Guns and butter

The blockchain design space bares some interesting resemblances to statecraft, with similar tradeoffs between features (like throughput) and security.

Dana J. Wright
11 min readMar 15, 2022


Image created by the author.

Haseeb Qureshi has a great thought experiment where he imagines if the entire Bitcoin network was just one giant Postgres database run by some dude in Canada.

In the experiment, Bitcoin looks exactly the same as it does right now. The monetary policy, the halving of block rewards every four years, the scarcity, all the same stuff. We still have China bans, we still have lazer eyes on twitter, we still have the lore of Satoshi, and we still have “the inexplicable link between Bitcoiners and carnivorism.”

Everything is the same, except for one difference: All the miners, the block explorers and the on chain metrics dashboards, they’re all just pinging the Canadian guy’s server. Whenever someone mines a block, they send it to the Canadian guy, he inserts the block into his database and “broadcasts” it to everyone else.

Haseeb then poses the question: How would this change Bitcoin?

The answer: Not at all, as long as it remains a secret.

In this hypothetical, bitcoin is just as fungible, portable and scarce. For all intents and purposes, it’s still “better than gold” and technically works the exact same way.

With one material difference: If it were ever discovered, the network would be completely defenseless against an attack. At any moment, the government could simply kick down the guy’s door and turn off Bitcoin.

Yes, and you’re under arrest.

Defense systems

The thought experiment teaches us something important about blockchains: Resistance to attack is the defining characteristic. Without robust defense systems, a cryptocurrency would just be a database. Like the ones that manage gift cards or frequent flyer miles.

Indeed, the first few attempts at creating an open, internet-based money were essentially that: DigiCash, E-gold and Liberty Reserve had many of the same properties of Bitcoin, but each was either shut down by the government or fatally flawed in other ways.

Satoshi didn’t build Bitcoin’s defense system for shits and giggles. Bitcoin is decentralized because it must be to remain protected from censorship, corruption and attack.


So the primacy of blockchain defense is not really up for debate. Without adequate defense, all other advantages are mute. Nation states must have militaries to defend their wealth, so too must blockchains.

But this feature does not come without a cost. Defense systems are in fact the most costly feature of any blockchain.

Chains pay their defense budgets in one of two ways: Issuance, where new tokens are minted and paid out to the military. Or through transaction fees, which are also paid to those who secure the network. For Proof of Work (PoW) blockchains, the miners are the military. For Proof of Stake (PoS) blockchains, the validators are the military.

Those are the two mechanisms. Every blockchain uses one or the other. Their relative merits and demerits are the source of some of the most contentious debates in the crypto space.

I’m not going to delve too deeply into those here, except to say that while they are technically very different, conceptually they achieve the same ends: Distributed enforcement of the network laws (consensus) and verification of network activity with no single point of failure.

Guns and butter

As was demonstrated with the Canadian guy, a robust defense system (or lack thereof) is completely invisible.

If it does its job well, the only evidence that it exists is the absence of catastrophic failure; your transactions always eventually go through, the network never gets taken out by a 51 percent attack, or a frontrunning vulnerability, or a fee sniping attack.


If security was the only feature we cared about, that would be the end of the discussion. Other chains would fail to improve upon Bitcoin’s security model and we would eventually have “one chain to rule them all.”

But it turns out defense is not binary. It’s a spectrum, and there are serious tradeoffs.

If you need the network to survive targeted attacks by three letter agencies, then defense is paramount. But many applications don’t actually need that level of defense for what they are trying to do. They benefit from some level of decentralization, but they also need transactions to be fast and cheap. And they want to be able to run far more complex operations on chain than UTXOs.

Those more “buttery” features are diametrically at odds with the strong defense system offered by Bitcoin. And pretty much impossible to build on Bitcoin, given the sparse, incomplete scripting language Bitcoin uses to minimize the amount of memory it takes to run a node.

Enter Ethereum.

A PoW blockchain designed to house smart contracts. With Ethereum, we saw the first attempt to tweak the guns to butter ratio. Ethereum emulated Bitcoin’s decentralization model but introduced a new language(Solidity) and framework (EVM) that made it possible for developers to execute more sophisticated operations on chain.

