An Ethical Framework for Artificial Intelligence and Sustainable Cities
Round 1
Reviewer 1 Report (Previous Reviewer 1)
The topic of the article is very interesting and it raises the reader's interest.
However, the results are a little disappointing.
The question is, "what's new"? What is different and innovative compared to the existing literature? Because no recognized method was used for the development of the conceptual framework such as: a systematic literature review, bibliometrics, meta-review or meta-analysis.
How do you justify the selection of references that were used?
The article needs a much greater degree of depth and a better alignment between literature-methodology-results.
As long as this alignment is not clear and the article does not present innovative results, it will be difficult to consider it for discussion and publication.
The author(s) made a good effort, but insufficient given the current state of maturity.
Author Response
Please find the response in the attached report document, thanks.
Author Response File: 
Author Response.pdf
Reviewer 2 Report (Previous Reviewer 2)
The paper is very well written and addresses an actual and interesting topic.
Author Response
Dear reviewer, many thanks for the review and positive view on it. We have improved substantially the manuscript with a major revision.
Reviewer 3 Report (New Reviewer)
A brief summary (one short paragraph) outlining the aim of the paper, its main contributions and strengths.
The paper aims to present an ethical framework for sustainable cities based on SGD-11, discuss the features of such framework, the relations with other SGD and how they can help each other to be achieved through combining ethical principles with digital innovation.
General concept comments
Article: highlighting areas of weakness, the testability of the hypothesis, methodological inaccuracies, missing controls, etc.
Wekaness: the title gives the reader an idea that is supported in the abstract, but the introduction section presents more general knowledge on Ethics and Philosophy than the connection with the topic of sustainable cities. It is not clear why the authors introduce general concepts and why to select SGD. More background on SGD in the Introduction, especially SGD-11 can help understand much better the contribution of the paper and strengthen its relationship with approach of this framework.
Review: commenting on the completeness of the review topic covered, the relevance of the review topic, the gap in knowledge identified, the appropriateness of references, etc. These comments are focused on the scientific content of the manuscript and should be specific enough for the authors to be able to respond.
The authors might mean to report a variate of ideas but in doing so, there are some arguments that seems contradictory (all SGD-11 targets can be achieved with AI – line 227, however in reference [37] there are some targets that are inhibit by AI), too praised (lines 58, 67, 84), or that can be questioned (pandemic slowed down 4th industrial revolution or it actually favored it – line 84). It is mentioned that digitalization is lead by technical and industrial principles, but it is not clear why fragmented principles should be denoted as practical guidelines (line 124). Authors mention risks and concerns about AI-based ethical approach, however they do not provide more details, specially to analyze them along with SGDs and targets. Figure 1, which aims to show the proposed framework and the relations of SDGs, lacks more detailed description. For example, the relations (represented with arrows) between SDGs.
Is the manuscript clear, relevant for the field and presented in a well-structured manner?
The manuscript raises a relevant domain such as sustainable cities, its relationship with SGD-11 and the need of framework that combines ethics and computational approaches to measure how the principals are achieved. It also mentions an interesting idea as to balance human-centered and collectiveness-oriented approach. However, the structure of the paper needs to be improved. The authors mention interesting concepts that could be expanded even more, such as Collective Intelligence and sets of ethical principles such as protective, actionable and projection or another set of 5 (transparency, justice and fairness, non-maleficence, responsibility, and privacy). These concepts seemed to be only listed and not completely analyzed and integrated with the framework.
Are the cited references mostly recent publications (within the last 5 years) and relevant? Does it include an excessive number of self-citations?
Arguments presented in line 76 are self-citations, elaborate a little more what is the relation of this paper’s argument with the citations and if all of them are appropiate. More computational approaches and models references can help strengthen the claim that computational social and AI-based perspective can help achieved the SGD-11-based ethical framework for sustainable cities and how. It is suggested to review examples provided in "Tackling climate change with machine learning" (Rolnick et al.), specially for SGD-7, SGD-13 and SGD-15. Two citiations seemed to be repeated- Ciudades y digitalización: construyendo desde la ética (lines 320 y 333).
