An essay on time preference and building future-proof blockchains.
The remarkable speed of change in the blockchain industry has increasingly been steering the space toward high time preference thinking. For investors and builders, chasing the “new meta” has supplanted more long-term oriented goals like building the future of finance and onboarding the next billion users to Web3.
This way of operating is completely opposed to the Nervos ethos, which has largely remained misunderstood and underappreciated. Our pursuit of the original cypherpunk goals has made us outliers in the industry, which is something we never imagined we’d become. Nevertheless, we believe pursuing these goals relentlessly is a noble and worthy endeavor, and we’ll explain why.
The scarcity of time implies scarcity of options, forcing all of us to choose between different alternatives constantly. All decisions have an opportunity cost, and even if there were no restraints on the number of alternative resources available, our decision on how to spend our time necessarily eliminates all other possible options for how we could've spent that same time.
Like a skill tree in an RPG game, every decision we make opens up a new branch of possibilities and closes all others. However, unlike in RPG games, economizing time in real life is unique because time passes and cannot be stopped or reversed. There are no checkpoints or restarts in real life, but only final decisions that can permanently alter the outcomes in our lives.
This makes time a commodity like no other. Because we can't choose between different quantities or periods of time, valuing time is more problematic than valuing any other free market commodity. The inherent uncertainty of the future makes us biased toward the present. After all, the future can only come through successfully securing survival in the present, making the needs of the present always more important than those of the future.
In economics, this phenomenon is explored through the conceptualization and study of time preference. Time preference refers to the degree to which individuals value the present compared to the future. High time preference implies heavy discounting of the future in favor of the present, while low time preference indicates lower discounting of the future or greater appreciation for it.
All living beings have a positive time preference, meaning they always discount the future compared to the present, albeit to different degrees. For example, animals have a much higher time preference than humans, as they act on their immediate instinctive impulses with little conception of the future. While some animals may be capable of building shelters that can last for the future or save a portion of their food for future consumption, their time preference still remains significantly higher than that of humans.
This is because humans are the only species that can arrive at their time preference intellectually rather than merely instinctively. Through logic and reasoning, we can compare future to present needs with much more nuance and therefore value them more appropriately. This capacity is, in fact, the mother of all capital creation. By reducing time preference, humans can carry out tasks over longer time horizons and produce goods not only for immediate consumption but for producing future goods—i.e., capital goods.
Once time preference drops enough to allow for any savings and capital formation, the process of civilization is initiated.
One Marshmallow Now, or Two Later?
The differences in time preference vary not only between species but also between individuals, with the impact on their lives being significant. This is best showcased in the Stanford Marshmallow experiment from 1960, where psychologists left children in a room with a marshmallow and told them they were free to eat it whenever they wanted. However, if they waited 15 minutes, they’d receive a second piece as a reward. In other words, the kids were given a choice between the immediate gratification of one marshmallow or delaying gratification and receiving two marshmallows.
The interesting finding from the experiment is that the kids who were able to delay gratification were found to have had better life outcomes, including higher SAT scores, lower body mass index, and lower proclivity to abuse drugs decades later. Beyond this experiment, everyone has seen and likely personally experienced the power of delayed gratification, either by forgoing indulging in tasty foods today to obtain the ideal beach body tomorrow or by forgoing excessive spending and consumption today to invest and build the desired retirement nest egg for tomorrow.
Sacrificing today’s needs and wants for future compounded gains is the essence of all progress. Yet, somehow, the crypto industry seems to have forgotten that, with high time preference thinking becoming the default among both blockchain and decentralized application developers. And while application development is generally more present-oriented and forgiving, building blockchains—especially Layer 1s—with the same high time preference mindset leads to unwanted outcomes that hurt the entire industry.
For example, this high time preference mindset gave us several Layer 1s that sacrificed security and decentralization for high throughput and resultantly experienced unplanned outages and deliberate halts during the last bull run. In a rush to capture ever more mercenary users flocking from Ethereum’s high transaction fees, these projects chose to eat the marshmallows immediately instead of delaying gratification and receiving two later. In other words, they captured a significant market share of users and developers at the time but sacrificed the long-term viability of their technology.
For us, “blockchains” that can be halted and restarted at will are oxymorons. Blockchains are, by definition, decentralized and immutable databases. If they’re maintained by only a handful of colluding validators and can be tampered with, then they are not blockchains but merely centralized inefficient databases. The high time preference thinking, therefore, led many crypto projects to recycle the old flawed, exclusionary, and vulnerable systems encumbering traditional finance.
Building Blockchains With Low Time Preference
Our low time preference approach to building blockchains is precisely what separates us from most Layer 1s in the industry today. Instead of rushing to market with an uninspired and unrefined technology stack, we took our time to design a blockchain that genuinely cracked the key problems saddling the existing solutions.
Resultantly, Nervos introduces the first blockchain architectured with the modular framework in mind from the get-go and the ground up. We realized very early on that the only way to scale to global ubiquity without sacrificing security or decentralization wasn’t through a monolithic architecture but a layered one. With most blockchains now embracing it, this design choice seems ordinary today, but at the time of our launch, it was completely unorthodox. More importantly, our low time preference paid off, as now our modular thesis is being confirmed and embraced by the broader industry, and we’re better positioned to bank on it than everyone else.
The decisions to base the CKB-VM on the novel RISC-V computer instruction set and develop the Cell accounting model were the results of long-term oriented thinking. The Cell model and RISC-V, in conjunction, make CKB the most flexible and interoperable blockchain in the industry, enabling support for all cryptographic primitives and building all kinds of novel applications that aren’t possible elsewhere. Because of these features, Account Abstraction, a notorious choke point for today’s blockchains, is possible in CKB by default.
However, to attain these characteristics and future-proof CKB, we had to make certain sacrifices in the present. For one, we alienated a notable chunk of the existing application developers accustomed to coding in Solidity and building on Account-based chains. The RISC-V-based virtual machine allows developers to build on bare metal, while the Cell model—which lends itself to a more functional programming style—grants them unlimited possibilities. While this is the only way to build a truly future-proof blockchain, it complicates onboarding less tech-savvy application developers. That being said, CKB supports all programming languages, meaning we can draw from a much bigger pool of developers. When the next generation of developers unencumbered by pre-existing Account model-based notions comes, we’ll be here to host them.
Our low time preference thinking also led us to choose the Proof-of-Work consensus mechanism over Proof-of-Stake, despite its waning popularity and the growing stigma surrounding its energy use. We knew that defending Proof-of-Work to the uninitiated would become increasingly harder, but we also knew that this consensus mechanism would provide CKB with superior security and decentralization. Instead of succumbing to the current societal and political pressures, we picked what was best for CKB in the long term. The correctness of this decision is already being proven today, as we see Proof-of-Stake-based chains becoming ever more centralized and vulnerable to capture and censorship.
Today more than ever, our ethos and approach to building blockchains appear fundamentally different, with the key differentiator being our significantly lower time preference. We're unapologetically sticking to crypto's original goal of creating a robust, open, accessible, and permissionless financial system. And the only way to get there is by forgoing some of the needs of the present and focusing on the needs of the future.
While much more work remains to be done, we're among the few on the right track. By lowering our time preference, we were able to get the fundamentals right and create the ideal, future-proof base layer for a modular blockchain network. Now, it's time to build on top.