Spamming Inside Bitcoin

There are many things wrong with the now infamous quote from Gloria Zhao (@glozow), Bitcoin Core Maintainer. In this post I will attempt to show why, exactly, it’s problematic. Admittedly, it is hard to see on first glance, so lets attempt to be systematic about it. 

The quote:

> “Like you, like if you don’t like cat photos. You don’t like wizards or whatever. That’s your choice. But I don’t think this is not a legitimate transaction. Or this is a waste of block space. Because NFTs are bad or whatever. I don’t think that language has a place. Well, I don’t think that should be considered. When you’re talking about writing policy code.”[1]

Lets first address the obvious thing that is wrong to get it out of the way so as not to be accused of any surface level analysis:

Nit: Whatever the point is, it is poorly stated. This is understandable as McCormack (the interviewer, @PeterMcCormack) is famous for a conversational style that puts people at ease. But any time we encounter a statement riddled with “like”s and “whatever”s we should seek to ignore them, and focus on the underlying points and issues. After all, it is a great joy, in both cases, to discover that someone who speaks like a hillbilly (like me) is actually brilliant (unlike me), and to discover that someone who speaks like a professor is actually mind blind. This doesn’t excuse poor articulation, of course, but it does cause us to ask, what are the underlying issues that Zhao, in this quote, is addressing?

Lets start at context. This quote is from a podcast [1] and the context is in regard to a section titled “inscriptions (spam) in mempool”[SIC]. The conversation between McCormack and Zhao revolves around the role of developers and maintainers in managing Bitcoin’s blockchain, particularly regarding transaction types and “spam.” Zhao explains that block space is a scarce resource, and its price is driven by demand, using economic principles. McCormack asks if developers should remain apolitical and focus solely on ensuring the software functions. Zhao argues that developers shouldn’t judge the legitimacy of transactions based on their use case (e.g., NFTs or “cat photos”), as defining “spam” in transaction relay code relates to resource usage, not content. McCormack acknowledges the decentralized nature of Bitcoin allows nodes or miners to make choices but expresses a personal preference for Bitcoin to focus on financial transactions, citing concerns about non-financial uses like NFTs potentially harming Bitcoin’s purpose.

In short, the heart of this clip is about the *purpose* of bitcoin. We can not define “spam” without defining purpose. McCormack says, directly after the infamous quote above, “I’ve ultimately come to the point that my preference would be Bitcoin was just used for financial transactions… of Bitcoin being money.”

Now, here is where it gets interesting: Both parties agree that this is the purpose of bitcoin. Zhao has been quoted many times saying (in various ways) the purpose of bitcoin is to “enable anyone, anywhere in the world, to send a payment.”[2] So what is the problem here? The problem here is that Zhao defines “spam” as the inappropriate or unauthorized use of network resources, such as computation or memory, rather than judging based on transaction use cases (e.g., NFTs or inscriptions).[3] But this definition is wanting and from it we get all our problems–it is missing the forest for the trees. There are several ways to define “spam” in Bitcoin that apply to consensus-valid transactions but do not fit Gloria Zhao’s definition.

So, let’s get into it.

1. Difficulty the First: This is a definition of spam that is far too limited. Here are some other definitions:[SEE TABLE]

2. Difficulty the Second: Now, many people defend Zhao’s view, and are quick to point out the problems in the above views. And they do exist. This was best articulated (as usual) by Andreas Antonopoulos speaking about “spam” transactions in 2016.[11] To paraphrase, there are two ways to define spam: 1. Top Down (clearly defined legitimate use cases), and 2. Market Driven (whoever pays the fees is not spam). The first way is problematic in that it enables the very system which is intended to fight centralize control, to be centrally controlled. This is where proponents of this view will be quick to point out limiting use cases is censorship, “distributed authoritarianism”[12], and the like. But like many of the best dilemmas, there is a third way poorly addressed by either extreme.