Is the manuscript scientifically sound and is the experimental design appropriate to test the hypothesis?
More details should be provided and consider commenting if the framework has been adopted or is being analyzed by any level of government, or if some potential users have analyzed or review it. On the other hand, the authors miss to provide more details on how the SGD related to SGD-11 were selected.
Are the figures/tables/images/schemes appropriate? Do they properly show the data? Are they easy to interpret and understand? Is the data interpreted appropriately and consistently throughout the manuscript? Please include details regarding the statistical analysis or data acquired from specific databases.
Describing Figure 1 and relations between different SDG and SGD-11 can help strengthen the proposed framework, especially arrows of interactions/relations.
Are the conclusions consistent with the evidence and arguments presented?
The conclusions introduce more details about Collective Intelligence than the previous parts of the paper, when it might be the other way around (more extended first and the summarized in the conclusions). The idea to combine ethical principles with Collective Intelligence can be further analyzed and described to make a stronger argument. It also talks about how different cultures can adopt computational approaches in different levels. However, this argument was not presented previously.
Author Response
Please find point-by-point response in the report document
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report (Previous Reviewer 1)
I appreciated the authors' effort in revising the article. However, I was expecting another kind of answer. I do not believe that justifying the selection of references using experience and individual education is an adequate argument.
There is relevant literature within the scope of the United Nations SDG and IA very relevant in this scope. Like, for example, the article you already cite: Vinuesa, R., Azizpour, H., Leite, I., Balaam, M., Dignum, V., Domisch, S., ... & Fuso Nerini, F. (2020). The role of artificial intelligence in achieving the Sustainable Development Goals. Nature communications, 11(1), 1-10.
But like the article above, there are others, such as: https://doi.org/10.1002/sd.2048
What I would have liked to have seen was argumentation based on existing literature, even if it is not specific in the realm of AI Ethics.
Although it is a study that follows a philosophical method, there are certain methodological principles that should be followed. With, for example, the objective definition of a research question. The alignment of this research question with the methodology used. How the results come to fill the gap that exists in the literature, and so on. Minor issue is the reference section, as some of the references are incomplete (vide sources from the internet, consultation date must be included).
According to my analysis, the article does not meet the conditions for publication. Although I recognize that the authors' effort has merit.
For this reason, I will inform the editor that if the other reviewers approve the article, I will not argue against it. Although I still believe that a philosophical study should follow the academic principle (e.g. research question, in-depth description of the methodology used, etc.).
Good luck with your research.
Author Response
Thanks for the valuable review, please find the attached report. We believe we have fixed the most relevant review comments.
Author Response File: Author Response.pdf
Reviewer 3 Report (New Reviewer)
Authors respond that their paper follows a philosophical approach. If so, there is more need to connect ideas among sentences and paragraphs of a sections to strengthen their claims. It seems like the authors are sure how they are related, but they need to explicitly mention when they are using an argument to support their claim, what parts of the papers are expository or their own contribution/reasoning (given what arguments), or why to mention certain ideas and references. For example, authors can extend how “…ethics was already related to the cosmical and urban order” in line 40 to really highlight why it is important to mention classic, enlightenment and modern eras. Idea in 63-64 is incomplete. Line 66 sounds weird, do the authors mean is responsibility of SDG? In line 81, why can the authors claim that? Can the authors provide some brief comparison with other tools to drive social development? Idea in lines 66-67 is interesting but seems not to be developed or justify. In line 202, Is this a list of what reference [27] mentions or an argument elaborated by the authors? Anyways, some examples about these features can strengthen the arguments so it does not seem only a list of desired features but a more developed and linked arguments. In line 212, can the authors discuss a more stronger argument on why to dismiss developing countries which might also be the most vulnerable populations, which the authors claim the framework might to work for. Section 3.2. looks more consistent, consider mentioning the name of the SDG again so the reader does not have to go back and for to the figure, and again the same review, connect better the ideas developed in each, i.e. sometimes metaverse seems to be mention randomly, justify and relate more.
Still concepts such as collective intelligence are more developed in the conclusions than in the main sections of the paper, and summarize in the conclusions.
