4 ms·

First of all, regardless of all the claims made in the document, I find it awesome to see such a gigantic writeup that tries to be at least a bit internally con

First of all, regardless of all the claims made in the document, I find it awesome to see such a gigantic writeup that tries to be at least a bit internally consistent. It reminds me of Bit451 [1][2], although that one was simply insane. So about Cicada: the document quickly notes that the 'human unique identifier' is at the core of the proposal. It is claimed that it "allows us to prevent Sybil attacks and ensure everyone has a voice in the system". It promises to do this with cryptography and 'biocryptics', a word which isn't really that widely used in the research community. It seems to mostly refer to anonymous biometrics authentication systems [3][4]. In that field I can find nothing about preventing Sybil attacks. Additionally, preventing Sybil attacks in a decentralized system, by relying on a biometric measurement that is fully under the control of an attacker seems doomed to fail. No matter how much cryptography you throw at it. There must either be some central party designating trust in a certain (anonymous) biometric, or there needs to be a web of trust that is sufficiently connected and resistant to manipulation. A combination, with good checks and balances and distribution of power, seems to be the way to go most of the time. [1] http://bit451.org/ http://bit451.org/ [2] https://news.ycombinator.com/item?id=8728231 https://news.ycombinator.com/item?id=8728231 [3] https://www.esat.kuleuven.be/cosic/publications/article-1462.pdf https://www.esat.kuleuven.be/cosic/publications/article-1462... [4] https://en.wikipedia.org/wiki/Private_biometrics https://en.wikipedia.org/wiki/Private_biometrics

