In this talk in Berlin, Andreas looks at the inner structure of bitcoin and how high-level financial and trust applications are composed from smaller elements. Using analogies from Lego blocks to a chef creating new recipes, this talk highlights the connection between creativity and the flexibility offered by fine-grained components.
Transcript
[AUDIENCE] This is a short question. What is your opinion on a block size debate? [ANDREAS] A short question indeed. Surprisingly enough, I have not been asked that yet...
this evening. [Laughter] I [recently] delivered a talk on scaling and why scaling is not something you "solve," but something that you continuously aspire to and push back on as a frontier. Scaling is the thing you can't yet do today, and how you keep pushing that frontier back... by failing to scale [over and over again], and then scaling a bit better to create opportunities...
for new applications. Scaling is not something we will solve, but what we will improve continuously, hopefully over decades, until we can [support] more and more interesting [applications]. In the [scaling] debate today, there is a lot of noise and drama. Part of that is because, we are [accustomed] to financial systems where [we perceive] zero noise [in governance].
In traditional financial systems, there is a committee of twelve people (usually white men) who gather together, smoke a couple of cigars, and ask, "Well, how can we best help our shareholders?" "I don't know. Should be drop interest rates again?" "We have already dropped interest rates one hundred times and the economy still dead." "Maybe one hundred and one times will do the trick." [Laughter] "Okay, let's drop them again and see." "I will be in the Bahamas next week, I don't give a shit." "Okay John, will you make the announcement?" "Sure, I will make the announcement." "After zero deliberation and absolutely no disagreement, we will now be dropping interest rates." Great... Oh, at least they are not fighting [with each other], because that would be bad! Democracy is loud, noisy, dirty, and weird.
Imagine if someone in Bitcoin could say, "It has been a year now since you all started talking about the block size limit." "But now it is time to make a decision. Here is what we will do." That is the one thing you can't do in Bitcoin. There is no one in control. We can't simply dictate our decisions to everybody else.
Consensus is an "everyone together or not at all" proposition, and that is not a bug. It is a feature. It is probably the most important feature in Bitcoin. People have not yet realized that we have reached...
a point where the distance between opinions is very small. A year ago, we were talking about a "20 megabyte block size limit with exponential increase until forever"... versus "no hard fork!" [Laughter] Now the debate is about a 2 megabyte block size increase plus Segregated Witness later, while the other side is [enabling] Segregated Witness now plus a 2 megabyte hard fork later. That is close.
If you haven't noticed, Classic just reached more than 6% [of the network]. Segregated Witness activation on BIP-9 requires 95% percent consensus, so Core has veto over the hard fork and Classic has veto over Segregated Witness. Interesting times ahead... When two parties hold a veto on each other's [desired outcomes], compromise is just around the corner.
It is an illusion that one party controls Bitcoin. You might appear to have a lot of power in Bitcoin... as long as you continue to do what the consensus wants. But when you stop serving consensus, your power evaporates.
That is what I think about the debate. I am not pro-Classic. I am not pro-Core. I am pro-consensus and consensus is coming; within the next year and a half, we will [activate] Segregated Witness, hard fork to 2 megabytes, thin blocks, invertible bloom lookup tables (IBLTs), and a number of other optimizations.
Then we will hit a scaling wall again, and the scaling debate will continue.