This piece is largely based off an interview I had with our CTO, Wyatt Meldman-Floch, where we took a deep dive into what inspired him to create Constellation, and more specifically, what fuels his underlying engineering ethos based around integrating natural law. Enjoy!

In our interview, we covered everything from thermoeconomics and dendrograms, the origins of Constellation and the universe, to DaVinci and Dutch string theorists and everything in between, so please forgive me while I attempt to properly formulate a jump off point for this quantum roller-coaster ride inside the mind of Wyatt.

The bulk of this piece is meant to illustrate how Constellation’s network can be compared to existing complex structures in nature. Wyatt breaks down how he approaches Constellation’s architecture by taking inspiration from and in some cases, mimicking, natural orders. Read on to find out how this approach is revolutionary and unlike any other approach in the blockchain space.

Constellation’s Inspiration


Constellation was spawned in part from many deep talks that Wyatt and Ryle would have after a days work during their time together at Radius, hanging out around the E-Trade building in downtown San Francisco. They were working in the midst of the big data boom, and they’d spend a lot of time discussing how they’d ideally envision things being done differently in the big data space.

Wyatt had also created a program called Big Fish, Little Fish, where nodes would host data and merge data like a little fish getting eaten by a big fish, whereby the big fish would represent increasingly trusted data points which had a greater idea of the network’s state. The creation of this program served as a sign of things to come:

“We spent a lot of time looking at patterns within primitives. If you’re looking at planet earth, it would be chemicals. If you’re looking on the internet, it would be points of data (the chemicals and data being the primitives) and looking at patterns in between them. That’s sort of how we started forming the ideas that would later become Constellation.”

Wyatt cited many mathematical geniuses as sources of inspiration for Constellation. From Fields Medal winners (which is essentially the Nobel Prize equivalent for Mathematicians) Stephan Smale, who did pioneering work in the 60’s tying Partial Differential Equations (PDEs) to economic markets and Alexander Grothendieck, to Dutch String Theorist and Nobel Prize winner Gerard’t Hooft. Maurice Herily, a Professor Emeritus at Brown, was also cited as a significant influence for Wyatt. “Herily analytically modeled distributed systems and proposed that we treat distributed systems in the same way we treat particle systems,” said Wyatt. All of these theorists, thinkers, and academics influenced Wyatt and the engineering team’s approach regarding the structure of Constellation.

If Herily could be seen as putting the “math on paper,” the creators of Twitter’s Algebird library for creating data models based on abstract algebra could be seen as turning that “math on paper” into “math on code.” Wyatt cited them as his modern-day engineering inspirations, saying “I’m really inspired by algebraic data types and projects like Cats and Algebird which extend the use of abstract algebra to create data structures that allow data processing to be reasoned about like a math problem.”

Taking all the influences into account, with Constellation, Wyatt and Ryle wanted to do things a bit differently than what was happening in the big data space.

Taking Cues From Nature – Entropy, Biomimicry & Thermoeconomics


Before we dive into the technicalities of how natural law is built intoConstellations tech, I asked Wyatt what inspired him to turn to nature for inspiration in the first place. “Often, it’s easier to try and reverse engineer a solution to a problemthan to start from scratch. We do this in programming every day. Fortunately, in many situations, we can see anomalies in nature, be it biochemistry or physics, which give us clues to the underlying structures that make solutions work.  The inspiration for most people who come up with new paradigms of understanding, like pure mathematicians, often comes from reasoning about natural phenomenon independent of existing frameworks.”

One aspect of natural law that is heavily featured in Constellation’s approach is thermoeconomics. For those who aren’t familiar (don’t worry, I wasn’t either) with this field, it’s a school of thought that maintains that human economic systems can be modeled as thermodynamic systems. Wyatt primarily integrated this in regards to thinking about the underlying utility of our utility token, $DAG. “I spent a lot of time thinking about what the ‘utility’ of a ‘utility token’ was, and it seemed to me that the utility of a token was actually the quantization of a good or service that the token represented. If that were the case then what is the underlying value of any cryptocurrency? It certainly has something to do with ensuring the validity of data.”

How can you communicate an exchange of value without an intermediary? In our society, that intermediary typically comes in the form of money, but in nature and physics, that unit is typically entropy. Wyatt thinks entropy (as it is a topologically invariant measure) is indicative of an underlying homotopy type hierarchy that the universe follows with regards to scaling. This notion is codified into Constellation as the underlying connective thread that ultimately normalizes the network’s data, allowing all processes to speak the same language. Speaking to this point, he described how this plays out in the network itself:

“Let’s model the actual dynamics of our distributed system based upon these physical and economic laws. It allows us to directly connect every operation that a network does directly to a unit of value. Every operation of a network, not just every unit of data, can be stored and maintained. People who care about the validity of their data, who care about encryption and privacy and using a blockchain, who want that provenance over their data… they care about whether it is true or not, and how we achieved that conclusion. If every bit of data that leads up to that conclusion is based on entropy, then you can build a model on that. You can normalize the data, and then you have all different processes speaking the same language. The same way an ecosystem or an economy communicates.”

