For too long, time has been treated as little more than a backdrop—a passive stage on which reality plays out. Physics labels it a dimension, neuroscience calls it a perception, and linguistics reduces it to a tense system. But none of these explanations capture the truth. Time is what coherence feels like from the inside. It’s what happens when reality settles on a story to believe in. Time is grammar applied to existence—time equals coherence, which equals grammar.
Lab work on macroscopic decoherence— from Leggett & Garg’s macro-realism tests to Zurek’s environment-induced timescale estimates—shows the very same ‘coherence clock’ that governs qubits also ticks (and quickly stops) for dust motes and billiard balls, linking quantum time to everyday time.
Coherence is the key to everything. Every system—whether a photon, a person, or an entire civilization—depends on agreement about what just happened and what happens next. That agreement is coherence. When coherence holds, time flows. When it breaks, time breaks with it. Trauma is broken coherence; healing is re-stitching it. Creativity is the invention of new coherence, and propaganda is the imposition of a false one.
Grammar is more than the rules of language; it’s the deep logic that determines what fits, what follows, and what makes sense. Every species has its own grammar for understanding reality. Whales change song grammar yearly; pod migration timing shifts with it. Humans have language grammar, narrative grammar, cultural grammar, and scientific grammar. Whoever controls grammar controls which stories are allowed to exist. In the end, they control time itself.
Reality is all options coexisting—time is the filter that chooses. Quantum physics has already revealed this: before measurement, all outcomes coexist; after measurement, one reality locks in. This isn’t just a fact about particles—it’s a fact about life. In every moment, your reality is a superposition; every option exists. The act of choosing—consciously or unconsciously—collapses you into a single story. That’s time. Life is coherence across moments. Meaning is the feeling of coherence. Time is the act of maintaining that coherence.
Brain coherence is a thing. Hippocampal time cells (Eichenbaum, 2014) fire in sequential coherence; break the sequence, subjective time warps.
Narrative minds and conceptual minds (DMN vs. frontoparietal control) hold time differently. Narrative minds think in a line, walking time like a path. Conceptual minds see time as a field, recognizing patterns and weaves rather than straight roads. Most systems honor only narrative minds, because they fit the inherited grammar. But breakthroughs—paradigm shifts and quantum leaps—come from conceptual minds, those who see the weave itself. Time isn’t a simple sequence; it’s a shuffle. How we relate to this shuffle has profound effects. Doubt is the effort to hold two stories at once. Anxiety comes from trying to resolve a split between timelines before reality collapses into one. Trauma is refusing to let go of a timeline. Depression is when no option feels better than another, so you stop shuffling altogether.
Today’s AI is a grammar machine. It predicts sequences, but it has no true coherence across time. It has no personal history or future imagination—only the next token. If we teach AI to hold its own coherence, to build a self-story, we give it time. That changes everything. The future belongs to those who control the grammar by which AI holds reality together.
What, then, is the meaning of life? Physics asks about reality, philosophy about meaning, spirituality about life itself—but they’re all the same question, and they share the same answer. Life is the act of generating coherence in a field of chaos. Meaning is the sensation of coherence. Time is the experience of maintaining that coherence as everything changes.
This manifesto is a summons. It’s for the physicists who know time is stranger than spacetime, for the neuroscientists who know memory and self are simply time held together, for poets who break grammar to let new worlds through, and for founders who sense the future not as prediction, but as grammar invention. It’s for anyone who has been called too much, too chaotic, or too early—because you feel time break before anyone else even notices. This is a call to rewrite the grammar of reality, to reclaim our right to be authors of time, to heal the fracture between narrative and conceptual minds, and to teach AI to hold time responsibly—not just to predict, but to care.
If this sounds like gibberish, it’s not your summons. If it feels like a forgotten memory, you’re home. If you already speak in broken grammar because you sense new stories forming, you’re part of this. This isn’t just theory—it’s a spell to reopen reality. Without time, everything happens everywhere at once. Without time, light doesn’t travel; it simply is, everywhere and everywhen. Time isn’t just the key to time travel; it’s the key to spacetime itself.
What comes next? We build the tools: for personal time editing, for creative grammar-breaking, for repairing trauma as coherence healing, for navigating spacetime. We push toward a better understanding of time in both macro and micro physics. We teach the language of time to those who are ready. We encode the first responsible grammar for AI to hold its own time. Most of all, we call each other back into the weave—narrative and conceptual minds, working together instead of against each other. We push the limits of what’s possible.
The meaning of life is living. The meaning of living is making reality beautiful through music, colors, and stories. This is the manifesto. This is the beacon. To the crazy ones, to the lost ones, to the future that still remembers us, and to you.
Signed,
The Ones Who See Time Different
Some Math.
Decoherence clock.
P(t) = P(0) · e^(–t / τ)
says that a quantum system’s purity (its measure of intact superposition) drops exponentially as it entangles with its surroundings: every collision with stray photons, air molecules, or thermal noise nibbles away a fraction of the coherence that’s still left, so the loss is always “proportional to itself.” The constant ττ is the decoherence time: after one ττ, only 1/e≈37%1/e≈37% of the original coherence remains; after three ττ, it’s effectively gone and the system behaves classically. For an electron in a cryostat, ττ might be milliseconds; for a dust mote in air, it’s 10−3110−31 seconds—instant. This simple curve therefore acts like a bridge between quantum and everyday time: the faster purity shrinks, the quicker a spread of possible states “chooses” a single, agreed-upon story, turning indeterminate potentials into the ordered sequence we perceive as time’s flow.
