Freedom transparency and privacy - Framing the dynamics

Summary of previous articles

First a brief summary of the articles which shows my understanding of what the authors said. I hope I have not misrepresented this. The articles are:

and

@dan poses the question which sees radical transparency and radical privacy as mutually exclusive, and argues that government use privacy and secrecy as means for disinformation and propaganda and argues that privacy is ‘the enemy of truth’. Looking at the conduct of politicians and governments quoting privacy as the reason for not sharing what they know about an issue and not disclosing information that is in the public interest would be a good example of Dan’s claim. The distinction between resist and aikido options for fighting for freedom are to increase our own privacy while expecting transparency in others or to turn privacy against tyrannical government by surrendering all our privacy which then affords total and radical transparency and forces the hand of those who wish to keep things secret. The argument is strengthened by the sense that our current belief in privacy is largely an illusion especially under the growing trend towards machine learning that may end up cracking even the hardest of codes protecting the thin veneer of privacy. Thus, sticking with privacy is likely to strengthen those with the power to apply algorithms for the purpose of breaking through the walls of citizens privacy protections. Conversely using the thought experiment of a software that could delete any online trace of our lives Dan suggests we would be completely free, although this comes at the cost of wrong-doing being wiped off the map and losing our ability to be aware of fraudsters and past criminal conduct, etc. This would then require us to rely on government to track down ‘bad actors’ and would hand corrupt governments the ability to use the technology to obscure their corruption from view. Transparency affords the ability to make everything visible and create social rather than physical defences against misconduct. If everything is transparent social pressures can be brought to bear on the individuals committing crimes. Transparency allows society to come to the aid of people who are hurt by governments and consequently government will lose power. Arguing on the basis of public blockchains as an exercise in transparency and freedom of speech Dan argues that a completely transparent public record can be created, but that the cryptocurrency community is letting government take the moral high ground by pointing at the lack of self-regulation to restrict freedoms and resit the disruption of transparency to government. The fightback to this has to be the requirement for radical all-encompassing transparency form which not even organised crime can hide. In summary you cannot have total radical transparency unless you have no privacy also.

@dana-edwards in reply to Dan has a different take on the question of transparency and privacy. Dana raises the objection that the question posed by Dan was mis-framed and should be about whether or not transparency and privacy can coexist rather than whether one exists necessarily at the expense of the other. This is also what I commented originally to Dan and what alerted me to Dana's excellent response. I would agree with this point for reasons that will become clearer further below. But essentially the question becomes more about how they might coexist. Dana asks us to imagine a world in which total transparency means our entire live will be open and rateable to others (IMHO similar to the dystopian future depicted in the Black mirror episode ‘Nosedive’).

The key part of Dana's argument seems to revolve around the idea that in Dan’s world humans watching other humans is fraught with problems. Transparency favors the elite wealthy segment of the world, the haves of this world, and in Dana’s words makes life a beauty pageant in which the have-nots lose out. In terms of opportunity and progress, transparency would have the impact of making lower power individuals more risk averse so as to be able to avoid losing reputation and other high stakes interests. Dana suggests AI has an important role to play here and does not object to being under the surveillance of AI provided humans are not accessing this. I imagine a scenario here where it is ok for your AI to predict that your current exercise regime means your life expectancy may be 2 years less than the average for people your current age and advise you what to do. But importantly for privacy this information should not be released to your health provider or insurer who would use it to discriminate against you and increase your fees. As regards trust, algorithms can be used to establish the basics without a human being needing to know much information (As a trust researcher I would argue that it depends on what you are trusting them with). In contrast with Dan, Dana argues that to have maximum freedom you need both maximum privacy and maximum transparency and this can be achieved using AI that keeps private data private (e.g. where and with whom you had your last date) and I assume important information that concerns all of us transparent.

The above articles are not without problems but have provided an inspiring theme and led me to do some more thing around a way of capturing the patterned nature of the issues as well as to make a contribution to thinking about privacy transparent and freedom next.

Thinking about a framework for privacy and transparency

The question of transparency and privacy is an interesting dynamic. I found both Dan’s and Dana’s views compelling. This made me think about a way of systematizing a framework for thinking about all the different scenarios and came up with this rather crude model that looks at the permutations of privacy and transparency which can then be interrogated about freedom. First privacy is something that characterizes the data of individuals.and I have made this assumption. Transparency on the other hand is a property of a system, broadly referring to how opaque its properties are. Transparency may be related to privacy of individuals within it (e.g. privacy of politicians in a low transparency system) but can be thought of as a general characteristic of said system. The following diagram clarifies these properties and their relations by giving examples of what exists in each quadrant.

This is really just an initial working model and I am not sure of what other examples of existing situations in which the conditions are met. Please comment below to add any that you think would be relevant. Both articles where about the question to what extent is freedom dependent on privacy and transparency. Taking each in turn and using myself as the agent for privacy and government for the agent of transparency here are the scenarios:

Low T High P society: Government/organisation would not disclose what it was up to, spending of money by the leadership would be difficult to track. Depending on corruption there may be a tendency to obscure the truth and refuse to investigate claims of government scandals. However in such a society I as an individual would have a relatively private sphere, and government would not be seeking to expose my own behavior to its view, but similarly it would not permit a clear view into its own inner workings.

High T and High P. This maybe the idea democratic society that operates through consensus. Citizen have their data protected and have full access to the performance of their government or institutions on which to base their decisions, without being influenced by the institution/government knowing anything about them

Low T and Low P. This is the kind of model we would expect to see in a dictatorship whereby citizen don’t know what is going on and the state know everything about them, surveys their actions for the purpose of oppressing those who dissent

High T and Low P. I struggled with this but wondered whether this was applicable to societies in which privacy or private property had little meaning (e.g. religious orders in collectivist cultures) and in which actions by the leadership are shared in an open way.

So how free would we be in each of these scenarios?

My argument would be that the freest would be the High T High P society. Low T and High P would run the risk of not knowing if I was not actually in a Low T and Low P scenario since in the absence of transparency it might not be apparent that my privacy had been breached. The High T Low P situation in Dan’s world is difficult to imagine as there are few real examples on which to base any meaningful image on. The idea of a religious order in which there was transparent information, but where privacy was unimportant seems to suggest that what would be required for this type of society would be a very different mindset perhaps moving towards a more collective way of perceiving the world and subjugating the needs of the individual to a higher cause. It maybe that such a society may be dictatorial in certain aspects in that claims of private property and privacy would not be acceptable. This might might be a form of society that emerges under certain conditions and crises (e.g. global warming, energy and food supply crises, Preppers' world) that demand a more collectivist social organisation.

So in synopsis based on a slightly different way of thinking about the same issues of transparency, privacy and freedom, I would concur with Dana’s view that a High T and High P situation is the most free.

The really interesting question will be how much AI will shape future societies in relation to privacy, transparency and freedom - AI is something I confess I know less about, but it seems to me that the idea of an all encompassing and all-knowing AI, whether running privately or publicly has an air of Panopticum. AI may end up imprisoning all of us by giving us the illusion that we are all free. Would we even know the difference?