Tezos: A Self-Amending Crypto-Ledger [pdf]

Proof of Stake seems like an interesting idea. I've often wondered if bitcoin's proof-of-work verification model would remain viable as the currency reaches the end of its generation phase and the incentives for hashing become negligible. Off-chain transactions will become more and more popular, which not only draws revenue away from the hashers but also undermines the consistency of the system.

Tezos' second innovation, its ability to incorporate changes to the protocol without forking, is also interesting but raises some questions. If I understand correctly, stakeholder consensus is required to integrate a change to the protocol. The author gives an example where a developer attempting to remove this safety mechanism would fail because the stakeholders, recognizing that the change is not in their interest, would reject it -- analogous to voters rejecting a proposal to eliminate democracy, I suppose.

However, this seems to rest on the assumption that a majority of stakeholders are familiar enough with the protocol to recognize which changes are beneficial to them and which aren't, in situations less obvious than the above example. Further, they must be willing to put in some time and effort to understand the amendment and cast their vote. My hypothesis is that this would lead to a disproportionate concentration of power in the hands of those who control large amounts of currency, precisely the problem that PoS purports to avoid.

The entities controlling large amounts of currency have plenty of incentive to understand the protocol well and vote on every proposed amendment. The average user, who controls a small wallet and uses the currency for day-to-day transactions, has much less incentive: the average user of HTTPS does not feel the need to educate themselves on the underlying technology, nor is the average user of dollars familiar with macroeconomic policy. So the amendments which pass "review" will be the ones most favourable to larger entities, even if these entities control less than half the actual currency. I don't buy into the idea that those with monopolies have little incentive to abuse them: even if a protocol change will eventually alienate users and devalue the currency, the controlling entities may still choose to implement it for short-term profit, undermining the currency in the process.

This seems like an amazing development.

If I read it correctly, Tezos aims to decentralize the core protocol development by allowing forks to coexist and peacefully merge in any innovations.

It is also in OCaml, which is great for the robustness of the code developed. Much less scary than C/C++. Tezos even goes to the point of suggesting formal proofs of correctness, using COQ proof assistant.

Very nice write-up on where are we standing currently, but that's pretty much it.

General idea seems great, but there's not much more than hand waving about it.

The name is most likely a reference to Leah McGrath Goodman (who, if you don't know, wrote the piece on Dorian Nakamoto in Newsweek).

Is there an implementation in the works? Also, if you're the author, find me on #tendermint on freenode. There are some things about PoS that I'd like to share with you.

who wrote this, Satoshi?