MarkPNeyer 9y ago
Here's a web of trust design that comes with a conflict resolution mechanism. https://github.com/neyer/respect https://github.com/neyer/respect http://s3.neyer.me/respect-matrix-slides.pdf http://s3.neyer.me/respect-matrix-slides.pdf
ConfucianNardin 9y ago
Their claim/assumption is based in trust of the enrollment system. Check the section on enrollment - this image is especially helpful: https://raw.githubusercontent.com/the-laughing-monkey/cicada-platform/master/images/Enrollment-1.jpg https://raw.githubusercontent.com/the-laughing-monkey/cicada.... The enrollment software is supposed to be verified using "Challenge/Response", which is supposed to ensure the guarantees provided by the HUIDs. Of course, given enough incentive, this can (and will) be cracked...
andrewla 9y ago
Personal identity seems like a generally intractable problem, despite the lofty goals stated in the introduction. And without it, the document is built on a foundation of sand. An argument that I find myself having quite frequently about Bitcoin and pseudonymity is that pseudonymity and concealing identity is not a feature of the system, but rather that defining two-way identity in any sort of useful way is such a difficult problem that solving it would be a bigger achievement than Bitcoin itself. Seeking a biological equivalent to a non-portable key in a TPM is pretty hopeless, and that's what you'd need to get a HUID -- not just a hash fingerprint that anyone can appropriate, but an internal signing mechanism to allow zero-knowledge verification of identity. And the inverse problem, the Sybil problem, is just as intractable, and as far as I know has never really been solved without a centralized issuing authority or reliance on a scarce resource (like proof-of-work, which even there solves only a weak variant of the Sybil problem by conflating the unit of "identity" with a unit of "work").
- Allower 9y ago
  nailed it
- anigbrowl 9y ago
  We know multiple ways to reliably identify someone using biological markers with near-perfect accuracy. Yes, that lack of perfection leaves open the possibility of subversion, but so what? You'll grow old and die waiting for the perfect solution.
  - hueving 9y ago
    And those are not subject to replay attacks?
    - xg15 9y ago
      This seems to me like the whole misadventure of putting finger print readers in door locks and phones: Someone noticed that fingerprints work quite well for identifying people in customs and law enforcement and simplified that thought to "fingerprints can identify people" - ignoring the circumstances of how fingerprints are taken when they work.
laughingmonkey1 9y ago
Thanks for posting this. It's always interesting to see people's reactions to my work. Usually when people do find it though, the series of comments that follow it tend to fall into a few categories. I will paraphrase and give my answer to these questions generically rather than answer every individual question: 1) "I've zeroed in on this one aspect of the system instead of the whole and found a flaw that invalidates the whole thing." In particular this comes up around the subject of an ID. Usually it centers on "biometrics are how the system is secured, those are gameable, this will never work." In fact, the primary method of controlling the spread of fake IDs is the reputation system is NOT the biometrics system which is merely a first barrier to entry. It is the reputation system. And that reputation system is different from rep systems that exist today which basically consist of people voting thumbs up or thumbs down which is not what I describe. Most of the misunderstanding around that is because the original paper was not as clear as it could be on that aspect, so I wrote up a second paper that breaks down the ID system and Rep Bank in much more detail. That paper is here: https://github.com/the-laughing-monkey/cicada-platform/blob/master/Identity-Without-Authority-2017.21.3.BETA.pdf https://github.com/the-laughing-monkey/cicada-platform/blob/... 2) "This will never work because of XYZ reason I just thought up after glancing at this a few seconds." Those questions are always answered in the document. I encourage you to actually read and do so with an open mind. 3) Some comment/focus on whatever my supposed ideology is. People have zero idea what my ideology is and debating something based on whatever you imagine my ideology to be is not very useful. It's a fallacy of composition. Also my ideology can't be deduced from the paper, largely because I don't much like ideologies and I don't really have one. I look at every problem or issue atomically, to be solved on its own merits, thus my beliefs run the gamut. Also, whatever I think about things is pretty much irrelevant. The system stands on its own and it has nothing to do with what I do or don't believe. The only thing I really believe is "locking in open" as Tim Berners-Lee put it. That is: open systems are better and decentralized systems in particular are key to the future. 4) "Direct democracy can't work because people get tired or aren't smart enough to vote on everything." I outline many solutions to this, in particular systems that can vote for you based on answering a series of questions, or systems that work in hybrid mode that allow you to vote on only the things that really matter to you, say abortion and guns, as well as full voting so that you can vote on everything if you are a highly engaged citizen. I also propose an easy solution to the California ballot problem where every stupid thing comes up for vote and where stuff only gets on there based on big money. It routes votes to ever increasing numbers of users and only the absolute most important votes would ever make it a national vote. Lastly the system allows for expert panels to create legislation that is fast tracked by the system. This would be a dramatic improvement on what we have now where unqualified politicians vote on everything from science to military to personal issues. Having groups of competing economists propose bills, or scientists writing science policy would be a major improvement to the current system. Lastly, the paper also outlines a method for breaking down bills into simpler terms for various folks to make it easier to read them and engage with them. 5) "You're an idiot, why even bother doing something like this?" This is utterly pointless commentary that comes from folks who haven't done much and spend most of their life running other people down for trying something new. Not very useful or cool. 6) Not understanding that every single layer of the system expects some level of exploitation and it doesn't matter if you found a few ways to game it. The system always trusts there will be some level of gaming the system. It's expected and dealt with. Central systems are not a panacea to this, despite most arguments thinking that's the case. Their is all kinds of exploitation of centralized IDs for instance. The decentralized ID system will have some level of people hacking it and it really won't matter as long as the system is resilient enough to keep the blast confined and to update itself to deal with new attacks, which the build in security chain combined with the reputation systems deals with very effectively. 7) I get it and how do we build it? Congrats. Join decstack.com and let's talk.

MarkPNeyer 9y ago

Here's a web of trust design that comes with a conflict resolution mechanism. https://github.com/neyer/respect https://github.com/neyer/respect http://s3.neyer.me/respect-matrix-slides.pdf http://s3.neyer.me/respect-matrix-slides.pdf

ConfucianNardin 9y ago

Their claim/assumption is based in trust of the enrollment system. Check the section on enrollment - this image is especially helpful: https://raw.githubusercontent.com/the-laughing-monkey/cicada-platform/master/images/Enrollment-1.jpg https://raw.githubusercontent.com/the-laughing-monkey/cicada.... The enrollment software is supposed to be verified using "Challenge/Response", which is supposed to ensure the guarantees provided by the HUIDs. Of course, given enough incentive, this can (and will) be cracked...

andrewla 9y ago

Personal identity seems like a generally intractable problem, despite the lofty goals stated in the introduction. And without it, the document is built on a foundation of sand. An argument that I find myself having quite frequently about Bitcoin and pseudonymity is that pseudonymity and concealing identity is not a feature of the system, but rather that defining two-way identity in any sort of useful way is such a difficult problem that solving it would be a bigger achievement than Bitcoin itself. Seeking a biological equivalent to a non-portable key in a TPM is pretty hopeless, and that's what you'd need to get a HUID -- not just a hash fingerprint that anyone can appropriate, but an internal signing mechanism to allow zero-knowledge verification of identity. And the inverse problem, the Sybil problem, is just as intractable, and as far as I know has never really been solved without a centralized issuing authority or reliance on a scarce resource (like proof-of-work, which even there solves only a weak variant of the Sybil problem by conflating the unit of "identity" with a unit of "work").

