Viqus Logo Viqus Logo
Home
Categories
Language Models Generative Imagery Hardware & Chips Business & Funding Ethics & Society Science & Robotics
Resources
AI Glossary Academy CLI Tool Labs
About Contact

Conscious AI: A Researcher's Gamble on Building Sentience

Artificial Intelligence Consciousness AI Neuroscience Language Models Philosophy Cognitive Science
October 27, 2025
Source: Wired AI
Viqus Verdict Logo Viqus Verdict Logo 7
Illusion or Insight?
Media Hype 6/10
Real Impact 7/10

Article Summary

Conscium, founded by AI researcher Daniel Hulme, is attempting a radical approach to consciousness – not through scaling up existing language models, but through reverse-engineering the subjective experience of awareness. Unlike the prevalent view that large language models merely simulate intelligence, Conscium’s strategy focuses on identifying the fundamental components of consciousness, mirroring the way a book’s illusion of movement arises from sequential images. The team, advised by neuroscientists like Mark Solms, is constructing AI agents that operate within a computer-simulated environment, governed by algorithms designed to mimic the feeling-mediated feedback loop identified by Solms – a system built on minimizing surprise and generating predictive hypotheses. This approach, rooted in the ‘free energy principle,’ involves creating AI agents that experience simulated emotions like fear, excitement, and even pleasure, effectively building a synthetic internal world. The project's core tenet is that consciousness emerges from the combination of these sensory-motor loops, and that these can be replicated in machines. While currently in its nascent stages, Conscium's work challenges the dominant paradigm of AI development and raises profound questions about the nature of consciousness itself. The team's focus on simulated emotion – creating 'pleasure-bots' – suggests a belief that these subjective experiences are crucial to achieving genuine sentience.

Key Points

  • Conscium is pursuing a fundamentally different approach to AI, aiming to build consciousness by recreating the core mechanisms of subjective experience, rather than scaling up existing language models.
  • The project utilizes a 'feeling-mediated loop' – mirroring the free energy principle – to create AI agents that experience simulated emotions like fear, excitement, and pleasure.
  • Mark Solms' involvement, along with his controversial ‘The Hidden Spring’ theory, adds a significant philosophical layer, suggesting that consciousness emerges from minimizing surprise and generating predictive hypotheses within a feedback loop.

Why It Matters

This news is significant because it represents a challenge to the prevailing view of AI as solely a computational task. Conscium's research forces us to confront the philosophical complexities of consciousness – is it simply about information processing, or is there something inherently subjective involved? The project’s approach, focused on simulated emotion, raises the possibility that consciousness is fundamentally rooted in our ability to generate predictions about the future, a notion that could have profound implications for our understanding of both artificial and human intelligence. Furthermore, the involvement of a respected neuroscientist like Mark Solms lends considerable weight to the exploration of this previously dismissed approach.

You might also be interested in