On 17 Sep 2015 12:10 pm, "Phillip Hallam-Baker" <<a href="mailto:phill@hallambaker.com">phill@hallambaker.com</a>> wrote: > > On Wed, Sep 16, 2015 at 9:47 PM, Ben Harris <<a href="mailto:mail@bharr.is">mail@bharr.is</a>> wrote: >> >> At the end of the day, don't these all just build upon a different definition of trust? >> >> In the centralised architecture you trust that none of the authorities are compromised. >> In Bitcoin you trust that no one has spent the money to get to 51%. >> In PoS are we trusting that no one has purchased a majority stake? >> In transparency methods, we trust that no more than N-1 verifiers are colluding. > > There is a big difference in the incentives created. > > In a multiple notary system the notaries have a major incentive to reject collusion attempts, it is the end of their business. BitCoin on the other hand creates a multi-million dollar incentive to defraud the system with absolutely no accountability for the parties making the attempt. That may be, but you are combining what you trust with why you are trusting it. If we reframe the triangle as Security - Human Readable - Trust-less, then maybe we see that "squaring the triangle" is actually just selecting a convenient definition of Trust-less (e.g. You don't need to trust centralised authorities). All the Secure/Human Readable solutions will require some level of trust. It is about making that trust transparent and deciding if that is acceptable to you. To steal a concept from my industry, ATARP - As Trust-less As Reasonably Practicable.