I still believe the topic is very interesting. Although the structure and the way it is presented is still confusing. From what I understand, you still can extend your work with more pages in case you need it to address these comments, but consult with the editors.
Author Response
Thanks for the valuable review, we have addressed the reviewer's comments.
Author Response File: Author Response.pdf
Round 3
Reviewer 1 Report (Previous Reviewer 1)
I think the article is much better now. The authors have made an effort to include what is in my view the most relevant literature. Therefore, I believe that the article presents a more solid theoretical support and that it meets the conditions to be published. Congratulations on your effort and resilience.
Author Response
Many thanks for the positive last review. We are very happy the reviewer liked our manuscript and work. Thanks!
Reviewer 3 Report (New Reviewer)
In general, the structure of the paper has improved. The goal of the paper is much clear now: design a framework that raise the individual-level ethics considering a broader and more general and collective approach to establish ethical principles based on SDGs, especially SDG-11 on sustainable cities and communities. Moreover, it is much better understood the reason of mentioning the different ages/schools of philosophy to raise the point that now a days, it is becoming more individual and subjective.
A suggestion is that when mentioning the need of a framework that learns from evidence, the authors could mention some of the evidence that will be need in each of the SDG related to SDG-11 or its targets.
On the other hand, section 3.2. where authors explain relations between other SDG and SDG-11, most of these descriptions repeats a same argument of need and request of AI-based, collective intelligence, etc. It would be more interesting to provide more examples of what kind of evidence, or information, is needed to encourage certain behavior, or avoid specific risk examples/problems related to that SDG and sustainable cities. Moreover, these examples can be categorized as protective, actionable, and projection.
Finally, authors must emphasize how their work is different from others analyzing general ethics in AI and why SDG-11 conveys really the other selected SDG. Consider analyzing:
- AI: a key enabler of sustainable development goals. A Khamis, H Li, E Prestes… - IEEE Robotics & …, 2019 - ieeexplore.ieee.org
- AI ethics for sustainable development goals. AM Astobiza, M Toboso, M Aparicio… - IEEE Technology and …, 2021 - ieeexplore.ieee.org
Other minor improvements:
Line 75, when introducing SDG-11 for the first time, it would be much helpful to introduce its name.
Line 100, certainly there has been several benefits from digitalization, but it also has presented challenges and risks, evaluate if to start with “Fortunately”. Perhaps, contrast benefits and disadvantages briefly, and how the former outnumber or is preferred compared to last. Something similar happens in line 292 when referring to Metaverse.
Line 112-113, any reference, or example of this statement?
Line 125, sentence is not clear
Line 165, authors mention that ethical principle must be revised, are they revising them?
Line 199, “have been proposed”, in this work? Mention explicitly
Line 212, can you provide some examples of this lack of technology governance and mechanism to evaluate digitalization? There exit some as Global Connectivity Index de Huawei, and Global Competitiveness Index (GCI).
Line 234, it would be interesting to know some examples of these advances on computational social science based on philosophical reasoning.
Line 264, perhaps authors should consider mentioning the reduced scope to developed world in the title and abstract. Although is highly encourage that if looking for a collective and to attend more vulnerable people, developing countries context should be considered.
Author Response
Many thanks for the positive feedback and the detailed review again. Please find the letter attached.
Author Response File: Author Response.pdf
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
The article focuses on an important topic and catches the reader's attention, however, it falls far short of expectations. Essentially, there are two main reasons that justifies rejection:
(1) A methods/methodology section is missing. This issue is what characterizes a scientific article.
(2) I cannot identify any framework. The article is 6 pages long and does not contain the most relevant information. I suggest that authors choose to carry out a systematic literature review, in order to identify the most relevant articles and thus develop a strong and solid framework.
Frankly, the manuscript does not have the expected degree of maturity.
Reviewer 2 Report
The paper introduces interesting questions in ethics. The paper would like to present a status of the art in ethics but does not provide an enough deep analysis.
Conclusions are missing
I would suggest to analyze deeper the problem and improve the presentation.