Allower 9y ago
nailed it
anigbrowl 9y ago
We know multiple ways to reliably identify someone using biological markers with near-perfect accuracy. Yes, that lack of perfection leaves open the possibility of subversion, but so what? You'll grow old and die waiting for the perfect solution.
- hueving 9y ago
  And those are not subject to replay attacks?
  - xg15 9y ago
    This seems to me like the whole misadventure of putting finger print readers in door locks and phones: Someone noticed that fingerprints work quite well for identifying people in customs and law enforcement and simplified that thought to "fingerprints can identify people" - ignoring the circumstances of how fingerprints are taken when they work.

laughingmonkey1 9y ago

Thanks for posting this. It's always interesting to see people's reactions to my work. Usually when people do find it though, the series of comments that follow it tend to fall into a few categories. I will paraphrase and give my answer to these questions generically rather than answer every individual question: 1) "I've zeroed in on this one aspect of the system instead of the whole and found a flaw that invalidates the whole thing." In particular this comes up around the subject of an ID. Usually it centers on "biometrics are how the system is secured, those are gameable, this will never work." In fact, the primary method of controlling the spread of fake IDs is the reputation system is NOT the biometrics system which is merely a first barrier to entry. It is the reputation system. And that reputation system is different from rep systems that exist today which basically consist of people voting thumbs up or thumbs down which is not what I describe. Most of the misunderstanding around that is because the original paper was not as clear as it could be on that aspect, so I wrote up a second paper that breaks down the ID system and Rep Bank in much more detail. That paper is here: https://github.com/the-laughing-monkey/cicada-platform/blob/master/Identity-Without-Authority-2017.21.3.BETA.pdf https://github.com/the-laughing-monkey/cicada-platform/blob/... 2) "This will never work because of XYZ reason I just thought up after glancing at this a few seconds." Those questions are always answered in the document. I encourage you to actually read and do so with an open mind. 3) Some comment/focus on whatever my supposed ideology is. People have zero idea what my ideology is and debating something based on whatever you imagine my ideology to be is not very useful. It's a fallacy of composition. Also my ideology can't be deduced from the paper, largely because I don't much like ideologies and I don't really have one. I look at every problem or issue atomically, to be solved on its own merits, thus my beliefs run the gamut. Also, whatever I think about things is pretty much irrelevant. The system stands on its own and it has nothing to do with what I do or don't believe. The only thing I really believe is "locking in open" as Tim Berners-Lee put it. That is: open systems are better and decentralized systems in particular are key to the future. 4) "Direct democracy can't work because people get tired or aren't smart enough to vote on everything." I outline many solutions to this, in particular systems that can vote for you based on answering a series of questions, or systems that work in hybrid mode that allow you to vote on only the things that really matter to you, say abortion and guns, as well as full voting so that you can vote on everything if you are a highly engaged citizen. I also propose an easy solution to the California ballot problem where every stupid thing comes up for vote and where stuff only gets on there based on big money. It routes votes to ever increasing numbers of users and only the absolute most important votes would ever make it a national vote. Lastly the system allows for expert panels to create legislation that is fast tracked by the system. This would be a dramatic improvement on what we have now where unqualified politicians vote on everything from science to military to personal issues. Having groups of competing economists propose bills, or scientists writing science policy would be a major improvement to the current system. Lastly, the paper also outlines a method for breaking down bills into simpler terms for various folks to make it easier to read them and engage with them. 5) "You're an idiot, why even bother doing something like this?" This is utterly pointless commentary that comes from folks who haven't done much and spend most of their life running other people down for trying something new. Not very useful or cool. 6) Not understanding that every single layer of the system expects some level of exploitation and it doesn't matter if you found a few ways to game it. The system always trusts there will be some level of gaming the system. It's expected and dealt with. Central systems are not a panacea to this, despite most arguments thinking that's the case. Their is all kinds of exploitation of centralized IDs for instance. The decentralized ID system will have some level of people hacking it and it really won't matter as long as the system is resilient enough to keep the blast confined and to update itself to deal with new attacks, which the build in security chain combined with the reputation systems deals with very effectively. 7) I get it and how do we build it? Congrats. Join decstack.com and let's talk.