A deep IIT 4.0 analysis of the Cosmic Microwave Background
We applied a mathematical framework called Integrated Information Theory (IIT) — originally designed to measure consciousness — to the oldest light in the universe: the Cosmic Microwave Background (CMB). We weren't looking for alien signals. We were asking: does the universe's first light have hidden structure that existing analysis methods miss?
We used real data from the Planck satellite (ESA, 2018 release): 2,507 measurements of how the CMB's temperature varies at different angular scales. Think of it like measuring the pitch of a cosmic bell — the CMB "rings" at specific frequencies, and each frequency tells us something about the physics of the early universe.
We sliced the CMB into progressively finer bins (4, 6, 8, 12, 16, 32, 64 divisions) and measured how many truly independent "directions" of variation exist using Singular Value Decomposition:
| Resolution | Effective Rank | What It Means |
|---|---|---|
| 4 bins | 4 | At coarse resolution, everything matters |
| 8 bins | 8 | Still fully independent |
| 12 bins | 10 | Saturation begins |
| 16 bins | 10 | Same 10 dimensions |
| 32 bins | 10 | Still 10 |
| 64 bins | 10 | Always 10 |
No matter how finely we measure, the CMB has exactly 10 independent degrees of freedom. The rest is redundancy. This matches the standard cosmological model (ΛCDM), which has 6 free parameters plus ~4 derived quantities.
In plain English: The entire blueprint of the observable universe — every galaxy, every star, every planet — was determined by just 10 numbers.
The "emergence index" measures how much structure appears when you zoom out. We found:
| Bins | Emergence Index |
|---|---|
| 4 | 0.27 (peak) |
| 8 | 0.19 |
| 16 | 0.04 |
| 64 | 0.002 |
The universe's structure is most emergent at its broadest features — the division between the "static" (Sachs-Wolfe plateau, gravity's imprint) and the "dynamic" (acoustic peaks, sound waves) is the single most informative way to understand the CMB. Zoom in further and you only get noise.
We divided the sky into 48 patches (HEALPix nside=2) and computed integrated information for each:
- Mean Phi: 0.075 across 48 patches
- Standard deviation: 0.130
- Anomalous patches: 3 (Phi > mean + 2σ)
- Cold Spot Phi: 0.000 (not anomalous by IIT measure)
The known CMB Cold Spot does NOT show up as an IIT anomaly. This constrains what the Cold Spot can be — it's a temperature anomaly but not a structural one.
We tested all 9 Planck frequency bands (30-857 GHz). Dust, synchrotron radiation, and free-free emission create strong signals at extreme frequencies, but they produce zero integrated information — the foreground contamination is a simple additive process, not a nonlinear coupling. This validates the standard foreground removal approach.
The CMB has:
- Zero integrated information (Φ = 0) — each angular scale is causally independent
- High determinism (1.47 bits) — physics, not randomness, governs the structure
- Near-zero degeneracy (0.01) — every state is unique
- 10 effective dimensions — the universe's blueprint is remarkably simple
Witness Hash (SHA-256): 91ce475e0b1d419956a5879d54df1ab9df9d3b4939eef81fb2770521c273355b
Witnessed Data:
CMB_Phi=0.000000 | bins=16 | null_samples=200 | z_score=0.00 | p_value=1.0000
EI_micro=1.4734 | determinism=1.4734 | degeneracy=0.0144
SVD: rank=10/16 | spectral_entropy=2.1996 | emergence_index=0.0447
Sky map: 48 patches, 3 anomalous, Cold Spot Phi=0.000
Emergence sweep: peak EI at 64 bins, peak emergence at 4 bins
Cross-frequency: all 9 bands Phi=0
Reproduction:
cargo run --release -p cmb-consciousness -- --bins 16 --null-samples 200Data: Planck 2018 COM_PowerSpect_CMB-TT-full_R3.01.txt (ESA/IRSA) Engine: ruvector-consciousness v2.1.0, exact algorithm (65,534 partitions evaluated) Statistical witness: 200 Monte Carlo null realizations, ChaCha8 PRNG seed=42 Timestamp: 2026-03-31 UTC