By copying certain aspects of Bitcoin and changing others, the early Etheruem community opened up a huge can of worms as they tried to determine what decentralization really means. Does it mean to have lots of nodes running the software, and for those nodes to be distributed across the globe? Does it matter if the hash power of miners is concentrated amongst a small group? What about Ethereum’s native coin, Ether (ETH)? Do Ethereum whales pose a threat to the network?

The decisions they made on these questions were put to an early test in June, 2016 when a thief made off with $50 million of ETH.

The infamous DAO hack happened shortly after the Ethereum community made history with the largest crowdfunding campaign ever at the time. More than $150 million in ETH raised from about 11,000 investors. The DAO was hoping to become an investor-directed venture capital firm, a brand new kind of entity made possible only by the Ethereum network.

The hack was an existential threat to Ethereum. The story goes deep, but I’m going to skip to the end: After a huge debate, The Ethereum community decided to roll back the blockchain and restore the lost funds to The DAO smart contract. While this solution made investors whole, it exposed what many believed to be a far more serious vulnerability in Ethereum: The governance and protocol development process itself was centralized.

The network architecture may have looked like Bitcoin, but there was an inner circle of ETH token holders who could unilaterally roll back the blockchain when it suited their interests to do so.

Not cool guys.

A multi polar world

With all the possibilities for how to design a blockchain protocol, it was perhaps inevitable that new networks would emerge, offering new technical approaches and new governing philosophies.

Indeed, we now live in a multi polar blockchain world.

A network can be decentralized when it comes to its token distribution but at the same time run on a centralized infrastructure. Binance Smart Chain is a good example of that.

Another network could be decentralized when it comes to number of nodes and physical locations of nodes, but centralized when it comes to governance. Cardano is a good example of that model.

On each chain, communities are forming and ecosystems are being built. They have their own leaders, their own sets of ideas and use cases that they prioritize over others. This then becomes their brand in the same way nation states market themselves to the rest of the world.

The frontier

If Layer 1 chains like Ethereum, Avalanche and Solana are nation states, then Layer 2 is the frontier. They go by many names: Rollups, wrappers, subnets, parachains, sidechains, etc. But basically a L2 is a mini blockchain that obeys the underlying rules of the parent chain.

Conceptually, L2s map pretty closely to the way the traditional financial system in the US works, where we have Layer 1: Fed Dollar, Layer 2: the ACH network, and Layer 3: banks, credit cards and finance apps like PayPal.

I could easily see Layer 2 becoming the go to solution for protocols that seek the security of final settlement on Layer 1, but with the speed and cheap transactions necessary to onboard the masses.

Indeed, rollups are already making enormous strides. Subnets on Avalanche are showing promise as well.

But where there is opportunity, there is danger.

The largest Ethereum L2 chain measured by Total Value Locked (TVL) is Polygon. Several decentralized applications built on Polygon suffered high profile exploits in 2021, losing tens of millions.

So while they may be easier for developers to build on and they may provide fast and inexpensive transactions, it remains to be seen whether L2s can provide the security assurances most people will need to consider storing significant value there.

Guardians with no teeth

Apart from scaling individual chains, the multichain world opens the door to yet another front in the blockchain development space: The much hyped idea of cross-chain compatibility.

Essentially, making it possible to move value and information between different blockchains. Of course, this is already possible. You can move assets between different networks on centralized exchanges like Coinbase or Kraken. But it’s a pretty crude method. Not all networks are supported, you pay fees to the exchange and you can’t do it programatically.

Several protocols such as POA, Cosmos, Polkadot, Komodo, and Ark are trying to facilitate interoperability in a trustless manner using different consensus mechanisms that would eliminate the need for an intermediary.

But it has proven to be a formidable challenge. As Layer 1 chains build up their own communities, economies, rules and processes, it inherently becomes more and more difficult to exchange information and value between them.

While Polkadot and Cosmos work on building comprehensive solutions, a bunch of individual decentralized application have stepped in to facilitate the movement of assets between chains.

