1. The Meaning of Autonomy
1.1. The Experience of Autonomy
Autonomy does not exist out in the world, an objective phenomenon, pi in the sky. It is always already in relation to others whether one is perceived as autonomous. In the absence of others, one neither has or lacks autonomy.
When a state asks for autonomy within the confines of a country, it is not so much saying, “I am autonomous so recognize me as such”, but “Recognize me as autonomous so I can be(come) autonomous”.
It is meaningless to state “I am autonomous”. Rather, the claim could be “others recognize me as autonomous” following which, specify who are the others in question. That is, the sense of autonomy originates from the outside (being another) looking in (at oneself) till experiencing the limits of that gaze.
The question of autonomy is not whether one is or is not autonomous but whether one’s peers recognize one as such, for it is only through the reflexive awareness of another’s recognition of one’s role, that one experiences oneself as autonomous.
This makes autonomy, as a category, more epistemic than ontic; it reveals itself by way of an “as-a” relationship over an “is-a” (am-a) relationship.
It is incorrect to state one is or is not autonomous as autonomy is not a standalone category. Rather, we can say, others see me as autonomous, thus, I experience myself as having autonomy. Autonomy is a role that our relationships take on, and not any kind of essence of our being.
1.2. The Topology of Autonomy
When a country recognizes a particular state as having decisional autonomy to a certain degree (such as by way of the Tenth Amendment of the US Constitution), it means anything because of the presence of other states and the federal District of Columbia, who are all competing for the ultimate decisional authority.
Here’s the text for the Tenth Amendment:
Each state retains its sovereignty, freedom, and independence, and every power, jurisdiction, and right, which is not by this Confederation expressly delegated to the United States, in Congress assembled.
Only in the presence of others who are not aligned with one, does one’s autonomy come into question. Autonomy is always from another’s perspective first and foremost, not from one’s own — as what others can not do to us, more than what we may do.
While autonomy expresses itself as both an autonomy-from and an autonomy-to, the former stands to be its originary movement in terms of proving one’s legitimacy as autonomous.
Thus, it is meaningless to frame any language of autonomy from a pure self-centred (i.e., other-agnostic) and enabling (“what one can do” — decide) lens. Instead, we should frame it only in a social (other-focused) and prohibitive (a la restrictive liberties: “what others can not make me do” — what is wrong for them to do instead my rights to defend myself) language.
- Other-centric: (outside gazing in)
- Preventive: There could be a government shutdown such as the US federal government shutdown (of 35 days) from midnight EST on December 22, 2018, until January 25, 2019, where the government was not enabled to do much, but it was still autonomous from other countries since they could not run its government on its behalf.
As an example, consider the two types of framing for an injunction to institute “autonomy”:
- “This law enables every state to be able to decide on their own.” — Seems incomplete, being focused on individual states and has an enabling style
- “No state can prevent another state from deciding on their own” — Is closer to the truth since what a state can do to another is a social realty
1.3. Autonomy contra Automation
A contrast to this language of autonomy can be found in the language of automation, which is an individualistic and an enabling one. Automation, as a category is more in the kinetic space than in the political one, as it deals with being able to run by itself — clockwork perfection.
Automatic = One can run by itself, with batteries, without needing a human hand; it’s opposite, being the mechanical
Autonomous = Others can not intervene in one’s running; it’s opposite, being the dependent.
It is for the benefit of the wearer of the watch — in that she does not have to wind it anymore — that it is used and not for any benefit to the watch itself, whereas for autonomy such as the autonomous state, the primary beneficiary should be the self over others, that is, the state itself and not other states.
The real benefit of automation is for its users and not for the automated thing. On the other hand, the real benefit of autonomy is for the autonomous entity itself, if not, it would like a puppet state with a fake sense of autonomy. When autonomy benefits others more than the autonomous, that amounts to exploitation.
While the meaning of autonomy is to be found in the non-interventionability from others in one’s life, the benefits of autonomy are geared for oneself over others. For automation, its the other way round, in that, its meaning and benefit lies in itself and others respectively.
Slavery is when we denigrate the autonomous to act as an automation.
1.4. Automation with Autonomy
Automation and autonomy are not islands to each other, they are in constant dialogue, re-shaping each other in a world with evermore technological automation.
When we use tools (software or physical) that automate parts of our life for the sake of higher efficiency, it takes our sense of autonomy out of our hands by giving it to those machines and software. But at the same time, it leaves us with the room to be autonomous in other parts of our life.
Consider creative tasks (such as developing a gaming world) where the non-creative parts have been automated. That leaves more room for the developers to be autonomous in their creativity around the conceptual building blocks of the game. Here, the automation in some aspects helps to further the autonomy in other aspects.
When it comes to blockchains, we want the consensus on the next block to be arrived at automatically (as promised, such as ~10 minutes on an average for Bitcoin), while the different units of software running on top of it be able to run with some autonomy from the constraints of the underlying consensus protocol. Here, the autonomy of the software on top gives the underlying automation of the consensus protocol that much more value, making it into a promising blockchain (e.g., what is Ethereum without all the dApps?).
