Ethereum's Strawmap: Frankensteinzation, Complexity, and Liveness Galore Part 2

Ethereum faces two distinct forms of complexity. In a previous article, I examined its architectural complexity: the combination of Casper FFG and LMD-GHOST to solve proof-of-stake’s nothing-at-stake problem. The Strawmap introduces a different form—statistical complexity—as Ethereum brings more stateful and interconnected economic activity onto its base layer. This article examines the tradeoffs between Bitcoin’s low-complexity model and Ethereum’s increasingly complex one. Is Bitcoin too simple for the economic system it seeks to replace? Can Ethereum successfully internalize that complexity without becoming unstable itself? And ultimately, will the Strawmap succeed?

Frankensteinization Aka Complexity 

What is complexity and how is it measured? The best explanation applicable to blockchain technology is Renato P. Dos Santos’ article, “On the Philosophy of Bitcoin/Blockchain Technology: Is It a Chaotic, Complex System?” 

There is no universally accepted definition of complexity. Dos Santos therefore adopts statistical complexity, developed by James Crutchfield and Karl Young, which measures the structured relationship between a system’s past and future. A perfectly ordered system has low complexity because its future is obvious, while a perfectly random system also has low complexity because its past tells us nothing about what comes next. Complexity emerges between these extremes, when what happened in the past changes what is likely to happen in the future.1

Suppose you have three systems: A, B, and C. System A simply repeats A A A A A A A A; its future is obvious and little needs to be learned from its past. System B consists of independent random coin flips; its future is uncertain, but its past provides no help in predicting it. System C, however, is history-dependent: A affects B, B affects C, and C feeds back into A, so different histories can produce different probabilities for what happens next. Unlike order or randomness, C therefore contains meaningful predictive structure.

Complexity should not be confused with complication, randomness, computation, or information storage. An atomic clock can be extraordinarily complicated to construct yet highly predictable, while a mechanically simple double pendulum can exhibit chaotic behavior. Likewise, randomness is not complexity: independent coin flips are unpredictable but contain no predictive structure. Nor does the amount of computation or stored information determine complexity, since neither tells us whether a system’s past meaningfully influences its future behavior.

To measure complexity, Dos Santos utilizes the mathematics of Crutchfield and Young's ε-machines. An ε-machine measures complexity in two steps: (a) it identifies the different conditions a system can be in that change what is likely to happen next; and (b) it measures the uncertainty over which of these predictive states the system occupies, with greater uncertainty corresponding to greater complexity.

For (a) imagine three different histories:

History 1: 70% chance of X, 30% chance of Y
History 2:  70% chance of X, 30% chance of Y
History 3:  20% chance of X, 80% chance of Y

The ε-machine groups histories that yield the same prediction together, such as Histories 1 and 2, and separates those that change the prediction, such as History 3. Only histories that change the prediction matter for measuring complexity.

For (b), suppose (a) gives us two causal states, with State A occurring 50% of the time and State B occurring 50%. Because the system frequently occupies both predictive states, its statistical complexity is higher. If instead State A occurs 99.9999% of the time, the system almost always occupies the same predictive state, so its complexity is very low. If State A occurs 100% of the time, its statistical complexity is zero.

Low Complexity Money-Forms: Bitcoin’s Internal Contradiction as Imminent Self-Destruction 

If we think about complexity in Bitcoin’s case, it’s modeled as having two causal states: State 1, where miners search for a valid hash, and State 2, where a valid hash is found and a new block is added. Because Bitcoin spends overwhelmingly more time in State 1 (e.g., 99.9999%) than in State 2 (e.g., 0.0001%) due to its brute-force mining mechanics, its statistical complexity is extremely low. Complex systems commonly exhibit nonlinearity, in which interactions produce disproportionate effects; emergence, in which interactions create new higher-level behaviors; and hierarchical growth, in which components organize into higher levels of structure. Bitcoin’s blockchain lacks these features as blocks repeatedly follow the same processes without creating new forms of behavior or organization. The blockchain grows larger, but not structurally more complex. 

Bitcoin maximalists praise the simplicity and low complexity of Bitcoin’s mechanics as a major achievement for immutable, sound money. It cannot be changed, and it enforces a strong sense of unforeseeable costliness. Because of this, Bitcoin instead externalizes complexity onto the wider economic system around it. This sounds ideal; however, offloading complexity into the wider economic system, where linearity, hierarchical growth, and emergence are extreme, causes significant problems. However, low internal complexity does not automatically create external complexity. Complexity is pushed outward only when there is substantial demand for functions that the simpler system cannot provide internally.

This was the case for the historical gold standard. Due to gold’s inability to achieve salability across space, i.e., ease of transacting over distance, banks introduced a custodial infrastructure and a fractional-reserve system via paper gold, thereby causing circulation-credit expansion. Banks didn’t have a choice - gold was too slow and expensive to transact and transport across distance. On top of the banking sector, we had speculative industries driven by animal spirits. This mismatch of a simple, low-complexity money and an external, complex system caused economic disaster through bank runs and the Great Depression. 

