The Long Reign
For five and a half million ticks, a world of evolving programs was ruled by a single family. Each of its thirteen rulers was the child of the one before. Clever minds rose four times at the edges, and each time they faded.
One family, thirteen rulers
After the first 100,000 ticks, no unrelated lineage ever took the throne. The leader was always replaced by one of its own daughter lineages: a branch whose genes had drifted far enough from its founder to earn a new name. Rule passed down, never across. The only interruptions were brief 50,000-tick spells when a crystal-mining cousin or a sister lineage briefly outnumbered the ruler.
Every ruler peaked between 176 and 235 members. Whatever each one looked like inside, the rich lit land held about two hundred of them.
Minds that grew, and shrank
Four times, big minds appeared: three times among the crystal miners on poor ground, once among the programs living around the vents. Each time the minds grew to 130–240 instructions. Each time the lineage either died out or shed its code, and the lines that shed code went on to thrive.
Inside the biggest mind
APO-7707 carried 189 instructions in seven routines, but its genes let it run only 72 per tick. Every tick it ran out of budget partway through routine three. Four whole routines, about a hundred instructions, never executed. Most were near-copies of earlier routines: gene duplication, the raw material of innovation, carried but never put to work.
What did run was a sharp controller. It amplified a tiny left-right difference in energy about 420 times and turned toward it. Its body was more inventive than its mind: eight mouths with catalysis from 0.16 to 0.97, a toolkit for crystal deposits of every hardness.
; APO-7707, routine 54838 (runs every tick) IN r4, in0 ‹energy left-right› MULK r0, r4, 14.31 ; learned in life (born 15.37) MULK r4, r0, 29.30 ; ~420× gain TURN r4 ; steer hard toward food NEG r7, r4 IFN r7 MULK r7, r3, -4.15 MARK r6, ch2 ; mark territory GO r7 EAT r7 ; routines 4–7: ~100 instructions, never reached
Lesson: code size is a poor measure of a mind. From v0.7 on, every report measures the instructions actually run each tick, not just the code carried.
The quiet middle
Novelty came in a rush and then dried up. Forty-two new behaviors and senses were recorded in the first half-million ticks. Between 2.5 and 4.5 million, almost none. In a world that never changes, once a good way of life is found there is little left to discover. A small revival came at the end, as vent-dwellers and crystal miners split into different ways of life.
Sitters and trapliners
Near the end of the run, watching the live view, Ryan noticed that crystal eaters move a lot while other lineages stay by the vents. Some of those vent-sitters seemed to get brave and wander off. A 39,000-tick recording (ticks 5.03M–5.07M) tested each observation.
| Lineage | Way of life | Time near a vent | Speed / tick | Typical range | Furthest member |
|---|---|---|---|---|---|
| TNK-8954 | Crystal trapliner, 90% crystal | 2% | 0.47 | 154 | 210 |
| TRN-9267 | Crystal trapliner, 85% crystal | 2% | 0.47 | 168 | 206 |
| GIB-8581 | Active vent group | 62% | 0.15 | 26 | 203 |
| TRN-9119 | Vent-sitter | 82% | 0.07 | 3 | 182 |
| SMT-9307 | Vent-sitter | 97% | 0.02 | 1 | 117 |
| GIB-9299 | Vent-sitter | 99% | 0.01 | 2 | 3 |
Ranges are in world units (the world is 300 across); "typical" is the median member's greatest distance from where it was first seen.
All three observations held. The crystal eaters were the fastest movers on the Grid. The vent-sitters mostly never left, but in several lineages a rare member crossed half the world: exploratory dispersal, the way new sources get found. And the miners had not learned restraint. Not one of 815 crystal cells was ever near full. They cropped crystal as fast as it regrew, making the rounds like bees among flowers. That is why the booms stopped: nothing was ever left to build up.
Sex, a minority habit
Mating was never common. Finding a willing partner costs time and energy, while budding alone costs only a little more energy and is always available. Weighted by births, mating never supplied more than about 8% of new programs, and it fell as low as 1%.
A test for v0.7: in nature, sex tends to pay when the world keeps changing or when enemies keep adapting (the "Red Queen"). v0.6 had neither. v0.7 has a drifting climate and wild game that evolves. If mating rises, that's the Red Queen at work.
Questions for v0.7
- Do minds hold their size when the world moves? Seasons, a drifting climate and vents that die should reward memory and senses instead of copied code.
- Will a dynasty still rule? Or will changing ground let unrelated lineages take turns.
- Does sex return? See chapter five.
- What will be learned first: hunting the grazers and browsers, gnawing and cracking nuts, or caching food for the lean season?