If automation does not help to further the goals of the autonomous using it, then the automation becomes an end in itself — automation for automation sake, making the automation a burden for the autonomous (e.g., owning a car too expensive to maintain, is a burden for the human owner; or, running a full node too expensive for one’s computer).
Often times, the burden of automation leads to its own emancipation as the owner decides to give up her ownership not being able to afford it even though the rhetoric is usually couched in liberation language (“look at me, I am so liberal. It took time, but I finally woke up”. This couched rhetoric was the general stance post-WWII of most European colonizers who had to decolonize not being able to afford the cost of domination after such an expensive war.
For an autonomous user, the cost of automation should not outweigh its benefits, however, in regards to autonomy, no cost is high enough as it is often perceived to be its own benefit. The factors that do put a cost on the need for autonomy are the dependencies that the autonomous has in order to keep running (such as for an autonomous state, it could be resources from other states, and thus, the more powerful may put embargos). For human autonomy, this is our fearful aversion of death and suffering.
By accepting death and suffering as options for us, no matter how much power someone has over us, the power to delegate our power to that person always lies in our hands. We lose our power to delegate as we start looking for trade-offs in an effort to minimize our chances of having to face death or suffering. In many ways, the cost of being autonomous is to be ready and accept death and suffering as part of life. This is encapsulated in Buddha’s first noble truth, “this is suffering” [this = life], which is a recipe for autonomy or, when Satre says, “Man is condemned to be free; because once thrown into the world, he is responsible for everything he does.”
When we go from software-based automation to software-based autonomy, we have to ensure whether and to what extent are the DAOs, DACs, and their ilk must be free enough to benefit from their own efforts over benefiting the automation of the underlying consensus protocol. In that way, we ought to
- be ready to give up some of our ownership of these units of software as a prohibitive measure against our power to decide for it
- ask ourselves at what cost, are we willing to acknowledge its autonomy from the underlying protocol
1.5. Privacy and Autonomy
Computer viruses hook itself onto another program and get itself executed when that program is executed, thereby, not needing anyone to explicitly execute it. It is not even possible to execute it intentionally since viruses keep their existence hidden, they operate in stealth mode. This makes it out of reach for regular users to prevent its execution as well. It executes all by itself upon the triggers as instructed by its code.
The execution of a virus neither needs any explicit human intention nor is it possible to provide one. No user can say one day, “today is a good day to double click on that virus file I downloaded”, and if anyone does have that thought, it is no longer a virus, but just another software albeit its harmful effects (such as the case for the creator of that virus).
To further its life, it can replicate itself by making copies which get sent to new computers by attaching to an email, all of this done in stealth mode.
Both for its execution and for its lifespan, it’s hidden nature ensures that it is unstoppable, except for an antivirus software capable of detecting and eventually removing it. The computer virus exhibits one form of autonomy, one where the autonomy is gained by staying hidden from those wanting to and capable of intervention.
There are positive sides to this form of autonomy, as found in private socio-legal orders in a plural jurisdiction. Consider the freedom to practice any religion in one’s own home as long as one keeps it in one’s home, much like the popular adage, “what happens in Vegas, stays in Vegas” or the Vegas Principle. One’s need for privacy should be proportional to the possibility of getting intervened when doing the same thing in public, while ensuring that one can keep all of the collateral within that space, with no spilling over.
Blockchains address this with its anonymity feature that some like ZCash and Monero have, or just the old technique of a new Private Key for every transaction.
In terms of autonomy, this is the justification behind the legitimacy of private spaces: Not everything should be public, somethings that are particular to one’s own taste should be hidden from those who could and would intervene. Even if we think we have nothing to hide, we can never be sure what might offend someone or what might not be in someone else’s benefit, and thus, get censored or objected to.
The Vegas principle is, however, different from the Tenth Amendment (as the Federalist principle) as the autonomy of the latter does not need private spaces. The Federalist principle is a perfect tool for actions that necessarily spillover, and thus, need a contractual understanding obliging each party to tolerate and even accept to a certain degree as per their need of reciprocity.
What is key to an ecology of jurisdictions, one where we strive to stay away from the monopoly of a single law, as is the case with International Law, is the need for acknowledging both the principles of federalism and that of Vegas, mixed in the right proportions as context demands.
1.6. Sovereignty and Autonomy
There is another form of autonomy, one that does not require its working to be hidden. Consider a car that can place an ad to rent itself out, and then use the service fees to repair itself — provided no rider can obstruct its autonomy. This does not mean a car mechanic can not obstruct its autonomy when it has to take it apart but can only do so if the car has reported the damage. The car mechanic will be operating under a mutually agreed-upon contractual relationship, one that acknowledges each others’ autonomy.
However, someone superior to the car, the owner of the car, can obstruct the autonomy with a one-way command, since autonomy does not preclude sovereignty.
Sovereignty is when you are your own master, no levels higher than you, whereas autonomy means no peer can decide for you unless by way of a mutual agreement.
