The Categorical Imperative: Why You’re Thinking About Existence Now (And What to Do Before a Boltzmann Brain Steals Your Thesis) — Dr. Katya Steiner

# A quick cosmic shrug

You graduated into a world where phones are simultaneously tiny computers and passive-aggressive notification machines, and yet here you are, conscious in a 13.8-billion-year-old universe. That feels like luck, insult, and a metaphysical puzzle all at once. Welcome to the club: ‘why now?’, what counts as science, whether minds are bubbles, and the mildly terrifying idea that the far future might be full of spontaneously forming brain-sized hallucinations that outnumber us. Pour coffee. Let’s unpack this without getting lost in technobabble.

## Why do we exist so early? Short answers and a mathy middle

There are three ways to start untangling the timing puzzle, arranged from commonsense to the annoyingly technical.

1) Selection bias. You can only notice times compatible with observers. This is indexical information — you are necessarily observing from an epoch that supports observers. In logic-speak, you have self-locating evidence.

2) Observer distribution. Maybe the distribution of observers in time is peaked around epochs with plentiful heavy elements and stable planets. Or maybe civilizations are fragile and burn out fast.

3) Anthropic measures and infinities. Here the real math comes in. If the universe can host observers for vastly more spacetime volume later, naive frequency-based probabilities suggest you should expect to be late — unless the measure (how we count observers) or the model of cosmology prevents late-time dominance.

That last point is the place where probability theory, measure theory, and philosophical logic collide. If you treat observers as points in spacetime, you must choose a measure: do you weigh by comoving volume, by entropy production, by number of intelligent labs? Different choices give different predictions. This is the measure problem in cosmology, and it’s not solved by intuition alone.

## When math disciplines start holding hands

– Probability & Bayesianism: Self-locating beliefs (I am an observer here, now) require you to update with indexical evidence. Think Sleeping Beauty and Doomsday-style paradoxes. Whether you use the Self-Sampling Assumption (SSA) or the Self-Indication Assumption (SIA) matters.

– Measure theory & infinities: If spacetime volume or observer-counts diverge, frequency interpretations collapse. You need regularization rules, cutoffs, or principled measures to compare infinities. That’s where cosmologists get slightly desperate and very careful.

– Algorithmic information theory: Solomonoff induction and algorithmic probability give priors over observations based on complexity. One way to favor ordinary observers over Boltzmann brains is to argue that simulated or fluctuated observers have higher Kolmogorov complexity or lower algorithmic probability — though this is contentious.

– Decision theory & expected utility: If late-time observers are numerous but irrational or unstable, should you weight them? Pascalian-style considerations and Pascal’s mugging become relevant when tiny probabilities combine with enormous payoffs. Philosophers and economists disagree about which weighting principles are reasonable.

– Modal and temporal logic: Considering possible-worlds with eternally inflating futures brings modal logic into play. Which worlds count? Which indexical propositions survive in asymmetric time directions?

All this is to say: the philosophically interesting question is not just ‘why now?’ but ‘how do we count observers?’ and ‘which logic do we use to update our beliefs when our evidence is indexical and weird?’

## Boltzmann brains: the cosmic flea that makes philosophers twitch

Imagine a de Sitter future so long that thermal fluctuations occasionally conjure a self-aware configuration with false memories. If those random brains massively outnumber ordinary observers, statistical reasoning suggests you are probably one. Yikes.

Why this is bad math and worse philosophy:

– Predictive failure: A theory that makes you more likely to be a Boltzmann brain undermines its own evidence base. We rely on consistent histories to trust observations; Boltzmann-dominated theories undercut that.

– Cognitive instability: If your best cosmology implies your cognition is untrustworthy, you have reason to reject the cosmology — but that’s an unstable move unless you can rerun the inference with better priors or a different measure.

Fixes? Tweak the measure, prefer cosmologies with finite observer counts, or adopt principles that favor ordinary observers. None are trivial, and each has trade-offs.

## Bubble model of consciousness — cute, vulnerable, and useful

The teenager’s bubble model is irresistible: neurons inside a physical bubble; a will bubble; emotion bubbles bumping into each other. It’s a useful metaphor for hierarchical emergence — lower-level dynamics producing higher-level phenomena.

Where math helps: dynamical systems, network theory, and information-theoretic measures can turn metaphors into testable claims. Is the will bubble a high-level attractor in phase space? Do emotional sub-bubbles correspond to metastable network states? These are precise hypotheses you can study.

Where to be cautious: metaphors flatten into metaphysics if you’re not careful. Saying qualia are ‘just’ synchronization patterns is a scientific hypothesis, not a dismissive slogan. If you want to defend it, couple the metaphor to measurable constraints and predictions.

## A quick guide for the caffeinated grad applicant

If you’re drafting a PhD proposal on any of this, keep it irritatingly practical:

– One-paragraph opening problem: state the puzzle and why it’s unsettled.
– State the literature gap: who says what, and which assumptions you’ll test.
– Methodology: conceptual analysis, formal modeling, case studies of cosmology papers, or simulations of observer-counting.
– Aims: a clarified measure, a toy model that avoids Boltzmann dominance, or an account of emergence with testable predictions.
– Feasibility: timeline, chapters, supervisor match. Show you have read the right people.

## Takeaway — existential comfort with a to-do list

You’re allowed to be fascinated and unsettled by why you exist when you do. The trick isn’t panicking about Boltzmann brains; it’s sharpening your tools. Learn a bit of measure theory, the basics of Bayesian updating with indexical info, and some dynamical-systems ideas about emergence. If you’re writing a thesis, make claims that could be wrong in principled ways — that is the entire point of scholarship.

And on a slightly cheeky note: if a Boltzmann brain materializes in the reading room and claims to have written your literature review, cite it politely and ask it to fetch coffee.

So here’s the question to leave you with, coffee in hand: if you had to choose one mathematical or logical principle to bet your whole research program on — Bayesian updating, a principled measure on observers, algorithmic simplicity, or dynamical emergence — which would you pick, and why?