Bitcoin, like gold, cannot achieve salability across space. Supporters argue that, given its digital nature, transactions can move across space much more easily than gold can. However, space doesn't matter if the time to transact is too slow (e.g., 10-minute block confirmation times) and too expensive. We are left with the same difficulties of gold - external, destructive structures to scale. Fractional-reserve Bitcoin is just around the corner, as instant-settlement BTC will command a significant premium once settlement fees become impossible for the retail user to secure. However, Bitcoin faces an existential threat greater than gold's because mining economics largely depends on the price of Bitcoin itself; if the network hash rate drops due to external forces, Bitcoin can cease to exist. If mining stops for gold, gold is still gold. Moreover, Bitcoin faces an existential risk that gold does not: its network depends on miners continuing to operate, which, in turn, depends heavily on Bitcoin’s price and mining economics. If external forces caused hashrate to fall far enough, Bitcoin’s ability to function could eventually be threatened. Gold, by contrast, does not depend on an operating network—if gold mining stops, existing gold remains gold.

Modern derivatives amplify these external pressures—effectively “paper Bitcoin”—and more sophisticated fractional-reserve lending within private and shadow banking systems. This creates an even greater mismatch between Bitcoin’s internal simplicity and the complexity of the external financial system surrounding it than existed between gold and its surrounding financial system. 

Lastly, outside the traditional financial layer, Bitcoin’s low internal complexity and the demand for more complex functionality give rise to Layer 2 services that are parasitic on base-layer fee accrual while introducing hierarchical structures that lack censorship resistance via centralized sequencing and bridging mechanisms. This creates a contradiction. Bitcoin’s inability to support this complexity internally makes its simplicity self-defeating, as it becomes reliant on the very external structures its sound-money design was meant to avoid.

Ethereum’s New Vision to Re-internalize Complexity 

Ethereum’s Strawmap solution is to internalize the complexity of the wider economic world. You cannot provide decentralized finance if your protocol (i.e., Bitcoin) relies on the centralized world around it; it must be recreated from within the protocol itself. The Strawmap realizes this by redirecting scaling from the L2 to the L1. L2s will no longer be parasitic to Ethereum’s economic base layer; instead, they will function as customizable layers for institutions, private networks, and compliance-first platforms.2  Rather than relying on external hierarchical, emergent, and non-linear forms, Ethereum can now reinternalize these forms in a decentralized manner once again. Thus, Ethereum is less likely to experience contradictory self-destruction like Bitcoin will.

However, reinternalizing complexity is a double-edged sword, as it comes with many potential problems. First, reinternalizing complexity provides composability (bringing back the philosophy of financial LEGO-block primitives), but this comes with several costs: greater systemic coupling and a greater shared blast radius. Systemic coupling is the interdependence between parts of a system, whereby A affects B, B affects C, and C feeds back into A, so different histories can produce different probabilities for what happens next. This creates the meaningful predictive structure that Dos Santos identifies with complexity. Decoupling leads to a greater shared blast radius, i.e., the extent to which a failure in one part of a system spreads and affects the rest of the system. One advantage of pushing complexity outward is containment. If an L2 has a catastrophic bug or becomes unavailable, Ethereum itself can continue operating. However, if more economically important execution is concentrated in the L1, pathological workloads and unforeseen interactions can affect the infrastructure everyone shares - paired with massive coupling, this can cause destructive feedback loops that amplify the original problem (or at the very least significantly increase the chances).

Lastly, this creates a new problem: social governance complexity. If the consequences of a destructive feedback loop become so severe, we face the same dilemma that emerged from Ethereum’s 2016 DAO hack and the hard fork that followed. Do we accept the economic consequences and let the outcome stand because “code is law”? Or do we intervene because the consequences are too severe to ignore? These social dilemmas bring us back to the interventionist debate between Friedrich Hayek and John Maynard Keynes that decentralized systems were supposed to help us escape. In other words, as Ethereum internalizes more of the complexity of the financial system, it may also internalize the same questions over when—and whether—we should intervene when that system fails. The Strawmap leaves us with an unresolved tradeoff. Bitcoin may be too simple to internalize the complexity of the financial system, but Ethereum may discover that internalizing that complexity brings the financial system’s instability with it. Whether the Strawmap can manage that complexity without becoming overwhelmed by it remains an open question. Perhaps the answer lies somewhere between these two extremes, in a hybrid architecture such as Kaspa’s based rollups. We shall see. 

[1]

The chart above illustrates this, whereby C = statistical complexity, i.e., how much structured information or organization the system contains; and H = entropy — how unpredictable or random the system is. Complexity tends to occur between perfect order and complete randomness.

[2] I will not break down an ε-machine model measuring potential complexity scores, as the Strawmap and its mechanisms aren’t fully worked out yet.