Cognitive Systems and Artificial Consciousness: What It Is Like to Be a Bat Is Not the Point

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear authors,

The article is well designed, however some flaws remain that need to be fixed. In particular:

The introduction section does not show clearly the practical need of the study: which is the gap it intends to fulfil? What is the main goal behind the study? What contribute to society in general? Also, you need to display its structure in terms of how the article is organized.

In the Materials and Methods you need to state crystal clear what the adopted methodology is: is it Ground Theory, Experimental, etc.

Section 3 should be renamed to ‘Literature Review’ or ‘Theoretical Framework’ and repositioned following the Introduction, as it widely argues upon theory and concepts.

The section 4 resembles a literature review on the subject: consider change methodology.

Conclusion neither present limitations, nor theoretical / practical consequences: consider adding two subsections. In particular, limitations must be included either in the main text or in a ‘Limitations’ section:

Implementing a functionally similar process in a cognitive architecture doesn't essentially mean the architecture encircles the theoretical concept: the risk of confusing simulation with explanation. Moreover, implementation does not equal realization: Just because an architecture has a module labeled “monitoring” does not mean it implements the cognitive or phenomenological function in question. Artificial systems may implement accessibility of information and monitoring modules, but compare these with true conscious access or introspective knowledge is a category error.

Also there is the risk of circularity—architectures are judged good because they fit the theory, and the theory is supported because architectures might implement it. In one hand, access mechanisms are different from access consciousness: A system may make internal representations available for decision-making or reporting, but that doesn’t show access consciousness as defined by Ned Block, in particular when no subjective experience is present. On the other hand, monitoring differs from self-awareness: many AI systems implement performance monitoring, or error correction. However, such mechanisms are mechanistic, not reflective in the metacognitive sense.

Finally, human metacognition involves high level double loop thought, while most artificial systems lack the architecture to reason about their own mental states in a context-sensitive way of complex social and developmental dimensions.

Those issues must be developed.

 

Regards,

Author Response

We are grateful for your thoughtful review and constructive feedback, which have helped us significantly improve the clarity and scholarly value of our manuscript. Please find our point-by-point responses below.

1. Practical Need, Research Gap, Main Goal, and Societal Contribution (Introduction)

We have revised the introduction to make the practical need, the main conceptual gap, and the societal motivation of the study more explicit. The introduction now clearly states that the absence of operational criteria for implementing functional consciousness in artificial agents constitutes the central research gap. We also emphasize the practical motivation to enable supervision, explainability, and auditability in cognitive systems, and underline the potential societal contribution: strengthening transparency and public trust in autonomous technologies. In addition, a closing paragraph has been added to the introduction, outlining the overall organization and structure of the article to guide the reader.

2. Methodology: Explicit Statement and Typology (Materials and Methods)

The Materials and Methods section has been expanded to explicitly state that the study employs a structured systematic review methodology, guided by PRISMA principles, with defined inclusion and exclusion criteria. We detail the databases consulted (SCOPUS, Web of Science), the search strategies used, and the multi-phase analysis protocol, ensuring transparency and reproducibility.

3. Section 3: Renaming and Repositioning as “Literature Review” or “Theoretical Framework”

We appreciate your suggestion. In the revised manuscript, we have added a clarifying paragraph at the start of Section 3 to justify why, due to the conceptual and analytic nature of the work, the theoretical framework and the results are necessarily integrated. While we have retained the current numbering for consistency with ongoing related work, the rationale for this structure is now made clear. Should the editorial team require further changes, we are open to adapting the structure as needed.

4. Section 4: Methodological Coherence

We recognize the possible overlap between the theoretical discussion and methodological exposition in Section 4. To address this, we have revised the section to make the transition from theoretical analysis to practical implications more explicit, and clarified how the discussion builds upon the comparative findings. The section now highlights the practical relevance and engineering applications that emerge from the conceptual results.