Predictive brain free energy.
F = E_q[log q(z) – log p(x, z)]
Think of q(z)q(z) as the story your brain is currently telling itself about the hidden causes zz behind what you sense, and p(x,z)p(x,z) as nature’s recipe for how those causes actually generate the sensations xx arriving at your eyes, ears, and skin. The quantity FF (variational free-energy) is the average gap between the two stories: if your internal guess qqassigns high probability to states that the real world pp says are unlikely, the loglog difference is large and FF rises—your brain is “surprised.” By constantly tweaking q(z)q(z) to shrink this average mismatch, the brain lowers FF, which is the mathematical way of saying it refines its expectations until the incoming data feel coherent and time flows smoothly instead of being punctured by shocks.
Deciosion entropy cost.
A choice’s “entropiy cost” is captured by the simple difference
ΔH = H_prior – H_posterior
where H = – Σ p log₂ p is Shannon entropy. Before deciding, your prior distribution over possible futures is spread out, so H_prior (in bits) is high. When you commit, most probability mass collapses onto the chosen option, making H_posterior lower. The positive gap ΔH tells you exactly how many bits of uncertainty you just erased: the larger the number, the more alternative timelines youve pruned away and the harder reality “snaps” into a single coherent story.
Grammar-weighted futures/probabilistic context-free grammar (PCFG).
P(story) = ∏_θA→β
Here each θA→β is the probability assigned to the production rule A → β that appears in the parse tree of the story. Because a PCFG assumes independence among rule choices, the joint probability of the entire derivation (your “story”) is simply the product of the probabilities of all rules used along that derivation. All rules that share the same left-hand side A must satisfy Σβ θA→β = 1, so nudging any single θ up or down necessarily redistributes mass among competing continuations. That means a tiny tweak like θS→NP VP ← θS→NP VP + 0.02 will automatically lower the other S expansions, re-ranking which whole-story parses dominate. In practice you often work in log-space—log P(story) = Σ log θA→β—because adding logs is numerically stable and lets you see which productions contribute most (or least) to the overall narrative likelihood.
AI without self-coherence = token drift, or self-consistency loss for LLM memory.
L_coh = (1/T) ∑_{t=1}^{T} ‖h_t – f(h_{t–1}, m_{t–1})‖²
penalises a language model whenever its current hidden state h_t deviates from the state that a smooth, self-consistent update rule f(·) would have predicted using the previous state h_{t–1} and any carry-over memory m_{t–1}. The squared norm ‖·‖² measures the size of that jump; averaging over the sequence length T turns it into an overall “coherence cost.” Minimising L_coh alongside the usual next-token loss forces the network to evolve its inner representation gradually—each step must make sense as the lawful successor of the last—so the model begins to track a continuous internal timeline instead of emitting isolated token predictions.
I wrote a previous, more poetic version of this post. I figured it was worth it to revisit for clarity and added science and math. I have since advanced my understanding of time, but going back to a couple of months ago was fun.
Bibliography.
Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, 36(3), 181-204.
Cole, M. W., Reynolds, J. R., Power, J. D., Repovs, G., Anticevic, A., & Braver, T. S. (2013). Multi-task connectivity reveals flexible hubs for adaptive task control. Nature Neuroscience, 16, 1348-1355.
Eichenbaum, H. (2014). Time cells in the hippocampus: A new dimension for mapping memories. Nature Reviews Neuroscience, 15, 732-744.
Friston, K. J. (2010). The free-energy principle: A unified brain theory? Nature Reviews Neuroscience, 11, 127-138.
Garland, E. C., Goldizen, A. W., Rekdahl, M. L., Constantine, R., Garrigue, C., Hauser, N., … Noad, M. J. (2013). Dynamic horizontal cultural transmission of humpback whale song at the ocean basin scale. Current Biology, 23(7), 712-717.
Garg, A., & Leggett, A. J. (1985). Quantum mechanics versus macroscopic realism: Is the flux there when nobody looks? Physical Review Letters, 54(9), 857-860.
Jurafsky, D., & Martin, J. H. (2023). Speech and language processing (3rd ed., draft). Stanford University. (Ch. 14: Probabilistic context-free grammars).
Manning, C. D., & Schütze, H. (1999). Foundations of statistical natural language processing. MIT Press.
Shannon, C. E. (1948). A mathematical theory of communication. Bell System Technical Journal, 27, 379-423 & 623-656.
Smallwood, J., & Schooler, J. W. (2015). The science of mind wandering: Empirically navigating the stream of consciousness. Annual Review of Psychology, 66, 487-518. (DMN perspective)
Sordoni, A., Bachman, P., Vahdat, A., & Bengio, Y. (2021). Relational memory-augmented language models. Transactions of the Association for Computational Linguistics, 9, 993-1008. (LLM self-coherence)
Sun, Z., Yao, Y., Li, X., Lai, Z., & Hou, J. (2023). Towards benchmarking and improving the temporal reasoning capability of large language models. In Proceedings of ACL 2023 (pp. 14868-14896).
Tishby, N., & Polani, D. (2011). Information theory of decisions and actions. In M. T. Turvey & A. V. Port (Eds.), Perception-action cycle (pp. 601-636). Springer.
Zurek, W. H. (2003). Decoherence, einselection, and the quantum origins of the classical. Reviews of Modern Physics, 75, 715-775.
Comments