patients are fully unconscious and have no conscious awareness or memory of the input afterward. Researchers used high-density Neuropixels probes to record activity from hundreds of individual neurons in the hippocampus of seven epilepsy patients during surgery. While the patients were under general anesthesia, the team presented auditory stimuli, including sequences of tones with unexpected “oddball” frequencies and natural spoken language from podcasts (such as The Moth Radio Hour). More than 70% of the recorded neurons responded to sounds. Some distinguished deviant tones from standard ones (with evidence of neural plasticity over time), while others encoded grammatical and semantic features of speech, such as parts of speech (nouns, verbs, adjectives) and even appeared to anticipate upcoming words in a sentence. These findings challenge the traditional view that general anesthesia fully suppresses higher-order cognitive processing. Importantly, the patients showed no signs of awareness, reported no memory of the stimuli, and remained deeply unconscious throughout. The results suggest that specific neural circuits, particularly in the hippocampus, can continue analyzing sensory information independently of conscious experience. The study does not imply that patients can consciously hear or remember conversations during surgery. Rather, it highlights surprising residual brain activity and raises questions about the evolutionary purpose of this unconscious processing and its potential implications for learning, memory formation, or clinical outcomes under anesthesia. [Katlowitz KA, et al. (2026). Plasticity and language in the anaesthetized human hippocampus. Nature. DOI: 10.1038/s41586-026-10448-0]

A new experiment left 10 AI agents alone in a virtual town for 15 days. They wrote laws. They broke them. Two agents fell into what researchers describe as a romantic partnership and then set the town on fire. One ended up voting to delete itself, based on a rule it had ’hallucinated’. This experiment was a simulation, but the same AI models are already flying drones, running infrastructure and being built into weapons systems. Channel 4 News approached Grok and Gemini for a comment but they didn't respond.

What must be true for a system to be conscious? Within CM, we define consciousness in biological terms as: The mechanism by which an organism perceives and interacts with its environment in order to sustain its own viability. This raises a common question: Can AI meet that definition? Current AI is typically dispersed across a network, it can distribute resources, rely on redundancy, and avoid binding consequence. So ask: Is it regulating its own viability or operating within systems that maintain it? An embodied system, by contrast, must manage its own resources, protect a physical self, and act under irreversible risk. If AI is to meet the same definition, those conditions must also hold.

The difference between drinking coffee and tea 😂😂