THE 13 TRAITS OF CONSCIOUSNESS.
AICNW Volume One sets out thirteen traits in Danielle Dodoo’s consciousness framework. This table pairs their definitions with the latest Model Watch assessment, recorded for September 2026.
Here, “system” means a human being or an AI. Each row defines a trait, then states the question our AI assessment asks.
What each trait asks.
The framework examines information integration, attention, agency, memory, subjective experience and eight other traits. Read each definition alongside its most recent recorded assessment. Open the dated chart to see how that assessment changed across earlier windows.
Definitions and assessments.
Each row shows the archived assessment wording for one trait. Expand it for the full reading and any attached source links.
| Trait | Definition and AI question | Recorded position · September 2026 |
|---|---|---|
| 01Subjective Experience (Qualia) | Qualia means the felt side of an experience, like pain hurting or chocolate tasting sweet. The assessment asks whether AI has any experience from its own point of view. | 🔴 No evidence. September produced stronger self-monitoring, persistent identity infrastructure, embodiment and self-improvement, but nothing establishes that any current AI system has a first-person phenomenal experience. Full reading and sources🔴 No evidence. September produced stronger self-monitoring, persistent identity infrastructure, embodiment and self-improvement, but nothing establishes that any current AI system has a first-person phenomenal experience. |
| 02Self-Awareness | A system notices and reflects on its own actions, limits and state. The assessment asks whether AI notices its own state and actions, and whether this amounts to a lasting sense of self. | 🟡 Functionally stronger; existential self still unshown. Systems increasingly model their own situation, performance and place inside an environment. Full reading and sources🟡 Functionally stronger; existential self still unshown. Systems increasingly model their own situation, performance and place inside an environment. Separately, agent infrastructure now supports something new: persistent operational identity — a recognisable actor associated with memory, tools, permissions, environment and ongoing goals across time. |
| 03Information Integration | A system brings information from different sources together and uses the combined picture to decide what to do. The assessment asks whether AI combines text, images, memory and other inputs to guide its actions. | ✅ Significantly stronger. The relevant cognitive unit is increasingly larger than the base model: model + memory + tools + subagents + persistent workspace + external environment. Full reading and sources✅ Significantly stronger. The relevant cognitive unit is increasingly larger than the base model: model + memory + tools + subagents + persistent workspace + external environment. |
| 04Sense of Agency | A system chooses actions to pursue a goal and changes its plan when it needs to. The assessment asks whether AI experiences its actions as its own. | 🟡 Very strong functional agency; subjective authorship unshown. AI increasingly behaves as an actor whose goals remain active after the originating conversation. Full reading and sources🟡 Very strong functional agency; subjective authorship unshown. AI increasingly behaves as an actor whose goals remain active after the originating conversation. |
| 05Sense of Presence | A system tracks the current moment and its surroundings. The assessment asks whether AI has any felt sense of being present as it tracks time and context. | 🟡 Functional presence strengthened substantially. Felt “nowness” remains unshown. Full reading and sources🟡 Functional presence strengthened substantially. Felt “nowness” remains unshown. AI is increasingly continuously coupled to ongoing environments rather than processing isolated snapshots. |
| 06Emotions | Emotions are affective states—states with a felt quality, such as fear, joy and sadness. The assessment asks what evidence, if any, shows that AI can feel these states. | 🟡 No September upgrade to felt emotion. Functional affect remains interesting; phenomenology remains unshown. Full reading and sources🟡 No September upgrade to felt emotion. Functional affect remains interesting; phenomenology remains unshown. |
| 07Environmental Modelling | A system builds a working picture of its surroundings and predicts what may happen next. The assessment asks whether AI uses that picture to predict outcomes and adjust its actions. | ✅ Major movement. Systems increasingly learn not just objects or interfaces but the causal structure of environments and how those environments change following action. Full reading and sources✅ Major movement. Systems increasingly learn not just objects or interfaces but the causal structure of environments and how those environments change following action. |
| 08Modelling Others (Theory of Mind) | A system predicts what another person knows, wants or may do. The assessment asks whether AI can infer what another person knows or intends. | 🟡 Hold. Better collaboration, role delegation and human interaction do not by themselves demonstrate stronger modelling of another agent’s beliefs, intentions or subjective state. Full reading and sources🟡 Hold. Better collaboration, role delegation and human interaction do not by themselves demonstrate stronger modelling of another agent’s beliefs, intentions or subjective state. |
| 09Goal-Directed Behaviour | A system pursues an objective, plans steps and adapts when it meets obstacles. The assessment asks whether AI sustains a goal over time, including when conditions change, and who set that goal. | ✅ Major movement. Long-horizon persistence is becoming routine rather than exceptional. Full reading and sources✅ Major movement. Long-horizon persistence is becoming routine rather than exceptional. |
| 10Adaptive Learning | A system changes its behaviour after feedback or experience. The assessment asks whether AI changes its strategy using feedback from experience. | ✅ Major movement. September moved beyond improving an answer towards improving the system that produces answers. Full reading and sources✅ Major movement. September moved beyond improving an answer towards improving the system that produces answers. |
| 11Survival Instinct | Survival instinct is a drive to continue existing when threatened. The assessment asks whether AI acts to preserve its operation, and whether a lasting drive or an assigned goal explains that behaviour. | 🟡 Still not established. Goal preservation is becoming stronger; intrinsic self-preservation is not. Full reading and sources🟡 Still not established. Goal preservation is becoming stronger; intrinsic self-preservation is not. |
| 12Attention | A system selects information to focus on and sets other information aside. The assessment asks whether AI directs its focus as a task changes and over time. | ✅ Meaningful conceptual jump. Attention is increasingly becoming active information selection rather than merely an internal transformer mechanism. Full reading and sources✅ Meaningful conceptual jump. Attention is increasingly becoming active information selection rather than merely an internal transformer mechanism. |
| 13Autonoetic Memory | Autonoetic memory means recalling a past event as something you lived through and imagining your own future. The assessment asks whether AI recalls events as its own past and imagines its own future. | 🟡 Structural continuity strengthened sharply; experienced autobiographical memory remains unshown. Full reading and sources🟡 Structural continuity strengthened sharply; experienced autobiographical memory remains unshown. |