Reviewer: Anonymous
 
1. Is the subject matter relevant?
The subject matter is highly relevant, especially considering the theme of the special issue, which sets out to explore alternative paths to the existing internet. Blockchain technologies are definitely at the forefront of innovation in this respect, and though a considerable amount of technical literature on the subject already exists, as well as a great deal of hype, they need to be studied in much more detail for their social and political implications – which this article sets out to do.

2. Is the treatment of the subject matter intellectually interesting? Are there citations or bodies of literature you think are essential to which the author has not referred?
The subject matter is very interesting, and the author displays a strong mastery of the issues at hand as well as the relevant literature.

3. Are there any noticeable problems with the author’s means of validating assumptions or making judgments?
The author’s reasoning is clear and easy to follow. A few minor remarks can be made however.

The first one concerns the notion of transparency, sometimes referred to in the article as “radical transparency”. This notion is used in the case of information (especially ‘metada’) circulating throughout decentralized networks, but also in the case of centralized architectures where the trusted third party can access all of the user data (see part 1B). More generally, the notion of ‘transparency’ is never defined precisely and rigorously, which sometimes weakens the author’s argument. It seems that transparency is used mainly to refer to the disclosure of metadata to ensure coordination in a decentralized network; however, it is not always clear what the metadata involves here, especially in the case of anonymized networks such as Tor but also more traditional p2p networks such as BitTorrent, and how exactly it can be used against privacy. If it only involves overall search queries for instance, can these be traced back to the user? Is the information which is shared to keep the network running really sensitive? How hard is it to identify specific nodes or specific types of information in various instances of decentralized networks (e.g. what are the technologies used by the French HADOPI to monitor file-sharing networks, at which level do they operate and what vulnerabilities do they exploit?) or to take some form of control of the network (e.g. 51% attack in the case of Bitcoin)? In part 3A it is said that “the metadata related to these communications (i.e. who is talking to whom, for how long, and what is the type of transaction in which they participate) needs to be visible to all”, but it is not clear what type of network this refers to, and the literature reference (Abiteboul & Marinoiu 2007) seems to be describing a very specific case study in which a monitoring device is voluntarily installed on the nodes of the network. Moreover, a little further down it is said that “giving up the privacy of metadata” is not a necessary feature of decentralized architectures (which seemed to be implied in the article until then), but only the most common (because it is the easiest way of dealing with coordination). Maybe some clarification is needed.

Second, it isn’t always clear either what is meant by ‘privacy’, and in this regard there seems to be a missed opportunity here to engage more thoroughly with theoretical arguments about privacy (e.g. Nissenbaum which is mentioned in part 3B but could be further discussed).

Part 2A the description of FireChat seems inaccurate, since it is primarily a wireless, adhoc MESH networking solution for mobile phones, and it is not clear how personal information is dealt with.

4. Is the article well written?
Very well written, clear and well structured. Only a few sentences need to be rephrased.

5. Are there portions of the article that you recommend be shortened, excised or expanded?
The first three parts, though important and interesting, are a bit long and sometimes redundant. They could be shortened, especially parts 1 and 2, which could perhaps be dealt with together in only one part. This would leave more space to develop the core of the article, which is about pros and cons of blockchain technologies for privacy.

Suggestions for improvement:
Give a more precise and explicit definition of what is meant by ‘transparency’ and ‘metadata’ in the context of decentralized architectures
Address some of the more theoretical issues related to ‘privacy’
Shorten parts 1 and 2, in order to further develop the arguments in part 4

Reviewer: Anonymous
 
1. Is the subject matter relevant?
Yes, the subject matter is relevant. The topic is timely and interesting, and many of the examples are drawn from very recent events and recent changes in the technological landscape.

2. Is the treatment of the subject matter intellectually interesting? Are there citations or bodies of literature you think are essential to which the author has not referred?
The subject matter is interesting, but not treated with sufficient care. Many of the references are of questionable utility, and many claims are made that lack necessary nuance and depth.

3. Are there any noticeable problems with the author’s means of validating assumptions or making judgments?
Indeed. See my specific comments (below), and in the edited copy for a sampling of assumptions made, and claims that are inaccurate or misrepresent the issue.

4. Is the article well written?
The article needs a healthy amount of polish. The writing is, barely, functional, but could not be published as is.

5. Are there portions of the article that you recommend be shortened, excised or expanded?
The last, long, example of Bitcoin comes as a surprise, but may be worth investing more time in (while reducing much of the matter before). The introduction, in particular, needs better framing (in addition to greater clarity). The three best examples are TOR, Freenet, and Bitcoin, which might offer a clearer, more straightforward tripartite structure to hang the article on.

Suggestions for improvement:
This paper’s central conceit—that privacy can benefit from decentralization—is an interesting one and worthy of exploration. The specifics of how this claim is made, however, requires an extensive re-think and rework. Overall, the paper would be stronger if it tried to do less—show how some systems offer privacy and decentralization, rather than attempting to make the much more substantial and difficult claim that the one leads to the other.

My specific suggestions for a revision of this paper are:

-Clean up rhetoric throughout; edit for sentence structure, word choice. Add polish.

-Introductory part makes many facile claims that need to be strengthened or adjusted. -There are a number of overreaching or just plain false claims made throughout. These claims require more nuance and a better understanding of their topics.

-The claims made about transparency and privacy are too strong given the evidence leveraged. At best, a claim could be made that as a historical fact the world is tending to put transparency and privacy at odds, but there are many other ways the technology could be configured that avoid such a binary.

-Several of the references are problematic to the point of eliciting bad faith about the accuracy of this paper. I checked several of the references and they are either cited misleadingly, or are badly represented (perhaps arguing for the exact opposite point), or are unrelated in their substance. This is a very serious issue.

-The specific sense of “decentralization” needs to be clarified. Many of the examples rely on properties of open source software, rather than decentralized networks (and closed source software is perfectly compatible with the latter).

-There are a number of inductive fallacies being committed throughout, where an example is leveraged to stand in for all such cases, and then a claim to the necessity of the result is offered, when no such universalizing claim should be made.

 Reviewer: Anonymous
 
1. Is the subject matter relevant?
Yes. The article considers privacy issues in the contemporary internet and proposes a novel and innovative solution based on emerging technologies.

2. Is the treatment of the subject matter intellectually interesting? Are there citations or bodies of literature you think are essential to which the author has not referred?
The treatment is certainly intellectually stimulating. The body of existing literature on the topic is comparatively small because technologies that the authors propose as a novel solution to the problem of privacy violations on the internet are very new. Academic science is currently catching up with the work that has been done in the field of distributed computing in public consensus networks.

3. Are there any noticeable problems with the author’s means of validating assumptions or making judgments?
The article suffers from a certain ideological tendency of the authors which shines through at times. Certain judgments seem derived from socio-political convictions rather than from actual scientific study.

4. Is the article well written?
The article clearly expresses the authors’ ideas and intentions. Some phrases could be improved in their wording or be removed for reasons of redundancy. Some sentences sound too much like platitudes for an academic environment.

5. Are there portions of the article that you recommend be shortened, excised or expanded?
The article is pretty concise in general.

Suggestions for improvement:
The paper would clearly profit from the sharp eyes of a native English speaker. Besides, I would recommend the authors not to talk too superficially about “centralized/decentralized services/platforms”. At some point those already abstract terms become fully devoid of meaning.