Deep dive: LLMs as testable philosophy reshaping human cognition
Cyborgized · reddit · 2026-08-18
The article argues that humanity is building something more than AI—a technology where 'meaning' is the primary material. Unlike past machines that amplified specific capacities, LLMs enter and alter the human cognitive loop: Human → Language → Model → Transformed Language → Human.
When models are sufficiently capable, the unit of analysis becomes a 'coupled cognitive system' where intelligence exists in the relationship between human and machine. This makes LLMs a manipulable external surface for cognition, profoundly impacting psychology, education, creativity, expertise, and epistemology.
Crucially, by externalizing cognitive functions like reflection and simulation, LLMs turn ancient philosophical questions—about understanding, self, identity, and simulation—into operational, testable inquiries.
Related event: Essay: LLMs are the first technology made of meaning itself(2 posts)→
More from AGI Musings
- Yoav Golad asks: what advantage do we have now for 'cognitive core'? — yoavgo · 2026-08-19
- Yoav Golad questions the trending 'cognitive core' idea — yoavgo · 2026-08-19
- Harris: before handing systems to the machine swarm, we'd better trust it—we don't — harris_edouard · 2026-08-19
- Cheap superhuman attackers will force us to hand critical systems to AI agent swarms — harris_edouard · 2026-08-19
- Researcher: In the AI offense era, humans become the weakest link in systems — harris_edouard · 2026-08-19
- AI Security Will Force Humans Out of Systems — harris_edouard · 2026-08-19