Autopsy of a quixotic dream

20 Mar, 2023

It's hard to write about moments of mass hysteria without falling prey to hindsight bias. We try to create explanations and narratives that fit our world view. They feel obvious now, but they were nowhere near as obvious when it was all unfolding.

I'll be careful with saying that any of this was unprecedented: perhaps it was in scale and speed, but not in spirit.

I'm writing a subjective post-mortem of a collective fever dream: the cryptocurrency-fuelled speculative mania that took place in 2021-2022. To be clear: I'm optimistically skeptical about blockchain and cryptocurrencies. I believe that there is some value in blockchain and distributed systems technology. By that I mean that their applications and use-cases are not as wide as some would argue, but that they have the potential to be transformative if applied in the right context: after all, I work in the industry. I have some skin in the game. I also think that most of what came out of this recent craze is complete bullshit, a nauseating combination of techno-utopioanism, fraud and pure stupidity.

Setting the scene

Let me tell you a story. It starts in 2008, an infamous year marked by a global financial crisis. The public has lost trust in the government, banks and all the other institutions that surround them. It exposed a status-quo that only seemed to reward greed, where banks manipulated the system to serve their own interests, taking asymmetric bets where they retained all the upside and outsourced most of the downside. "Too Big To Fail", or why everyone loses but only the few get to win.

October 2008. The idea of Bitcoin is formulated for the first time, with a pseudonymous Satoshi Nakamoto dropping its white paper. It sets a vision that would permeate widely to eventually define many key tenets of the blockchain industry. The idea of a peer-to-peer electronic cash system, where trust could be entirely replaced by cryptographic proofs and economic incentives, rendering the corrupt intermediaries that plagued the current monetary and banking system entirely obsolete.

This was a compelling narrative to many, and in many ways marked the birth of cryptocurrencies, blockchains and "web3" as we know them today. Distributed systems themselves weren't exactly new and had been actively researched for a number of years prior to the creation of Bitcoin. What Bitcoin introduced was the concept of a blockchain (with some neat, albeit extremely niche technology to back it up) and the promise of a trustless, self-custodial currency and payment network.

In the first few years following its launch, Bitcoin amassed a modest but dedicated following and found some initial uses in "anonymous" online payments (to be clear: it was never anonymous, only pseudonymous). Ever heard of the Silk Road? Not the one from history books.

In 2013, Vitalik Buterin enters the picture by publishing the Ethereum white paper. It's a Eureka moment: what if you could now build complex applications on top of a blockchain? He introduces the concept of "smart contracts", the foundation of a built-in programmable layer that could take the usefulness and applications of blockchain technology to entirely new heights. To simplify: You can imagine Bitcoin as a distributed spreadsheet, whereas Ethereum is more akin to a computer (VBA might be a better analogy, if you're familiar with it). Note that "ledger" and "state machine" are the more appropriate words in this context.

A Cambrian explosion

With this leap in technology, anyone could now create their own tokens or currencies with custom rules. And that's what we saw: coins popping up at every corner, offered to the public through Initial Coin Offerings or ICOs. It was easier than ever to create magic internet money. Smart contracts allowed people to program more complex financial applications and enabled the advent of decentralised finance (or DeFi). You could now exchange tokens using Automated Market Makers (AMM). You could even lend them to other people by pooling them in smart contracts, or endlessly re-use your fairy dust as collateral.

Oh, and why stop at one blockchain? If Ethereum could be its own separate ledger, why can't we all have our own separate ledgers? And now you see new blockchains popping up everywhere. They each have their own flavour and narrative, their own set of jargon and "innovation". It's time to start adding words to the acronyms, POS becomes dPOS, NPOS, LPOS, PPOS. BFT becomes PBFT, SBFT, LibraBFT... tired of the acronyms yet? It doesn't stop here.

Wait! There are blockchains everywhere now. How do I move my tokens between them? Just build bridges with pseudo-blockchains in-between. And now too many people are using my blockchain... it's ok, we can scale them! Split the validators, try "sharding"... no even better, "Proto-Danksharding". Maybe we can layer them like a sandwich? Layer 1s and Layer 2s? Why not Layer 3s too? It's ok trust me, why be pessimistic? I'll make sure that whatever you do on my Layer 2 is reflected on that Layer 1 in 7 business days.

Anyway, my entire point is that the (fabricated) complexity increased exponentially without much of an increase in practicality and usability. Everyone wanted a piece of the pie, and wanted to sound smarter and more sophisticated than whoever preceded them. So we ended up with absolutely meaningless jargon that no-one can agree on, and pretty much nothing to show for it. Believe it or not, people even arranged in pseudo-cults they deceptively call communities to defend their own acronyms.

