Thread by @hus_qy

@DesheShai Btw. what we discussed is actually just going to be the pre-consensus of the entire protocol.

We decided to use PoS as a pre-consensus as it is more efficient and secure than PoW when it comes to producing agreements (you can slash attackers and so on) but the entire reason ...

@DesheShai ... why we want to enable users to issue their own blocks is because we actually want to tap into a completely new form of sybil protection for settlement that we call "social consensus" or "subjective asymmetric trust" (name tbd.) and that we believe to be superior to PoW ...

@DesheShai ... when it comes to establishing an "objective settlement security".

The idea is essentially very simple. Whenever users issue a block, they do not just attach their block to their preferred part of the DAG and therefore contribute to the decision making process but they ...

@DesheShai ... also include a cryptographic commitment to the state of a confirmed epoch in the past.

By chaining these commitments, we produce similar cryptographic structures as a blockchain (just delayed as everybody needs to wait for consensus to know "what to commit to" next).

@DesheShai Since every user that issues a block, signs his block, we collect statements not just by the "largest stake holders" but also by all of the users of the network that run nodes and attest to decisions being taken.

If the stakers would ever collude to rollback the chain, then ...

@DesheShai ... they might be able to create another fake version that has more stake associated weight (the typical nothing at stake and long-range attacks of PoS), but in our system this chain will lack all of the statements of the honest users.

Similar to PoW where an ever growing ...

@DesheShai ... amount of real world commitments (in the form of burned energy) piles up on past decisions, here also an ever growing number of attestations by real world actors piles up on past decisions (with some delay as we use PoS as a pre-consensus which takes some time).

Now it ...

@DesheShai is of course impossible to "objectively measure" the opinions of "all honest actors" of the network in the same way as with PoW but it is very possible to "subjectively measure" the opinion of a "locally chosen subset of all honest actors" that each node operator personally ...

@DesheShai ... chooses and that actually gets very close to PoW.

It doesn't really matter who I trust (be it my grandma, big institutions or only nodes that I operate myself). As long as my chosen subset is a "good random selection of all honest actors" that are active in the system ...

@DesheShai ... and that regularly create commitments, it becomes a good enough random subsample to "directly measure social consensus" (the quorum can be arbitrarily big due to the use of virtual voting) and I can build a consensus around that.

The chain-selection rule will essentially ...

@DesheShai ... be: I only switch to a stake-heavier chain if this heavier chain is also supported by a majority of my locally chosen trusted actors and if I sync with the network without "knowing anything", I will sync with the majority chain of my trusted actors.

This entire system ...

@DesheShai ... has very similar properties to traditional PoW systems and I think it is even superior as it not only becomes more and more decentralized with increasing adoption (more potentially trustable actors with a real world reputation use the system), but it even allows node ...

@DesheShai ... operators to "manually evict" misbehaving trusted actors from their trusted set if the trusted actors would ever collude in a big conspiracy and turn malicious.

An "external attack" that "mines things in secret" becomes conceptually impossible and the only way to break ...

@DesheShai ... the system would be to not only corrupt a majority of the stake but to also corrupt a majority of the trusted actors and the question remains if such an agreement exists, then there are maybe very good reasons to roll-back the state (i.e. because we faced a catastrophic ...

@DesheShai ... event like the infamous BTC inflation bug).

It is important to note that this should be able to even reconcile disagreements in very hostile network environments like large scale network splits.

If we for example assume a 70 / 30 split of the network then we will most ...

@DesheShai ... probably not just see a split of stake but an equal or at least related split of activities of honest actors.

If nodes choose the heavier chain accordingly, this whole system should gradually converge in the same way as the heaviest chain rule in a PoW system.

As a ...

@DesheShai ... proponent for PoW - would you be interested in setting up a call to discuss these ideas and maybe have a public debate about remaining benefits of PoW over such a proposed system?

I would really like to learn if you see additional benefits of PoW that go beyond just ...

@DesheShai ... establishing a superior settlement security.

@DesheShai @Justin_Bons this might also be interesting for you as you previously discussed the benefits of PoS over PoW but you kind of ignored the increased settlement security that PoW based systems usually offer.

With the proposed model of users issuing their own blocks, you can get ...

@DesheShai @Justin_Bons ... the best of both worlds and you no longer need to ignore these drawbacks of PoS based systems when discussing PoW (maybe you would also be interested in joining such a public discussion)?