These are known as bridges. Every major chain has at least one bridge and some chains have many bridges.

Bridges use wrapped tokens, which are tokens from one blockchain that are locked into a smart contract on a different chain. The smart contract gets the asset data from a cross-chain oracle called a “guardian,” which certifies that the coins have been properly locked on the parent chain. The bridge then mints or releases tokens of the same value onto the other chain.

And surprise surprise, bridges are perhaps the least secure category of decentralized applications, with losses totaling around $2 billion so far from various hacks and exploits.

Polychain was hacked in August, 2021 in the largest exploit in crypto history measured by amount of money stolen (about $600 million). The Solana x Wormhole hack from late January, 2022 resulted in about $340 million extracted through a vulnerability in that bridge’s code. And Rari Capital, another cross-chain protocol was hit by an attack in May of 2021, losing about $11 million.

In his lengthy analysis posted on Twitter a few hours after the Wormhole attack, samsczun explained how their backend failed to properly validate its guardian accounts.

Basically, Wormhole didn’t properly validate all input accounts, which allowed the attacker to spoof guardian signatures and mint 120,000 ETH (worth about $323) million on Solana. They then bridged about 93,750 tokens into a private wallet stored on Ethereum.

“As evidenced by all the exploits we’ve seen, cross-chain is a very hard area … with the added complexity of connections with every other chain and all their idiosyncrasies,” according to Ryan Watkins, a research analyst at blockchain data firm Messari.

The bridge dilemma

In a recent post on Reddit, Vitalik warned that “fundamental security limits of bridges” made them particularly vulnerable to a 51% attack.

This kind of attack opens the door to something known as double spending, which allows the attacker to make two or more payments with the same tokens.

“No one will 51% attack Ethereum just to steal 100 Solana-WETH (or, for that matter, 51% attack Solana just to steal 100 Ethereum-WSOL),” Vitalik writes.

“But if there’s 10 million ETH or SOL in the bridge, then the motivation to make an attack becomes much higher, and large pools may well coordinate to make the attack happen.”


So basically, the demand for moving assets cross chain is very real, the narrative of blockchain interoperability continues to gain steam, and the pressure is on for protocol developers to ship fast.

But the more bridges there are and the more they are used, the greater the incentive is for nefarious pools to coordinate and 51% attack the bridge.

The mineshaft gap

If you haven’t seen Stanley Kubrick’s 1964 film Dr. Strangelove or: How I Learned to Stop Worrying and Love the Bomb, you should.


In the film, the Soviet Union implements a dead man’s switch that, in the event of an attack, automatically detonates fifty H-bombs in the hundred megaton range and jackets them with cobalt thorium G, producing a doomsday shroud which kills all human and animal life and renders earth uninhabitable for 100 years.

I do not know what the equivalent of that situation would be in crypto, but given our track record so far I would not put it past us to find out. Much like the Cold War contest between the United States and the Soviet Union, crypto is narrative driven phenomenon. And deeply tribal.

At the end of the film, after the doomsday machine is triggered and the generals have conceded that earth will be destroyed, they begin discussing plans for how the survivors can continue living underground long enough to repopulate the planet.

As each country divulges their respective underground capacity, one of the generals exclaims that they must not allow for a “mineshaft gap.”

This is of course a play on the “missile gap,” a term coined by John F. Kennedy to suggest that the nation with the most nuclear missiles was the most powerful, and to convince the American people that the US must ramp up production.

In this conversation though, the balance of power was no longer defined by the number of nuclear missiles each country had. But rather by the capacity of underground shelters each country had that could keep the few survivors of those nuclear missiles alive.

I don’t know what the equivalent to that scenario would be in crypto, but I wouldn’t put it past us to find out. Crypto has some of the brightest minds of my generation working on it, as well as some of the most maniacal.

This technology has the power to solve some of our most pressing problems. But like any great power, it has a dark side that can also be harnessed to do great evil.

And we ignore that part at our peril.

Thanks for reading until the end. I work in crypto and think about it non-stop. You can find me on twitter @dappbeast



Dana J. Wright

Designer · dreamer · dilettante · djw.eth