What has grown out of the essays

A public forecast of one question about one life. The name comes from the Belters, the asteroid-belt workers in The Expanse, the science-fiction novels by James S. A. Corey. The forecast breaks the question into eight factors, the way the astronomer Frank Drake's equation breaks the number of civilisations in the galaxy into seven, and under each factor sits a tree of concrete events, about seventy in all, each with a plain test of what would count and a probability. The headline is worked out from the tree and never typed in. Every week the site reads the news for anything that bears on an event and records what moved, and every January the forecast is scored against what actually happened. A visitor can put in their own birth year and see what the tree says for them.

A dated and sourced record of how expert and public opinion about artificial general intelligence, meaning AI that can do most of the intellectual work a person can, has moved since 2005. Every statement, survey and market reading links to where it came from, and the whole dataset can be downloaded. Beside the record sits a test of a different question: has the world entered the singularity, the idea that machines get smarter so fast that the world changes faster than anyone can keep up with? The test has four parts, each with a line fixed in advance (how fast AI capability is doubling, whether forecasts of it still work, how much of AI research AI now does, and whether the world economy has changed speed), and the site scores them from public numbers.

A public explainer and tracker for the oldest question in this project. The galaxy is old enough and big enough that someone else should have shown up by now, and nobody has. The site takes the Drake Equation, the astronomer Frank Drake's 1961 way of writing the number of detectable civilisations in our galaxy as seven quantities multiplied together, gives each quantity the range the published estimates support, plays the multiplication out two hundred thousand times, and reports the spread rather than a single number. Every serious explanation for the silence has a page of its own at an address that never changes, and every paper read against the numbers is logged whether or not it moved one.

A public count. It lists the specific ways a person can be changed on purpose, from a cochlear implant to a brain chip, sorted under the seven things about a person that can be changed: what they sense, how they move, how they think and remember, how they feel, how long they last, what they pass to their children, and how directly they can be joined to a machine. Each upgrade carries a plain test of what would count as a person having it, the public record that settles it, and a stage on a five-rung ladder from idea only to in wide use. There are no probabilities: a stage, a date and a record say what has happened, and the site does not guess about what will. Beside the count sits a dated log of what people have claimed about merging with machines, checked against the recordings.

A public tracker of the race between medicine and ageing. Every year you live, medicine gets a little better, but not yet fast enough: for a person in middle age, the years they can expect to have left fall by about a year every year, less the small gains new treatments add. The site asks when that stops being true. It breaks the question into eight branches, seven kinds of damage that build up in a body over a lifetime and the tools that would show whether any repair worked, and under each branch the specific treatments that would have to arrive, each with a plain test, a stage, and the project's estimate of when it reaches the clinic. The year at the top is played out from those estimates two hundred thousand times. Beside it sits a dated record of who promised what and when, held against what has actually been shown. It recommends nothing and nothing on it is medical advice.

The experimental side of the project. The essays argue that consciousness is something a system does, the ongoing work of binding its moments into one point of view, and if that is right it should have parts that can be found and tested. The programme builds small systems and measures them against that account, one component at a time, and every experiment says beforehand what result would count against it. None of them claims to settle whether a machine is conscious; the honest output is a position on a scale, never a verdict. The site is the programme's log: where it stands, what is waiting on a decision, what it has learned so far, the plan ahead, and every dollar spent.

The project's first book, and the place the argument is made at length. The classical safety story asks how to make AI serve us; the book works the harder questions underneath it: how to tell what a system is when its inside cannot be verified, and what we owe it before the metaphysics resolves. Attributing experience to anyone runs the same inference from outward signs to inner state; machines only make the bet visible again. The full text can be read here while the book is being finished, and every substantive revision is dated.