How Proof of Work (Bitcoin) is Wrong

April 1 2026

(This is a blog post transcribing a previously published video (25th of Sept. of 2025, see the companies' socials))

1. What is Proof of Work?

So what is proof of work? Here we are considering it in the context of Bitcoin basically because of the relevance of Bitcoin. Proof of Work is a small but nevertheless important part of the consensus mechanism of the Bitcoin protocol. Basically, miners compete to solve cryptographic puzzles, and the first to succeed earns the right to add a block. The purpose is to make altering the blockchain costly, since it would require redoing all of this so-called “work.”

The question here is: what exactly is meant by “work” in this context? Why use this word and not others? The word also invites a discussion of the foundations of Bitcoin. This discussion is relevant because of the relevance of Bitcoin and the observed public discourse (scientific institutions, technological centers, companies, financial institutions, governments, etc.).

2. Possible Interpretations

There are basically two possible interpretations. These are the two most straightforward.

There is the physical interpretation of work.

There is the economics/production/labor interpretation of work.

3. Proof of Work in the Physical Interpretation

First, I am not a physicist, but the details are not important here.

In physics, work has been defined, and the argument is that since Bitcoin mining involves running computations on digital electronic computers, then this qualifies as work because it consumes (electrical) energy. But this interpretation has issues.

Even in physics, the concept of energy is not universally defined. There is no single, general theory of energy (abstract, general, without adjectives). Instead, energy appears in different forms (gravitational, electromagnetic, ...); unified mainly through conservation laws. So even if we accept the reduction of work to energy, the term energy is not completely defined in physical terms.

In other words, given that the concept of energy is already problematic, if we put bitcoin on top of that, so to speak, we get greater confusion. Maybe an acceptable naming would be Proof of Electrical or Electromagnetic Energy. But even this has some challenges I will talk about in a couple of minutes.

4. Proof of Work in the Production/Labor Interpretation

Another possibility is to interpret “work” in the economic or production or labor sense of the word. Mining certainly involves costs: producing hardware, producing electricity, and employing people to maintain operations.

But here, too, the connection is not straightforward. Energy costs, for example, vary greatly depending on location, infrastructure, and regulation. If Proof of Work is meant to prove that energy was spent, the variation in how energy is priced and sourced makes the “proof” highly contingent, and more generally the variation of economic conditions. In other words, if the concept of energy interpreted in the physics domain is problematic, the concept of energy understood as a sector of the economy or as some economical concept is even worse (in the scientific sense). The content of the proof is not really clear, the concept is not clearly defined.

5. Possible Solutions

What are some possible solutions? Some have already tried.

First idea that may come to mind is renaming it. I think there is nothing really wrong with Bitcoin in itself or with the implementation of Proof of Work. The problem is the naming is misleading. We could rename Proof of Work to Proof of Computation. After all, what miners actually do is compute hashes repeatedly until a valid one is found. This is a clearly defined process, and the proof is that the computation has been carried out.

However, even this framing has complications. Computation is not absolute; it depends for instance on computer architecture (at the outset, we cannot ignore this understanding of computing). The same puzzle is approached differently by CPUs, GPUs, and ASICs, each with different levels of efficiency. Basically, variations of the architecture of a computer and further specifications we can consider. A task that is computationally costly for one architecture may be trivial for another. This means that even “Proof of Computation” is not a universally defined concept, in this interpretation (where we consider computer in its most general understanding). This also has to do with the issue of Proof of Electrical Energy, and I will come back again to Proof of Computation in a minute.

6. Solution

A more accurate naming is Proof of Hash, or more precisely, Proof of Signed Hash. This is not pretentious. What miners actually prove is that they have produced a hash that meets the difficulty target, and that it is cryptographically linked to a block and their signature. Both hashing and signing are mathematically well-defined functions. There is no ambiguity here.

Beyond this, further labels such as Proof of Work, Proof of Computation, or Proof of Energy are a stretch if we want to be rigorous. Work and Energy are specifically problematic because of the meanings and connotations.

7. Coda: Proof of Stake

It is useful to contrast this with Proof of Stake. The terminology is more consistent. Proof of Stake literally demonstrates that validators have locked or “staked” tokens. In this case, the name matches the mechanism, and the description is more scientifically accurate. Definitely less inaccurate.

In particular, we can look at Ethereum’s Proof of Stake, the most popular blockchain implementing Proof of Stake. From a computational standpoint, Ethereum is more complex and allows more complex verification processes (which can be specified computationally). Bitcoin’s mechanism, by contrast, is relatively simple: it amounts to proving that a hash with certain properties exists. In fact, the Ethereum Virtual Machine is a Turing complete machine (one of the main proposals in the origin of Ethereum) and this potentially has to do with the verification processes and the definition of Proof of Stake. So Bitcoin’s while consensus is more simple (the machine that processes Bitcoin transactions inside the Bitcoin blockchain is not Turing complete) and narrowly focused on hash computation, Ethereum’s protocol allows a broader set of computational proofs as part of its consensus and execution layers.

8. Conclusion

So in conclusion, how is Proof of Work wrong? It is a terminological problem, that can lead to misleading interpretations in the context of Bitcoin and the more general context of the global economy and commentary by all sorts of organizations (scientific, technological, companies in general, governments, financial, etc.). Bitcoin’s consensus mechanism is technically effective, but the way it is described is misleading. Proof of Work is neither work in the physical sense nor in the economic sense. For scientific integrity, it would be more precise to change the name, maybe call this Proof of Signed Hash.

Finally, this critique is not a claim that Bitcoin cannot serve as a digital coin or payment system. It is only an attempt to clarify misconceptions about its foundation, so that discussions of its role can be based on a more accurate understanding.