When certain peers are autonomous from each other with no superiors to take the autonomy away, those peers are then classified as sovereign but only within those members; and when that applies to every member of the network, it is a self-soveriegn network. There are no indivdual self-sovereign members in a network, it depends on the condition of autonomy between every member.
Autonomy does in no way make hierarchies irrelevant, on the contrary, hierarchies reinforce autonomy by enforcing the autonomy of those under it. Self-sovereign in a network each peer is autonomous from the other, where there are no hierarchies.
2. The Question of Autonomy
2.1. The Politics of Autonomy
If I was the same rank as you, that would deter me from meddling in your business. Otherwise, having a rank is useless. And if I tried to meddle, you can always appeal to someone of a higher rank to settle the matter.
The question of who is autonomous from whom is one of rank for between autonomous peers in rank and strength, there is neither the presence nor the lack of autonomy.
The question of who is autonomous from whom is one of who is more autonomous from both. It is the ones with a higher rank, as the more-autonomous ones, who are there to enforce that everyone below one’s rank be respectful of the distances to keep between their peers.
The question who is autonomous from whom is one of who others are as autonomous as others, as well as who all are more autonomous than us.
Between players of the same rank, if there be no clear power structure defining their roles and privileges, it’s all up for grabs as to who is worth how much to whom. Autonomy does not remove hierarchies, rather, it demands to stay within hierarchies of appeal and enforcement when resources are shared between equally autonomous entities.
Having power demands of us that we actively protect the power of those under our power. This means disempowering those under us who are using their power to disempower others. While doing so, one can not be cautious enough to not end up removing the political opportunities for the disempowered to self-empower themselves against those oppressing them.
Having power demands of us that we be careful of not enjoying our exercise of power to the point that we help recreate the inequalities over and over again in the hope of coming along to protect against it. This happens all too often after the first few years of honemooning when revolutions are won; point in case: the Bolsheviks.
The ones with power-over (the higher rank ones) do not have to be people, humans or living entities as we usually think, but could be a rule or set of instructions as commonly agreed upon by the peers, much like a holy book as the word of the G-d or the Rule of Law, or the consensus rules of a blockchain.
However, even in those cases, it is people who are to decide the meaning (as argued in the §2.2 of Part II of this series) interpretive hacks like what lawyers do, framing the law in terms of which section comes before and where are the commas) of such a code.
Even for a computational rule-based system, such as Bitcoin, Nakamoto decided some rules, Intel decided some other rules for the processors, Cisco decided how to packet switch, the power companies on how to negotiate with miners, core devs on protocol updates , and so on and so forth — all people, all humans.
Although the humans still remain in a rule-based systems, the difference from human-only power structures is that the rules once transcribed in a computer code or a book, serve as an in-between, almost as if the transcription (the constitution, for example) becomes another kind of species working with humans to sustain the power structure. For all successful and scale-free systems, the life of this “in-between” is so persistent that the human meaning-workers come and go, while the rules have its own ultimate purpose beyond the scope of any particular person. Thus, it’s so hard to change the rules once it becomes a tradition.
2.2. The Specifics of Autonomy
An adult human is considered to be autonomous, but that does not mean that it does not have to follow the socio-political rules of the society she lives in. In fact, because adulthood is conceived of as the state where one is mature enough to follow the rules, the reward of which, that adults acknowledge a certain degree of autonomy towards each other as peers. This means every adult agrees to be brought under arrest when breaking a given rule.
Autonomy is always a rule-based (conditional) experience — as set by the agreement among peers to follow those rules, and as is the agreement enforced by those higher than one’s peers, such as subjects and institutions of authority.
Nothing is absolutely autonomous, except for an autonomous sovereign, such as G-d or the Absolute Monarch of a Kingdom.
But even then, nothing is completely autonomous; there are always externalities that every autonomous entity depends on as a necessity. The car of our example does have to depend on those running the electric or oil company for if they all shut down or deny services, it can not do anything. It can come up with strategies to compel the power companies to lift this embargo but its autonomy, essentially, ends there.
You are autonomous to the extent that your peers are unable to replace you, take your place, become you. Each one us have multiple areas of autonomies, where each area is founded on its own peer group with overlaps in between and their own set externalities they have to depend on.
If we claim that a blockchain-based software is autonomous, we must specify
- Who are the Peers?
- What is the Conditions (checks-n-balances by way of rules) under which its autonomy is upheld?
- Who are the Superiors who can obstruct that autonomy?
- What are the Externalities necessary for that autonomy?
- Which of its parts are not autonomous?
- Which of its parts need to be private from what other parts in order to hide from getting intervened?
- At what cost of running the autonomous software, are we willing to make sacrifices to its autonomy and how so?
When we are classifying a piece of software as a DAx, we have to clarify as to which peer group it is autonomous from, under which ruleset, and where does its autonomy end.
As there are multiple sets of peers for every piece of software, each set representing a different stakeholder type, there are multiple instances of autonomy, where each instance has its own set of externalities and conditions that make it so. It is as if there are multiple souls to every software with some overlaps among them.