With the increasing adoption of Bitcoin, cypherpunks have once again risen to prominence.
As Gregory Maxwell and Peter Todd, self-proclaimed cypherpunks, and Adam Back, the inventor of Hashcash, pave a path for Bitcoin, their vision highlights a cautious approach in an attempt to not repeat the mistakes of the 90s.
Are they right or are they about to repeat the precise mistakes they wish to avoid?
Gregory Maxwell, one of the five core Bitcoin developers until recently, revealed in a post two months ago that he and Todd had a long history with Bitcoin.
“In 2001 Peter Todd was outlining the primary challenges (search spend-twice) in the construction of a Bitcoin like system with Hal Finney and Adam Back; in 2004 I was using and collaborating with Hal on the RPOW reusable proof of work system.”
While Finney was the first to embrace Bitcoin, neither Todd nor Maxwell are known to have communicated with Satoshi Nakamoto.
Debate over Bitcoin’s scalability
Over the past year the duo has been at the forefront of an intense debate over Bitcoin’s scalability. The differences, surprisingly, seem minimal.
A question of “timing and priority”, says Gavin Andresen, who is of the opinion that Bitcoin should primarily scale on-chain in the short term.
Gregory Maxwell disagrees, stating in a post that “the decentralized Bitcoin blockchain is probably the most insanely inefficient mode of communication ever devised by man.”
Some suggest that this may be a narrow view, looking at the code only, rather than its wider effects, such as potentially replacing bank branches or employees. Nonetheless, every node has to store and communicate every transaction.
The use of layer two systems
The idea is to slow down demand on nodes by the use of layer two systems, such as Lightning.
According to Maxwell, from its very beginning Bitcoin was design to incorporate layers in secure ways through its smart contracting capability.
In effect we will use the Bitcoin system as a highly accessible and perfectly trustworthy robotic judge and conduct most of our business outside of the court room.
He continues by saying that this is possible precisely because of the core properties of Bitcoin.
A censorable or reversible base system is not very suitable to build powerful upper layer transaction processing on top of and if the underlying asset isn't sound, there is little point in transacting with it at all.
Things that need fixing
Eric Lombrozo, the main developer of segwit, an important component for Lightning, states in reply to a number of our questions:
“[W]e don't want to keep transaction capacity small - the problem here is that block space costs resources and adds load to the network. We'd prefer to move as much offchain as possible but we would also like to increase on-chain capacity as much as the network can bear without sacrificing security, privacy, decentralization, etc...and there are things that need fixing so we can really do this well.”
Lombrozo concludes that on-chain capacity determines how many users can join the network whereas offchain protocols allow these same users to transact without costing the network almost anything.
Slowing down the blockchain growth
Bitcoin’s blockchain currently stands at around 70GB, increasing by 5GB a month. As capacity will increase in the next few months, in the short term it may grow by 100GB a year. Leading to Satoshi’s conclusion that nodes would be run on servers.
There are many technical solutions, such as pruning or sharding or Lightning which may delay such eventuality as lower value transactions, some 70,000 daily for less than $5 according to Rusty Russel, move to the Lightning Network.
Cypherpunks’ focus is to slow down the blockchain growth, but it is not clear how the inevitable conclusion of Satoshi that nodes would be run on servers can be avoided.
The blockchain may be 200GB next year, at which point it is likely already too inconvenient for most. It will then continue to grow by 100GB a year.
Thus one can easily extrapolate, as Satoshi did, that in 5 or 10 years it is unlikely anyone would be running a node on a laptop even if the blocksize is forever kept at 2 or 4MB.
A lack of vision
In our three-day research we were unable to find Cypherpunks’ long term vision of how Bitcoin will operate in 10 or 20 years – a lack of vision which may have contributed to the intensity of last year’s debate.
Although they emphasize nodes should not be run on data-centers, we could not find any articulation of how that can be avoided in the long term.
Instead, the focus seems to be on keeping the blockchain as small as possible, for as long as possible so as to ensure that running a node, either on a laptop or server, is as cheap as possible with no clear end game if, say, a country adopted Bitcoin.
In contrast, Satoshi clearly seems to have thought that software is cheap and technology keeps improving.
Therefore, although investment to run a node may be required, it would be affordable.
Thus, Bitcoin would remain decentralized as long as there is demand for its use as service providers would be incentivized to invest in nodes.