5. Conclusions: Limitations and Theoretical/Practical Consequences

We have reorganized the Conclusion to address both the principal limitations of the study and its theoretical and practical consequences. Two dedicated subsections are now included: one discussing the main constraints and risks of the functionalist approach, and another outlining the conceptual and engineering implications of the findings.

6. Simulation vs. Genuine Explanation and the Risks of Circularity

A new paragraph has been added to the Discussion section, explicitly addressing the distinction between functional simulation and genuine conscious experience, and the risks of category errors and theoretical circularity. The manuscript now clarifies that while artificial systems can implement modules labeled “monitoring” or “access,” these do not equate to phenomenological or metacognitive states as understood in human cognition. We further stress that such architectures should not be conflated with true conscious access or introspective knowledge, as highlighted by Ned Block.

7. Human Metacognition vs. Artificial Systems

We now address the limitations of current artificial systems in replicating advanced human metacognition. The revised manuscript clarifies that our study does not claim equivalence between artificial and human metacognition, and that the boundaries of advanced self-reflective cognition—such as double-loop learning and complex socio-developmental reasoning—remain open challenges for artificial architectures. These limitations are explicitly acknowledged and presented as avenues for future research.

Once again, we sincerely thank the reviewer for their insightful comments and for contributing to the improvement of our work.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

Thank you for this interesting contribution.

The article deals with the idea of artificial consciousness as focussing on its functionality and ability to work as opposed to the idea of phenomenal consciousness (the subjective experience) and access consciousness (functional processes).

It says modern AI and cognitive architectures attempt goal-oriented, measurable capabilities like reasoning, self-monitoring, attention, memory, and adaptive behavior but not to mimic human subjective experience.

The text encompasses avoidance of practical realizations of these features in several fields, e.g. healthcare, industry, environmental management, and promote the pragmatic approach towards the construction of autonomous, self-evaluating AI systems based on the technical transparency and efficiency of functioning.

Some suggestions are:

• In the introductory stage explain the difference in phenomenal and access consciousness so as to make a clear conceptual foundation to the reader who may be new to the subject.  
• To increase reproducibility and transparency, clarify the description of research methodology, especially standards and method performed in the comparative literature research and functional modules identification.  
• Incorporate additional examples or case studies of the mechanism in which the modules of functional consciousness can (or already are) applied in existing cognitive systems or AI.  
• Enhance the writing and clarity of the English language throughout the manuscript to be able to communicate the complex ideas in the better than in the possible case of professional language editing.  
• Deepen the argument about practicalities and limits of using functional consciousness in real-life AI (such as the ethical aspects).  
• Add an expanded statement of how the prospective functional modules communicate inside of a cognitive framework to obtain an all-encompassing picture of artificial consciousness.  
• It may be useful to include some visual diagrams or flow-charts of how the proposed modules of consciousness interact with each other and how they relate architecturally to the rest of the possible system.  
• Fill in gaps considering the possible criticisms or alternative approaches to the paper on the operational approach to artificial consciousness to enhance the scientific character and equilibrium of the paper.  
• Modernize and expand reference base so that it contains latest news in the areas of AI explainability, AI self-awareness, and AI cognitive architectures so that the source material reflects the state of the field presently.  
• Point out possible directions of further research/open questions generated as a result of the work, in order to provide future work and progress in the field of artificial consciousness.

Kindest regards

Comments on the Quality of English Language

The English language used in the paper is mostly comprehensible and easy to follow, as it puts to paper very complicated concepts regarding the idea of free will of artificial consciousness.

Some remarks are:

• There are sentences which are long and may be simplified to be more understandable and easier to read through.  
• Technical language is used in an appropriate manner but sometimes is insufficient to the reader, who has less knowledge of the field.  
• Obvious slips in grammar and some instances of woodenness are likely to come across here and there and can only confuse the readers and make the text less clear.  
• The changes of sections might be better and more gradual to improve continuity and logical flow of arguments.  
• It would be in the manuscripts interest to have a professional language editing done to smooth out syntax or style and punctuation.  
• To keep the reader interested and highlight the main ideas better, it is necessary to improve the brevity of some of the paragraphs.

