Human Joy and Happiness

Motivation, scope, and a restrained technical framing

Dot Theory is presented on this site as a research programme in structured representation and model revision. That can read as dry. In practice, the reason anyone cares about better models is not abstract elegance. It is that better models change outcomes, and outcomes change lives and bring joy.

This page explains the human motivation behind the programme without turning it into a manifesto. The link is simple:

• We live inside partially observable systems.

• We make decisions under uncertainty.

• Our wellbeing depends on which decisions we make, and on the environments those decisions co-create.

• Not including this page would technically make the proposal incomplete.

In that sense, questions of joy and happiness are not outside natural philosophy. They are one of its end uses.

This page does not attempt to “prove” a theory of consciousness, or sentiment nor to settle disputes in philosophy of mind. It offers a narrower claim:

If you model humans and institutions as agents acting under partial observability, then improved representation of context and feedback can improve decision quality. Improved decision quality can improve experienced wellbeing.

That is the most that can be responsibly said at this level.

A short explanation of Dot Theory in this context

Dot Theory treats “dots” as real units of meaningful information: data points recorded from observation, measurement, or self-report that construct decisions.

By itself, a dot is not meaning. Meaning arises when dots are interpreted within a context and is demonstrative.

This context can be formal and measurable, for example time, location, sensor calibration, selection effects, incentives, and constraints. It can also be personal and experiential, for example memory, attention, affect, and values. The former category is straightforward to formalise. The latter category is harder, and should be treated with humility.

The operational thesis is:

When a model uses relevant contextual information that it previously omitted, its predictions and decisions can improve in regimes where that context matters and can be accessed.

The wellbeing thesis is:

When decisions improve under a stable value specification, expected measurement outcomes and wellbeing can improve.

This is a claim about modelling and decision-making under uncertainty. It does not require the assumption that “reality is only data”, nor that consciousness is fundamental in the physical sense. Both these claims are superfluous to the inquiry, but worth mentioning as objects, as they may be positioned as ways to stimulate popular understanding of the concept of relative realism.

Glossary of key terms used on this page

Data (D): recorded informational units such as measurements, events, and reports.
Metadata (M): contextual qualifiers of data, such as time, location, measurement conditions, and interpretation conditions.
Observer (O): an agent, human or artificial, that selects data and acts on it.
Utility (U): a value function over outcomes. On this page, “happiness” is treated as a family of experienced utilities, not a single scalar.
Creativity: the generation of novel candidate interpretations, options, or actions through recombination and exploration.
Partially observable system: a system where the underlying state cannot be fully measured, so agents infer it from incomplete signals.
Recursive interpretation: repeated updating of beliefs and plans as new information becomes available.

Six restrained premises

These premises are intentionally modest. They aim to be contestable rather than proclamatory.

Premise 1: Partial observability is the default

Human agents act with incomplete information about the world and about themselves. Most meaningful decisions are therefore made under uncertainty.

Premise 2: Context changes meaning and action

The same recorded datum can imply different actions depending on context. If context is omitted, error rates increase in the regimes where that context is relevant.

Premise 3: Human wellbeing depends on decision quality under a value specification

Given a value function, better decisions tend to yield better expected outcomes. This does not imply perfect outcomes or moral certainty. It is a structural claim about choice under uncertainty.

Premise 4: Humans generate context, not just consume it

Humans do not only ingest signals. They generate salience, goals, and narratives. They decide what is relevant. This is not a mystical claim. It is an empirical fact about attention, memory, and action.

Premise 5: Creativity is an exploration operator

Creativity can be treated as the production of candidate interpretations or actions that are not obtained by direct extrapolation from the current frame. In practical terms, it expands the search space of viable options.

Premise 6: Institutions and tools can support or suppress creativity and agency

Systems of education, work, governance, and technology shape the space of options people perceive and the feedback they receive. The design of those systems therefore affects experienced wellbeing.

How this links to happiness without metaphysical inflation

“Happiness” is not treated here as a metaphysical essence. It is treated as an experienced signal that tracks, imperfectly, whether an agent’s interaction with its environment is going well according to its values and constraints.

In that framing:

• better context representation can reduce avoidable error

• better feedback can improve correction

• better option generation can improve agency

• better agency can improve experienced wellbeing for many people in many environments

This is compatible with pluralism. Different individuals and cultures have different value functions. A modelling programme does not replace those values. It can only improve decision making relative to them.

What this page does not claim

To keep these statements accountable, it is important to state boundaries.

This page does not claim:

• that consciousness is fundamental in the physical sense

• that quantum mechanics “explains” subjective experience

• that AI systems are or are not conscious

• that happiness is a single measurable scalar

• that improved models guarantee good outcomes

It claims something narrower:

Better representation of relevant context and better feedback can, in many settings, improve decision quality and therefore improve expected wellbeing under a given value specification.

A minimal technical appendix

For one formal lens for the above, a possible expression is:

Let hidden state be sₜ, observation be oₜ, and action be aₜ. Let the agent maintain a belief bₜ(s) over hidden states. Let utility be U over trajectories.

A standard POMDP framing gives:

• belief update: bₜ₊₁ = Update(bₜ, oₜ₊₁, aₜ)

• policy: aₜ = π(bₜ)

• objective: maximise expected utility 𝔼[U | π]

Dot Theory’s representational claim, translated into this language, is:

There exist contexts where the observation model and belief update are systematically improved by including additional formally specified metadata μ, so that:

Update(bₜ, oₜ₊₁, aₜ) → Update(bₜ, oₜ₊₁, aₜ, μ)

and this yields measurable improvement in predictive and decision performance in those regimes.

That is a computational claim. It is testable.

Closing

I care about this programme because representation determines what becomes thinkable and therefore doable.

In human terms: the way we structure information shapes the decisions we make, and the decisions we make shape our lives.

If Dot Theory ends up contributing anything valuable, it will hopefully be because it helped build tools and institutions that reduce avoidable error, improve feedback, and expand agency under real constraints.

That is where joy and happiness become relevant to a technical programme without turning it into ideology.

Stefaan