There is beauty in simplicity. Any practical system or idea should be easy to explain to anyone at a high-level. Most "thought-leaders" (read: grifters, scammers) in this space decided instead that they could get away with selling snake oil by adding layers and layers of convolution. It made them sound smart, and they could now say: "trust me, I know better than you", or "let me educate you". Bluntly, the need for "education" everyone keeps talking about is primarily an excuse for their inability to create or articulate intuitive systems. It's easier to convince people if you force them to adopt your arbitrary framework and constrain the conversation. If all else fails, you can always endlessly debate the semantics.

This masquerade proved to be appealing. Charlatans were rewarded based on perceived sophistication and were seldom measured against concrete results. There was very little incentive to stay simple and sober about technology when there was widespread appetite to throw money at anything that sounded remotely novel. All one had to provide was a steady stream of inflated, non-sensical promises, wrapped in imaginary words and sprinkled with the scent of easy money.

The zero-sum game and the line that goes up

It's time to address the meat of the matter: the deeply flawed, pervasive incentives and mechanics that defined and enabled much of the frenzy of 21-22. This took place in a uniquely favourable macro-economic environment, in the midst of a pandemic that drove large-scale government and central bank intervention and subsidisation. The money printer was running hot and negative interest rates went from being a theoretical anomaly to a plausible reality. The effects of this shift to extreme dovishness were not constrained to the crypto markets as they caused highly unusual behaviour in financial markets, driving a wave, or really a tsunami, of speculation and unfettered risk-taking. Put simply, things were starting to break.

By design, most cryptocurrencies, and more generally tokens, were set up to reward early adopters. It starts with giving them an early allocation at much more favourable prices (or devising mechanisms by which they can receive emissions in the early days). It is then combined with the introduction of artificial scarcity: supply-caps, theoretical "deflationary" pressure, burning and locking up tokens. The more opaque the better.

The next step is to craft a careful story around a given token and its future utility, in an attempt to attract a set of unique naïve suckers to purchase the token (at an already inflated price) and to even provide their own tokens as liquidity for swapping out of this token (read: exit liquidity for early investors). The more devious schemes will even set arbitrary rewards for parking these tokens in specific smart contracts, adding a promise of passive income on top. At the launch of the token, the liquidity is kept limited and the entire focus shifts on creating as much demand as possible, causing prices to soar rapidly.

These token economics, or "tokenomics" are now transparently referred to as "ponzinomics" by insiders (which some early investors see as a good sign; the term was widely robbed of its negative connotations in the eyes of the people who only stand to benefit from these mechanisms). They keep getting more complex in attempts to abstract away the underlying pyramidal structure, and are coated by - once again - what seems to be an infinite supply of newly produced meaningless jargon.

The particular scheme pictured above attracted over $153m in deposits, promising an eye-watering rate of return (a large number similar schemes were being created at the time with participants rotating between them. The first one, Olympus DAO, reached a peak of $5B in deposits and its runner up, Wonderland, came close, amassing nearly $3B before eventually cutting out token holders and seizing over $600m as of the time of posting).

It was obvious to many participants that this was a Ponzi scheme, but they still tried to game it. No one was in it in the long-term: the goal was just to come in early and pull out before the inevitable collapse. For anyone that was in it, their only incentive was to bring more people in (clueless or not). Admittedly, these are extreme examples, but the same underlying mechanics are ubiquitous in the crypto space.

The best way to my illustrate my point is to look at the rise of non-fungible tokens (NFTs). An extremely bland, non-existent technical "innovation" that was blown completely out of proportion. I could dive into specifics, but I've rambled enough. Instead, let me recommend "Line Goes Up" by Dan Olson. While I do not fully agree with him on every point, I think he perfectly lays out the skeptic's argument to NFTs (as we saw them used at the time. They can be useful in the right context but are ultimately a boring technicality).

The core of the issue here remains that most of the crypto-economic experiments we saw during this period were zero-sum games. They produced nothing of practical economic value. To re-iterate: there was simply no value to be returned to investors, the only inflows were those of new investors piling in.

Losing the plot: distraction in the face of existential threat

Many have mistaken this sudden spurt of growth for success, interpreting it as a sign of approval and misconstruing it as adoption. The uglier truth is that blockchains and cryptocurrencies have mostly failed to fulfil their initial goal. If anything, they strayed away from it. In an amazing display of mental gymnastics, misguided founders and developers are trying to justify new use-cases for the technology, applying it to everything they come across.

The only tokens that currently satisfy the requirements to be a currency (medium of exchange; store of value; unit of account) are centrally-issued stablecoins, pegged to existing fiat currencies. They are centralised, censorable, often unregulated and opaque. In many ways, due to the absence of oversight, they are weaker intermediaries that the ones this industry set to replace. All of the major "decentralised" or algorithmic fiat-pegged stablecoins today are mainly backed by centralised stablecoins.

There have been fiat-pegged stablecoins with no centralised fiat backing that grew in popularity. Instead, these were backed by extremely volatile tokens with no intrinsic value, beyond a promise of "governance power". Terra (UST) is the most extreme example: it was mainly backed by the value of Luna, a volatile token that was prone to speculation, and a small relative amount of Bitcoin and Avalanche tokens. Over $40B worth of UST was issued before its spectacular collapse.

