OVERWRLD
Next

The data center is out of room.

Power, interconnect and floor space are the binding constraints on AI infrastructure, and none of them can be bought on the timescale that demand is growing. Compression is the only lever that makes what is already built carry more.

01/The economics

You cannot pour concrete fast enough.

New capacity arrives on a multi-year schedule and is gated by grid connections nobody controls. Meanwhile the bytes keep growing: bigger datasets, bigger checkpoints, more traffic between clusters that were never meant to be one machine.

Interconnect is the bottleneck

Training at scale is increasingly limited by what moves between nodes and sites, not by what any single node can compute.

Datasets are written once and read forever

That is precisely the shape the engine is built for: spend the compute at encode, keep every read cheap.

Every byte carries an energy cost

Storage, transfer and the cooling that follows them all scale with volume. Fewer bytes is less of all three.

02/Deployment

Where it fits

Data movement inside an AI estate
  1. 01deployable now
    Training corpora

    Written once, read across many runs and many clusters.

  2. 02deployable now
    Checkpoints

    Written constantly, read on restart and for evaluation. Pure encode-once traffic.

  3. 03deployable now
    Cross-site transfer

    Moving datasets and weights between regions, billed by the terabyte.

  4. 04standards-bound
    Live inference path

    Latency-critical serving, where decode budget is measured in microseconds.

Highlighted stages are inside infrastructure you operate, where a decoder is a library you link rather than a standard you wait for.

03/Outcomes

What changes

  • 01More effective capacityThe same disks and the same links carry more, without a procurement cycle.
  • 02Cheaper movement between clustersCross-region and cross-site transfer is a line item that scales directly with volume.
  • 03Headroom bought with compute, not concreteEncode compute is elastic and schedulable. Building a data center is neither.
04/Evidence

What we have measured, and what we have not.

The public result is on text data. We have not published figures for training corpora, checkpoints or any other AI workload, because we have not measured them under a protocol we would be willing to defend. If your data is the interesting case, the honest answer is to run it: your corpus, your tuned baseline, decompressor counted.

05/Next step

Measure it on your own data.