AI-Driven Productivity Gains: Artificial Intelligence and Firm Productivity
Round 1
Reviewer 1 Report
The authors prepared an article on gains due to AI advancements and deployments. The article seems useful. However, following concerns may be addressed:
1. The abstract may be rewritten to highlight AI deployment in specific fields.
2. The test case of any area may be included to reflect how AI has changed the picture in that area.
As far as English is concered, the authors are suggested to go through the article for minor English corrections.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Dear Authors and Editor!
After analyzing the article, I make some notes:
The title is objective. The topic is interesting, even more so at this time when artificial intelligence has attracted more attention, in various aspects, such as academic, security, industrial, among others.
Abstract: I missed the methodological procedures.
Introduction: well contextualized, it presents the objectives, justifies the study and brings its contributions.
Literature review: presents three topics relevant to the content, they are objective and with updated references.
Data, Variables and Identification Method: described in three parts, it clearly presents the procedures, the dependent and independent variables, how companies were selected and the data and analysis methods used.
Empirical Results: The beginning gets a little repetitive with the methodology. At most, it has a strong statistical description and the data are presented objectively.
Mechanism and Heterogeneous tests: complementing the previous item, it emphasizes the statistics part to show the robustness of the data used in the research; I think it could be next to the results item.
Conclusions: adequate with what was presented in the results, presents the limitations of the study; could include suggestions and future studies, although the limitations of this study can be used as inspiration.
* The negative point of the article is that it lacks a discussion of the research findings with the supporting literature, thus making the study more academically robust.
* I missed some acronyms being presented before being used, because the article may reach people who may not know them.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Submitted article is very well organized and logical. The main research question addressed was impact of AI on firms productivity growth. For which huge data was collected i.e., from 2899 manufacturing firms with sample size of 20,904 which seems suspicious and needs to ensure its credibility.
Abbreviations used should be explained in the sentences where they first appear not in the later paragraphs, i.e., TFP in line 11 but explained in line 268 (Total factor Productivity) and so on, like CSMAR in line 210, IFR in line 218.......
The topic is relevant for industry 4.0 concept to enhance its usability.
Conclusions are supported by the results of the research.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 4 Report
Authors formulated the thesis on impact of AI on Firm Productivity Growth. That thesis seems to be trivial, therefore I would expect systematic literature survey if it (as above) is to be the research study.
Unfortunately, the literature review for that study is rather poor. Authors should take into account that AI has been developed for many years. Hence the analysis of its impact was the subject of study many years ago, not just now.
Authors are expected to expand the acronyms e.g., TFP , IFR, even if they use it as a keyword.
In the title you have impact od AI on productivity’s growth, in abstract impact of AI on productivity, sorry, it is not the same and that interpretations are to be explained.
The sentence on AI impacts on productivity looks like hypothesis , therefore I ask about falsification,
Would you consider as possible that AI has no impact on productity. If it is possible so the next question is why AI is developed if it does not increase productivity.
The discussion on what AI is seems to be very poor, I would recommend authors to read Herbert Simon, the scope of AI solutions is wide and as such should be discussed.
Answer is needed how Authors understand productivity and which particular variable is responsible for that.
Authors declared they found “the impact of AI on the structure of labor demand as well as AI increases 443 firms' demand for overall labor while also increasing the demand for high-skilled workers.”
I’m not sure if authors verified what they wanted.
Lack of explanation what software and estimated method were applied.
Lack of explanation how that research work concerns the sustainability issue.
Hypotheses and research questions are not clearly formulated and that explanations should added.
no comments
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 4 Report
In comparison with earlier version, authors have changed the paper towards an essay, they have provided much more descriptive arguments, hence they paper is becoming more persuasive.
My requests:
Please, add references to Solow paradox
Len(MD&A) - What is this function , please, explain
Please, explain the AI impact on productivity and well as on sustainability,
At least , taking into account the literature survey, and because this paper is submitted to Sustainability Journal therefore , the issue of sustainable development supported by AI should be considered.
If you consider that Solow paradox, can you formulate the Solow paradox in discussion on sustainability , and not only productivity?
no comments
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
