The distributed autonomous corporation (DAC) scares me. The fundamental idea of federating services across to miners responsible for running the blockchain and larger network is very interesting, but the implications appear to me to be more dystopian than utopian. Maybe it is just latent pessimism, but for every good example I can think of, I can think of some truly terrifying ones. For those unfamiliar with the characteristics of a DAC, I’ve cribbed this verbatim from the Coinwiki page on the subject and added commentary regarding each attribute.

  • They are corporations, they are free and independent persons (but don’t have legal personality).

Railings about the evils of corporate entities aside, the corporation has been frequently compared to a pathological individual. It has no moral standing except for the pursuit of profit. This isn’t really a commentary about whether that sole purpose is a good or bad thing, except to say in the physical world there is a framework of laws and people that can ultimately be punished when a corporate entity behaves poorly. Since people are responsible for the operation, society has legal, moral, and physical levers available to alter behavior.

  • They are autonomous once up to speed, they no longer need (or heed) their creators.

Once the code hits the network, it can function independently of human authority. It responds to stimuli, unhampered by any semblance of moral code except that imbued by its creator.

  • They are distributed there are no central points of control or failure that can be attacked.

Once developed and dispersed, the entity cannot be removed from the network. There are many questions about how bad behavior can be squelched in this system. If anything I would imagine it would be encouraged as the DAC itself could pay miners/users to keep itself running regardless of externalities.

  • They are transparent and their books and business rules are audit-able by all.

This I generally think is a fantastic feature. Public auditing of transactions, and the ability for software to introspect what is happening at a fundamental level could be a boon for legal compliance. However, most of that benefit is really a function of pro-grammatic money.

  • They are confidential and customer information is securely (and incorruptible) protected.

So the entity can be paid for its work without revealing who is paying it, or in any way prying into their motivations.

  • They are trustworthy, because no interaction with them depends on trust.

Trustworthy in this sense means that they do what they do. You can audit what they do. It doesn’t mean what they do is particularly nice.

  • They are fiduciaries, acting solely in their customers’ and shareholders’ interests.

This is a bit of a stretch. Companies don’t act entirely in their customers’ and shareholders’ interests. I think because of the transparent nature of the entity it would be difficult to fund something that actively subverted the interests of it’s shareholders, but not impossible. An entity will pursue profit, or survivability in some context.

  • They are self-regulating – they robotic-ally obey their own rules.

Self-regulating means that they do what they say they are going to do, and you can audit those rules, inputs, and actions. It may be obvious to point out, but this has nothing to do with real world ‘regulation’

  • They are incorruptible and no one can exercise seductive or coercive influence over them.

Their programming cannot be changed, but they could be thoroughly corrupt to begin with.

  • They are sovereign over their digital resources (but don’t have legal capacity).[2]

Moneys in their control are used to exercise the purpose for which they are created.

What could possibly go wrong. Their are numerous people in the community that look at this list, think about all of the corruption of the physical world, and yearn for a perfect merger of form and function. Something that binds human labor to a constant, irrevocable, self sustaining and incorruptible force. A guardian if you will, enshrined in code that optimizes, protects, and provides necessary services to all. Without further ado, a list of very bad things, in no particular order that DACs can make a reality.

  • Real-time Assassination (or binary options for evil)

Let’s mix a binary option with a data feed that injects data to the DAC on whether an individual is alive or dead. You pay into the address representing the person you want killed. Potential assassins can place a bet with the DAC that the person will die within the next second, minute, hour, or day. Accumulated bets go the pay-out. If the option expires true on consensus that the feed event occurred and the target destroyed, the money pool is paid out to people who took the timed bets. Anonymous financial incentive to murder, completely outside the rule of law.

  • Identity Theft Collective

How much is an identity worth? Let the DAC decide. Pay the collective to obtain fraudulent identities, and incentivize human labor to do the dirty work. The more data delivered, the better the pay-out.

  • Hackers For Hire

Have a target you want to hack? Let the DAC federate that work for you. Incentivize it to find skilled human labor to do the job, put in parameters for successful completion, and let it run.

  • RansomBot

Why not a viral DAC? Let’s mix the Cryptolocker virus with a DAC responsible for distributing the keys. It incentivizes humans to subvert machines for a cut of the profit relative to the number of machines they have compromised. It ransoms the hard drive contents back to individuals and uses the profit to fund itself forever and ever.

  • Coup Machine

What about pesky governments? A variant on the real-time assassination market incentivizes all sorts of real life physical interaction. Like the change of major government bodies, or the subversion and stalling of regulatory processes.

All right. Clearly there are good uses also. I am definitely playig devil’s advocate. You could imagine incentivizing charity in an equitable fashion that dramatically reduces waste as an example. The Internet itself has been an instrument for all sorts of terrible things, and yet the net good that it has generated has been incalculable. I think the most important difference is that the Internet does not bind human labor with financial incentive in the same way a DAC potentially could. This leads to a very dangerous synergy, where real life actions can be aligned, or constrained by code that is not responsive to social law. Food for thought. Let me know what you think in the comments, and if there are some overarching examples of social good that you think a DAC could create I would love to hear it.

Comments