3. Difficulty the Third: So lets look at this third way. In information theory, a low-entropy, high-fidelity carrier signal (the blockchain) is essential for reliably transmitting high-entropy (surprising) information (high value txs). Noise or unpredictability in the carrier introduces errors, reducing message trustworthiness. Steganographic content–hidden information in open data (arbitrary content in bitcoin)–disrupts systems requiring low-entropy carriers, potentially impairing their ability to deliver surprising information. Since user-defined entropy (required for PKI) makes steganography unstoppable (in bitcoin), we can define steganography as spam (Please read this additional piece [13] to get a better sense of this). Neither a Top Down, nor a Market Driven approach address this issue:

 –  The top-down approach conflate a tx from A to B with amount X+arbitrary data and the same tx from A to B with amount X sans arbitrary data, erroneously assuming that we must BAN BOTH txs. 

 – The market based approach *also* conflate a tx from A to B with amount X+arbitrary data and the same tx from A to B with amount X sans arbitrary data, erroneously assuming that we must ALLOW BOTH txs.

 – This third ways does not conflate these two transactions: given a tx from A to B with amount X+arbitrary data and the same txs from A to B with amount X sans arbitrary data, ban the one with arbitrary data, allow the one without it, in so much as you can be sure a sufficient amount of entropy is protected. 

This is not censorship of monetary transfer, nor is it sanction of ill intent. If we do not define steganographic content as spam from the perspective of the free market, why define valid transactions at all?–just make everything valid, even malformed txs, not my business! Total Laissez-faire (and wildly problematic). If we do not define steganographic content as spam from the perspective of top down, why run bitcoin at all?–after all, the purpose can be corrupted, so there is no purpose to it!

4. Difficulty the Fourth: Steganographic spam has three ways to be dealt with: 

 – A. Redefine system purpose: Allow steganographic content, risking conflicting uses.

 – B. Restrict system purpose: Limit steganography by restricting use-cases or user entropy within sufficiency restraints.

 – C. Ignore the problem: Viable only if carrier signal entropy degradation is unproven; otherwise, system collapse may be likely.

In the case of Bitcoin, we can prove that Steganographic spam bloats the blockchain, so we can eliminate C. This leaves B and A. And, I think, that with regard to the debate, we can evenly put Zhao is the A camp. By limiting a definition of spam to not include steganographic content (but instead strictly stick to market based), we definitionally must include a cross purpose of the protocol as a condoned purpose of the protocol. In our case, setting the default for OP_RETURN to unlimited to carve-out a space for arbitrary data. And this is deeply problematic:

> In logic, a system’s telos (inherent end) cannot sustain contradictions without collapse, as per non-contradiction (a thing cannot be and not be simultaneously). Thus, Option A–redefining the system to accommodate steganographic content–is untenable. Steganography involves information used against the system’s purpose. Permitting this content makes this antithetical purpose legitimate, no longer steganographic, but forces the system to adopt a contradictory purpose in addition to its original purpose–it asks the system to both go forward and backward simultaneously. A system cannot pursue opposing ends without losing coherence.[13]

5. Difficulty the Fifth: From this Fourth Difficulty, we get the social difficulties and the reason this debate has become so heated. The first item lost is the intention. If we do not define steganographic content as spam, then we must define it as legitimate use, and we lose the defense of Mens rea–”Guilty Mind” we are intending to use software for purposes contrary to the software, *as defined by whoever puts content on chain*. This is a moral and ethical failing.

6. Difficulty the Sixth: This introduces a clear rejection of reason, and introduction of anti-Aristotelian logic, post-modernism, Newspeak, and general Communist Retardation. As I have noted elsewhere[14], the now infamous quote codes exactly like a line out of Animal Farm. The attempt to be “apolitical” is itself a political stance, and self-undermining. A is A.

7. Difficulty the Seventh: Failing these ethical, moral, logical, and political purposes, we then, finally, get to introduce legal consequence: 

> “There is no excuse for the Bitcoin reference client to expose us to the wide variety of extreme legal risks that come with arbitrary content.” – Nick Szabo (@NickSzabo4)[15]. 

There are more difficulties of course, but this is all the time I have for this today.

Fund of Founders

Founding Stanford EIR

@duck9 alum, Deeply Understood Capital Credit Chinese Knowledge 9

Solo Founder Uber API

650-566-9600 Office

650-566-9696 Direct

Cell: 415-720-8500