Sequelograph

A real event, read carefully. Then a fiction that does not pretend to be the news.

The record, 23 Sep 2026

Agents flag a phage enzyme system with DNA repeats

On 23 September 2026 Anthropic described a life-sciences group in which model agents searched reverse transcriptases and noticed, next to an unusual one in a jumbo phage, a repeat array reminiscent of CRISPR plus a neighboring gene of unknown function. The underlying enzyme had been seen before; the lab says the array and partner had not been described as a system. About 950 agents spent 21 hours and roughly 210 million tokens. Early human experiments in the lab, at biosafety levels 1 and 2, found the array is expressed as distinct short RNAs. The function is unknown, and the write-up points to a preprint rather than a finished mechanism.

Read the source at Anthropic

The record ends here. Everything below this line is invented.

What if a gene-editing tool was named before anyone knew what it did?

Part one

The Repeat Array

Theo ran the gel the agents could not run. They had found the pattern, a reverse transcriptase in a phage, a partner gene, a stack of repeats spaced like a memory bank. He had the protein in a harmless lab strain and the RNA on a gel, short bands, distinct, one for each repeat the way a programmable system would advertise itself.

He did not know what the bands did. Cutting, copying, and pasting were words from the cousin systems, not measurements from this one. The morning note from the patent desk asked for a use. A use was how a filing became more than a sequence.

Theo labeled the bands with their lengths and wrote, under the last one, function unknown.

The partner protein, which the note assumed would sit with the enzyme, came down in a different fraction of the cell, away from it.

He photographed the gel before anyone could crop that lane out.

What’s real

  • Agents linked a phage reverse transcriptase to a repeat array
  • The enzyme itself had been reported before
  • Early experiments found the array makes distinct short RNAs
  • The biological function is still unknown

What’s invented

  • Theo's patent desk asking for a use
  • A partner protein in the wrong fraction
  • A photograph taken to stop a crop
  • A filing that wants a function by morning

Part two: The Lane They Would Crop

Theo has the gel showing short RNAs and a partner in the wrong place, and the patent desk still wants a use.

Part two isn’t on sale yet. Check back soon.

Filed under biology, genomics, enzymes, phage biology.

More records

How it works

Each record starts with one real science event, with a link to where it was reported.

A short story then asks what happens next. Part one is free to read.

Part two costs a few dollars, paid once. You keep it.