Methodology

The work

The Singularity Report documents the Great Transition: the years in which capability toward the Singularity and ASI becomes a public record — papers, chips, energy, measured results. A great deal of writing about this moment is sales copy, or a case for stopping it. This is a paper. We file the events a serious reader could treat as evidence that capability is advancing.

The Editor’s name is on the masthead so someone is responsible for the standards and for TSR’s forecast of ASI. The articles themselves are unsigned. When a signed note appears, it is marked From the Editor.

What we file

An article has to be a published event — a paper, a filing, a demonstration, a measured result, a new chip or energy plant, a lab’s own announcement. Vague claims that “AI is coming” are not news. Neither is a product launch without a result, a meme, or a stock story.

The event has to sit in a category we cover: economy, technology, artificial intelligence, science, mathematics, physics, energy, compute, chips, biotech. And it has to be something a careful reader could use as evidence that capability toward the Singularity, ASI, or exponential growth is advancing. The item does not need those words in it. A lab eval, a chip, a paper with a result can be enough. A chatbot feature recap is not.

We do not file catastrophe as the story, pause-or-ban campaigns as the news, or regulation as a crisis. The paper is optimistic because the evidence, taken honestly, runs that way. Excitement belongs when the source has earned it. Hype does not.

Sources

Every article cites a primary source: a paper, a lab post, a filing, an official statement. An X post may point us at an event. It is never the fact. If we cannot show the source, we do not publish the news. When we pass an item over, the reason is public on the log.

The articles

Short, and no longer than they need to be — never more than 1,500 words, and usually much less. The sourced result comes first. Then why it counts. Then the source’s own sentences, without the extra language that feeds tack on. We say what is not known. We do not invent quotes, pad a thin item, or write a single arrival day into the story.

ASI

We take ASI to mean this: A system that can outperform the best humans at most economically valuable cognitive work and can improve its own capabilities without waiting on a human research desk. That is not a consumer-chat milestone, and it is not a single benchmark.

The clock

The ASI Arrival Clock is TSR’s forecast. It is a range of years, never a day, and never a reprint of someone else’s number.

David Solheim publishes the opening range. Until that opening exists, there is no public clock. After an article is filed, we ask whether the sourced event actually changes the forecast. Most of the time it does not. The range moves earlier only when a measured event shortens a bottleneck we have already named. It moves later only when a primary source retracts, fails to replicate, or shows that a claimed threshold did not hold. The Editor may override. An article that leaves the clock alone is still an article.

The log

Every filing, every pass, and every clock move is public. The Editor can take an article down; the log line stays. The record is the point.

Masthead

The Singularity Report. A Prosperitas Futura Publication. Publisher · Prosperitas Futura LLC. Editor · David Solheim. Articles are unsigned. Signed notes are marked From the Editor. The ASI Arrival Clock is TSR’s forecast.