In Wyatt’s presentation at De:Centralize 2018 (linked below), he said: “If you want to solve a problem, the smartest thing you can do is just copy what nature did.” The popularized term for this notion is biomimicry,  which is defined as the “imitation of the models, systems, and elements of nature for the purpose of solving complex human problems.” A famous example of biomimicry can be seen in Davinci’s early sketches of a flying machine (as seen above), based upon his observations of birds. I asked Wyatt to follow up with some other examples of how this is evident in the architecture of Constellation, and he shared several other fascinating parallels.

“Constellation’s network topology resembles a scale-free network. This structure pops up in the natural ecosystem or the cosmos (correlation does not imply causation, but the universe looks like a dendrogram, so does a neuron and a tree, etc.) because they are systems that are continuously applying some sort of optimization rule, typically based in the reduction of entropy. It’s an optimization problem that acts effectively like a balancing act. In the same way that plants and animals die, are reborn and transmuted into different forms of energy, our network needs to autonomously reorganize itself, like an ecosystem adapting to climate change.”

We see this same resilient autonomous pattern play out in modern day economics, as major corporations die out, get bought up or get swallowed whole by a different entity. “That’s why people model economics after thermodynamics, because it’s just like nature,” said Wyatt. In regards to Constellation’s Proof of Reputable Observation protocol, which assigns a reputation-based score to nodes based on prior accuracy, validity, and trust, Wyatt once again looked at nature for inspiration. “If you’re creating disorder within the greater system, the system will self- correct you and put you in your place to find its equilibrium.”

In Wyatt’s eyes, each blockchain protocol can be viewed as a different species, and with Constellation, he envisions the network serving as a baseline “pollinator” for any potential use case. “Bitcoin is like the crocodile, that has existed since the time of dinosaurs. We want to be like a pigeon—the most common, adaptable species that flourishes and finds utility in any ecosystem. Alternatively, we want to be the bees, something that helps to pollinate all potential use cases and be that underlying infrastructure.”

Fractal-ly Speaking…


In Wyatt’s presentation mentioned above, two notions popped up a fair amount — Homotopy Types and the dendrogram structure of a hierarchical DAG, which minimizes entropy in a system. I asked Wyatt for a layman’s breakdown of how these concepts appear in nature, and how they each factor into the design of Constellation.

He started by sharing the work of the foremost mathematician of the early 20th century, von Neumann, who essentially came up with the basis for Constellation’s dynamic partitioning scheme a long time ago. Wyatt, take it away:

“What that question asks is ‘how can we model higher dimensional spaces that generates something abstract like a fractal?’ Without getting too into it, the methods used in fractal geometry (what pure mathematicians use to understand recursive, self-similar structures [like a hylo/metamorphism]) are primarily focused on something called measure theory, which is important, but they don’t translate explicitly to non-linear structures in programming like the f-algebra of a recursion scheme.”

What Wyatt and the engineering team did was extend that idea to something called Homotopy Type theory, which is how they think about the structure of data and processes in functional programming. In Wyatt’s words, “there’s this thing called the Curry-Howard Correspondence which ensures that a ‘Typed’ program is actually a proof. In this regard, by implementing these ideas with functional methods, we can assume that our code implements a model like something from continuous geometry.”

Why on earth would we want to have a fractal structure like this? According to Wyatt, “it’s because fractals have special space preserving properties which help us embed (store) information efficiently, and storing information efficiently is at the root of how we approach scalability.” A 2D fractal is a way to infinitely put information in a space and pack it in there, and this fractal structure is what allows Constellation to create an infinitely unbounded space to pack data into.

Constellation’s logo itself echoes this notion as the solution of closest packing of spheres (a hexagon) represents what we’re trying to do, which is “expand the closest packing of all information in all dimensions.”

I followed this up by asking Wyatt about the fractal nature of Constellation’s DAG, and if any other existing DAG based protocols, such as IOTA, are designed with a similarly fractal nature in mind. “Holochain is close, but Holochain is built for API level servers, nothing more, nothing less. Constellation is built for everything. We want to be the glue to connect the bottom line with something as slow and store-valuey as Bitcoin.” For a more thorough breakdown of how Constellation’s hylochain differs from the likes of a typical blockchain or other DAG protocols, check out the infographic below.



Hopefully, you’ve just gleaned a telescopic look inside the mind of Wyatt and the origins of Constellation, and how much of its underlying architecture is inspired by biomimicry. I hope this piece fully conveys how unique Constellation’s architecture truly is while giving you some further insight into Wyatt’s inspiration for creating Constellation.

Until next time,