Centralised exchanges, such as Binance and the defunct FTX, are the house of cards that holds the system together. They are absolutely essential today, but are amongst the most corrupt, fraudulent and opaque organisations. Instead of getting rid of intermediaries, we have created an entirely new set of corrupt yet necessary insiders that only seek to extract value from the system.

Blockchains are still an ideal playground for experimentation and provide powerful tools to digitise value and create programmable money. They have not yet proven that they are capable of creating a decentralised, or at least significantly distributed and trustless, "peer-to-peer electronic cash system". However, the large amounts of funding that flew through the industry have created an incentive to find new "niches", to differentiate, and what we are seeing today is an explosion of startups that are operating based on a false premise.

Before we build on-chain dog-walking apps and dating apps, let's maybe focus on the problems blockchains and cryptocurrencies were initially set to solve.

On centralisation and why blockchain applications often don't inherit their underlying properties

I will refrain from saying that blockchains are generally decentralised, it's a meaningless term. They are distributed, some more than others. What matters is that they have the potential to be more significantly distributed (this statement only applies to the newer consensus methods, such as Avalanche Consensus).

More importantly, this is not a property that is automatically transferred to any applications built on top of a blockchain. A considerable part of the infrastructure and tooling that has been developed to make development easier are points of centralisation. Consider oracles, which are decentralised only in name and serve the purpose of making off-chain data accessible for smart contracts. The same applies to commonly used APIs & RPC endpoints, that allow applications to query blockchain data and display it, or enable wallets to post transactions on-chain.

The smart contracts themselves might run on a distributed ledger, but they still often give special rights to specific addresses (used to upgrade contracts or halt them for example). It turns out that continued development and iteration is extremely challenging and time-consuming when working on distributed systems. There is a natural incentive for developers to retain control.

Many blockchain applications will also completely rely on off-chain inputs or centrally operated infrastructure. This is currently necessary in many cases such as asset tokenisation, complex DeFi applications requiring heavy computation or even anything that requires a proper database.

In this article, Moxie Marlinspike, founder of Signal, discusses back-end and client centralisation and suggests that achieving true decentralisation will require us to fundamentally alter our relationship to technology. It imposes trade-offs that many users might never accept.

A sober look at blockchains and where they can be valuable

By now, I hope to have picked your brain on the flaws of the industry. They can (nearly) all be attributed to a rampant, selfish desire to engineer explosive growth in the short term, with little to no regards to long-term sustainability. It is rooted deep in the culture.

It's a Faustian bargain: this willingness to sacrifice the values that underpin the blockchain space, taking shortcuts and putting growth and individual enrichment above all else sets us on a dangerous trajectory. That being said, I remain cautiously optimistic and believe that there is a clear, albeit tenuous, path forward.

The primary hurdle is to find, develop or adopt appropriate currencies. Early experiments in creating a non-government issued currency have failed but blockchains can still offer a powerful and flexible platform for experimentation. We are still far from seeing the Hayekian vision of private instances of money competing in a free-market materialise, but we have made strides on a technological standpoint. The alternative would be to create more robust systems to tokenise fiat currencies: central bank digital currencies, if implemented in a sensible way, could eliminate the counterparty risk introduced by private stablecoins.

Implementing transparent and auditable frameworks for the tokenisation of assets on blockchains can help address significant inefficiencies found in traditional means of settlement. The number of intermediaries can be reduced, and the distribution of powers and functions in the system can be clearly defined.

This will pave the way for payments-related use-cases to thrive, especially in markets where banking infrastructure might be lacking. Blockchains today have become much faster and much more efficient, allowing them to facilitate cross-border payments and in-person payments at scale with a fraction of the overhead and in a completely non-custodial environment. Of course, special care will have to be given to security, user experience and general usability. These elements will have to be developed without re-introducing centralisation: distributed infrastructure and tooling will be essential.

Distributed finance, powered by immutable smart contracts, can evolve to become more capital efficient and robust enough to provide necessary financial services at scale, in a completely non-custodial environment. This will require continued work on improving exchanges, lending protocols, futures and options markets to name a few. Regulation has to be taken into consideration here: we will need to drop the ponzinomics and build products that reflect our values if we want to convince regulators to take a sensible approach to regulating DeFi. Yes, it might already be too late.

As a closing thought, I want to bring up the topic of scale and complexity. Blockchains should be implemented in limited, localised contexts before being generalised. The complexity of a system increases exponentially with its scale. We struggle to even observe - let alone forecast - these complex interactions: replicating and codifying them in a generalised context is a doomed endeavour. The whole is more than the sum of its parts.

This will likely be the subject of another blog post, I'm still toying around with these ideas myself: from complexity theory to localism and fractal theory, observing how things change (or remain the same) with scale is fascinating.

Enjoyed this blog post? Consider subscribing:

#featured