In general, the level of English is sufficient enough, although it could be enhanced as a way to more effectively convey subtle ideas and reinforce the punch of a paper.

Author Response

We thank the reviewer for their thorough reading, positive remarks, and thoughtful suggestions, all of which have been instrumental in further enhancing the clarity and scientific value of our manuscript. We address each comment below:

1. Clarify the difference between phenomenal and access consciousness in the introduction for conceptual foundation

We have revised the introductory section to more clearly explain the distinction between phenomenal consciousness (subjective experience) and access consciousness (functional, measurable processes). An accessible analogy has been added to assist readers who may be new to the field, thereby laying a stronger conceptual foundation.

2. Clarify methodology, standards, and methods for reproducibility and transparency

The Materials and Methods section has been expanded and clarified to state explicitly the methodological standards applied, the systematic review protocol, inclusion and exclusion criteria, and the steps followed in both comparative literature research and the identification of functional modules. This revision aims to maximize transparency and reproducibility.

3. Add examples or case studies of application of functional modules in cognitive systems or AI

While this article primarily focuses on conceptual analysis, we have integrated concrete examples and references to practical implementations where appropriate. For instance, we now cite real-world applications in fields such as healthcare, industry, and education, indicating how certain cognitive architectures already incorporate modules like attention, self-monitoring, and reasoning.

4. Enhance clarity and quality of English throughout the manuscript

The entire manuscript has undergone careful revision to improve clarity, grammar, and readability. Long sentences have been shortened or restructured, transitions between sections have been made more gradual, and technical terms are now more clearly explained when first introduced. Where necessary, external language editing tools have been used, and peer feedback has been incorporated to further polish the academic English.

5. Deepen the argument regarding practicalities and limits, including ethical aspects

A dedicated paragraph has been added in the Discussion to address both the practical benefits and the inherent limitations of applying functional consciousness in real-life AI systems. Ethical implications, such as responsibility, transparency, and the societal impact of autonomous, self-evaluating agents, are now explicitly discussed.

6. Expand on communication among functional modules in a cognitive framework

The revised text now contains an expanded section that explains how the prospective functional modules interact within an artificial cognitive framework. A narrative example illustrates the typical flow of information among modules such as attention, working memory, reasoning, and monitoring, providing a more comprehensive perspective on the internal dynamics of artificial consciousness.

7. Include visual diagrams or flow-charts

In response to this suggestion, a block diagram (Figure 1) has been added to the manuscript. This figure visually depicts the layered organization of the proposed cognitive architecture and clarifies the interrelations among the functional modules discussed in the text.

8. Address possible criticisms and alternative approaches to the operational model

The revised Discussion now includes a section acknowledging criticisms and alternative theoretical perspectives regarding the operational approach. We explicitly discuss risks such as circularity, category errors, and the limitations of functionalist models compared to alternative views on artificial consciousness.

9. Update and expand reference base with recent advances

The bibliography has been updated and expanded to include recent, high-impact sources (2020–2025) on AI explainability, self-awareness, and cognitive architectures. This ensures that the references reflect the current state of the field and support the arguments presented throughout the article.

10. Indicate directions for further research and open questions

A final section now explicitly highlights open questions and avenues for future research, including empirical validation of proposed models, exploration of social and ethical ramifications, and the need for further development of adaptive and metacognitive mechanisms.

Comments on the Quality of English Language

We thank the reviewer for these remarks. The manuscript has undergone a thorough revision for language quality, grammar, and clarity. Sentences have been simplified where possible, and paragraphs have been made more concise. Technical terminology is now introduced with brief explanations for non-specialist readers. Additional professional editing has been applied to ensure smooth syntax and logical flow, improving both continuity and the overall readability of the paper.

We appreciate the reviewer’s guidance, which has greatly contributed to strengthening both the content and communication of our research.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The article is fit for publication