Alternative consensus mechanisms/ algorithms are a new science, started in January 2009. The number of academic papers and PhD dissertations focused on consensus algorithms is increasing. This is not a theoretical science, this is an experimental science which can be run on a global scale.
We need to see large-scale deployments of practical use, with real actors and real money at stake. Different consensus algorithms produce different characteristics and features. Proof-of-work produces thermodynamically guaranteed immutability. This is not a zero-sum game, there is plenty of room for competition and niche applications.
Transcript
[AUDIENCE] Hey Andreas, thanks for the talk. Do you mind commenting quickly on... alternative consensus mechanisms that are being proposed in different blockchain environments? How does that impact your view on the private versus public blockchain space?
[ANDREAS] I think consensus algorithms are a very interesting area. It is basically a new science. [It has been growing] since January 3rd 2009, when there was [only] one [blockchain]. If you look at the the number of academics and papers published [on this topic], it has been growing...
at a hyperbolic rate [recently]; hundreds of PhD dissertations and papers on consensus algorithms, in the last year. I only expect that pace to increase. I expect to see entire masters degrees, post-graduate or PhD [programs]... focused on consensus algorithms There are interesting experiments in that space.
To me, the important [part] is to run your experiment... in the wild, at scale, with real actors and real money at stake, and see how it survives in those conditions. This is not a theoretical science. It is an experimental science.
We can run experiments on a global scale, and prove these systems [work] in real terms. To use the trite expression, that is 'where the rubber meets the road.' Until we see large-scale deployments of novel consensus algorithms, I know only one that works: That is proof-of-work. We will see what comes next. That doesn't mean you can't do it another way, but [we must] recognize that other consensus algorithms...
will produce different characteristics and features. Proof-of-work produces thermo-dynamically guaranteed immutability, something you can't do any other way. With proof-of-stake, we will see different results, features, and characteristics. This is a broad ecosystem in which different systems can specialise for fitness in particular niches, applications, and use cases.
There is plenty of room for competition. This is not the old zero-sum game of national currencies, where there can only be one in any particular domain, and it is winner-takes-all. That is not what we are doing here. It is a completely different [game].