Prof. Pandu Rangan gave an engrossing lecture on Ahaa! Blockchain! on Oct 20, 2018 at LNMIIT as part of LNM Distinguished Lecture Series. The lecture was well attended even though it was scheduled during the long Navrathri days. The talk was supposed to be from 2 to 3 pm but it went until 3.30pm and actually the listeners at the end of the talk got the feeling that they needed more!
Professor started the lecture with a question – where should I begin? He answered with a popular quote let us begin from the beginning! He mentioned that Aha moment is an intellectual moment. It occurs when an idea strikes like a lightening and Blockchain is such an Aha moment that has the potential to change the world into a completely new direction. In the talk he mentioned several Aha moments that I shall try to summarise it here. In no way this blog covers all the points mentioned in the talk but it gives an overall summarisation of the whole 1.30 hours talk. I am not an expert in this area so if there are any mistakes here I take responsibility – please mention the mistake I shall correct and edit it here.
He first touched upon the double spending issue in the financial world which had a big cascading effect especially in the Western World that brought the bank systems to almost a halting state. He added that this happens mainly because entities which receives payments are unaware of the fact if truly sufficient source money is available or not. This is similar to giving cheques for ₹50 each to two people when there is only ₹60 is available with you. Hence if the information about the availability of source money is there with all, in a decentralised manner, this double spending need not be a big problem – since such transitions could be rejected at the first instance itself. But if all the financial data is going to be with all, then we need to take care of the security, privacy, and anonymity of all the transactions. Also, when the network is distributed it requires consensus.
Distributed consensus is a well studied problem and we have good results for the same. But here we are talking about decentralisation and not distributed computing, where we know various information about the systems involved like the name, the location, how many are there and so on. In a decentralised network it is on a completely nameless network with no information about how many systems are there (of course, we can give an estimate but cannot be exact), which are the systems are active at the moment and so on. Hence the complexity is more in the case of decentralisation with incomplete information about the network.
A man virtually known as Satoshi Nakamoto came up with this ingenious solution. Aha moment! He blended two concepts – bitcoin (a virtual currency) and a distributed ledger that stores the transaction information in chain of blocks. Each block created for a transaction has to be approved and that requires a consensus. Blockchain is the abstraction for several transaction pooled with time stamps. Here the sequence is strictly maintained. Moreover, here the only operation allowed is Append – that means we can only append to the existing sequence with a new block once it is accepted through a protocol. Also, block rejection will be a loss of money for the transaction initiator. To avoid the inconsistencies the chain is built in a unique way. But, in reality, that may not be the case as there may be a chance that a fork might happen in the chain that may create two or more parallel chains from the forking junction that may lead to inconsistencies. Here again, Nakamoto’s work provides a way. Another aha moment! His theoretical work guarantees that by accepting the longest chain at any point of forking instances we will be able to achieve the uniqueness of the chain and thereby avoiding inconsistencies.
Professor quoted here we are humans who build solutions for the problems but at the same time we are humans who find problems in the solutions! Consequently this problem is not yet fully solved. It is evolving and new ideas emerging almost every week. The biggest open problem to solve here is can we prevent the 51% attack – a hypothetical, but still a possible scenario where more than 50% of the bit coin miners can combine together and pool their resources in and prevent new transactions from gaining confirmations. This will result in halting payments between specific, any, or all users (denial-of-service attack). This scenario might also lead to reversing the transactions that were completed before and make double spending possible. To a query from a student Professor mentioned that blockchain transaction would not be ideal for transactions involving smaller money. It is mainly meant for only transactions involving huge money. He also briefly mentioned about Ethereum project and its fully open source framework to do bitcoin transaction using blockchain technology.
Lot to be done, lot to be learned and lot to be experienced – hence this is an exciting area to do research and Professor, at last, invited all of us – welcome to the world of blockchain!