“Our awareness seems to shrink in direct ratio as communications expand; the world is open to us as never before, and we walk about as prisoners, each in his private portable cage.” — Arthur Koestler, 1944

Day by day, humans are becoming a more entangled global organism. We’re deeply dependent on external systems (supply chain crisis), yet we’re increasingly divided, clouded by mistrust in all directions, and driven into tribal camps. We find consensus on goods and services through markets, for better or worse. But the concept of “consensus” (or synonymous; agreement, accord, congruence, etc) can be expanded.

A healthy person’s organ systems have quicker responses and more capacity. Societies could optimize in similar ways. This is for consensus amongst sovereign free-thinking individuals with no coercion. Obviously, you could “find” consensus with a dictator. But I’m more interested in mechanisms (technical, cultural, ritual) that facilitate genuine consensus-making.

Thinking up the human “stack,” there’s the consensus of perception, information, interpretation, value, objective, and action.

Imagine you’re with a fellow hunter and hear a snap in the jungle. Between each other, you move through each layer of consensus. “You heard that too?” (perception) “It came from the left.” (information) “It might be a tiger.” (interpretation) “We should pay attention.” (value) “We should get out.” (objective) “Let’s start running.” (action).

Each stage is a process of agreeing upon each other’s responses. Disagreement at any one stage halts moving up the latter towards action.

Consensus of Perception

The consensus of perception is basically: what raw data do we perceive in the world together? While disagreements on correct information are increasingly plentiful (due to the amount to sift through), they’re often not beyond simple searching/sourcing/verifying. Many disputes stem from a lack of tools to develop other types of consensus alongside information.

This is arguably a crumbling enterprise that’s never really been dependable. The majority of perceptions we gather about the world are filtered through countless brains and media organizations. Human communication is like a big game of telephone. We share what we heard on a podcast from someone who read an article where the reporter talked with an eyewitness that read something on Twitter (ouch).

How do we capture raw data of events happening in the world? How do we avoid selection bias (eg sourcing) and capture a uniform distribution? How do we verify data/metadata authenticity? How do we do this without creating a surveillance state? These are all extremely difficult problems, especially in the wake of deep fakes (not just faces but everything).

I believe to tackle the problem of misinformation, you must tackle the problem of misperception. If you can barely see, your information will be noisy at best, muddy on average, and harmfully false at worst.

Consensus of Information

The consensus of information is basically: what does that raw data represent? This is an extremely dense topic (multiple textbooks) that I’ll leave for others to expand on. (Maybe I’ll come back to thinking about this later).

Some honorable mentions for consensus of information are the amount of information increase, compression, overfitting, biases in data, propaganda model, replicability crisis, and kinda most of this list.

Consensus of Interpretation

This is an intermediary category that’s not completely necessary, but I believe it’s relevant when talking about consensus between humans, creatures that interpret facts differently.

Interpretation is essentially the question: why is that information true? It reflects a deeper understanding of a piece of information (the general pattern it exists within) and the associations it has with others. Interpretations are very much dependent on one’s base assumptions about the world.

The consensus among interpretations comes from clean information, shared assumptions, open discourse, and the patience/ability to explain/explore associations.

Consensus of Value

A consensus of value is essentially: What information/interpretations can we all agree should be valued the most? Value consensus is arguably the role of religious behavior. We connect on shared values in conversation, yet it’s fragile. Often, we hide values or even lie about them. Alongside bandwidth limitations/data loss of language, media filters, ideologies, cognitive biases, etc, highly held values aren’t proportionally represented in human output.

We live in an endless sea of data, information, and voices. Values act as filters on this mass of data. Malcolm Gladwell’s distinction of puzzles vs. mysteries makes this clear. We don’t have a lack of information, we have too much information. Values allow us to direct attention towards only the information that matters. They also do this.

Consensus on mutually constructive values is important, because apart from biological drives (food, reproduction, etc) what we value is often an imitation of what other people value. We outsource value hierarchies to other people or institutions that represent authority. We want a “cool” car because other people do too. This is called the mimetic theory of desire.

Finding consensus on values is the arena of argumentation, rhetoric, activists, and mission statements. But we have few hard tools for clarifying personal values and quantifying collective values.

How do we find population-wide value alignment? How do we agree upon where to allocate resources? What do people see as the highest good, for themselves, their family, their community, humanity, and the earth? How do we filter for information that serves that highest good without just confirming our biases?

There’s a lot to write about here (eg. the meaning crisis, Montessori schools, new voting systems, psychedelic experience, and more) but I’ll leave that to another post.

Consensus of Objective

The consensus of objective is: with these shared values and this shared understanding, what is the best goal to pursue? This is the role of leadership. Different leaders market different objectives, then people vote with their feet, time, and resources on select aims. We demand an objective, more than there’s supply. Often, objectives are communicated with a worldview and line of reasoning. This is to find consensus among the lower levels before agreeing on where to go. Like most things, finding consensus on objectives is bottlenecked by natural language.

Often many different objectives have to be amalgamated together. Serving a person’s self-interest, community interest, corporate interest, social interest, national interest, and global interest.

How can we anonymously (or pseudonymously) uncover people’s desired objectives for themselves, their community, their higher values, the larger population, and all of humanity? Finding consensus among objectives is a real-world variation of the value alignment problem previously mentioned.

Consensus of Action

Then finally consensus of action, which is basically: how do we implement that objective? Often I’ve found people completely agree on a goal but everything falls apart when they act. They disagree on the plan or they get caught in integration hell. This seems the most difficult. It’s the arena of clear plans, skilled logistics, and effective management.

Lots of people are playing with new social organization structures. DAOs are an exciting way to organize complex projects across a global community where everyone has skin in the game. But often, an effective plan is best formed by a single person or small group, as this brings focus and clarity.

Action is the end goal of forming consensus. There’s no point in consensus unless it helps formulate how to act together.

Conclusion

These are some categories from unpacking the concept of consensus. The process of agreeing with each other on how to act. Starting with correctly perceiving the world. Once we agree that our raw data roughly matches, we can derive information, come up with explanations, and agree upon what the data represents. Then, with a shared set of values, we can dig deeper into relevant areas. Eventually, we’ll become more informed on how to respond, and we can hash out what should be done. Finally, we’ll come upon something to do, and formulate a plan to do so. This will complete our consensus-making.

I’ve found these ideas interesting to ponder what implementation could bring better consensus for each level. Of course, these ideas are mostly semantic, but they are interesting frames of reference to think from.

American institutions we’re built to impose consensus amongst hundreds of millions of people from birth to death. But our institutions are crumbling, so we need better mechanisms to find alignment across groups of people. Not for the sake of kumbaya (although that’s sweet), but to be resilient, at peace with each other, and lead meaningful lives.

Hope you found something interesting there!

Aaren