Decentralized Autonomous Corporations

From P2P Foundation
Jump to: navigation, search


Vitalik Buterin:

"Decentralized autonomous corporations/companies are a smaller topic, because they are basically a subclass of DAOs, but they are worth mentioning. Since the main exponent of DAC as terminology is Daniel Larimer, we will borrow as a definition the point that he himself consistently promotes: a DAC pays dividends. That is, there is a concept of shares in a DAC which are purchaseable and tradeable in some fashion, and those shares potentially entitle their holders to continual receipts based on the DAC’s success. A DAO is non-profit; though you can make money in a DAO, the way to do that is by participating in its ecosystem and not by providing investment into the DAO itself. Obviously, this distinction is a murky one; all DAOs contain internal capital that can be owned, and the value of that internal capital can easily go up as the DAO becomes more powerful/popular, so a large portion of DAOs are inevitably going to be DAC-like to some extent.

Thus, the distinction is more of a fluid one and hinges on emphasis: to what extent are dividends the main point, and to what extent is it about earning tokens by participation? Also, to what extent does the concept of a “share” exist as opposed to simple virtual property? For example, a membership on a nonprofit board is not really a share, because membership frequently gets granted and confiscated at will, something which would be unacceptable for something classified as investable property, and a bitcoin is not a share because a bitcoin does not entitle you to any claim on profits or decision-making ability inside the system, whereas a share in a corporation definitely is a share. In the end, perhaps the distinction might ultimately be the surprisingly obscure point of whether or not the profit mechanism and the consensus mechanism are the same thing.

The above definitions are still not close to complete; there will likely be gray areas and holes in them, and exactly what kind of automation a DO must have before it becomes a DAO is a very hard question to answer. Additionally, there is also the question of how all of these things should be built. An AI, for example, should likely exist as a network of private servers, each one running often proprietary local code, whereas a DO should be fully open source and blockchain-based. Between those two extremes, there is a large number of different paradigms to pursue. How much of the intelligence should be in the core code? Should genetic algorithms be used for updating code, or should it be futarchy or some voting or vetting mechanism based on individuals? Should membership be corporate-style, with sellable and transferable shares, or nonprofit-style, where members can vote other members in and out? Should blockchains be proof of work, proof of stake, or reputation-based? Should DAOs try to maintain balances in other currencies, or should they only reward behavior by issuing their own internal token? These are all hard problems and we have only just begun scratching the surface of them." (