| paper | d4-journals-jet-2018-0058 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
Proposition 1 characterizes preference properties; it does not assert that an object can be computed or that computing it is hard. Replacing the finite sum with an integral gives a plausible analytic extension, while adding resource constraints and an optimization objective would introduce a new computational problem.
fails bit a — no named computational result to mirror
The proposed population extension covers Proposition 1's characterization of time-consistent and time-invariant welfare weights; it yields no computational result.
The paper has a natural population-continuum extension, but it contains no named computational result of the kind ChoCo requires as an anchor. So I can’t honestly give a continuous computational problem per anchor.
The closest result is Proposition 1, proved in this paper. It says that if at every history at least two individuals receive positive welfare weight and have different discount factors, then utilitarian social preferences are time consistent if and only if the weights take the form \(w_i(h_\tau)=w_0^i\beta_i^\tau/\sum_j w_0^j\beta_j^\tau\); they are time invariant if and only if the weights are constant, \(w_i(h_\tau)=y_i\). This is a characterization of preference properties, not a claim about an algorithm, complexity class, or computational problem.
A plausible continuum setting would be a large polity managing a shared resource: citizens have types \(\theta=(U_\theta,\beta_\theta)\), and \(\mu_\theta\) is each type’s share of the population. For a common consumption path \(c\), aggregate welfare could be written as
\[
W_\tau(c)=\int w_\tau(\theta)\sum_{t=0}^{\infty}U_\theta(c_t)\beta_\theta^t\,d\mu(\theta).
\]
A continuum analogue of Proposition 1 would characterize when this social preference is time consistent or time invariant. That is a plausible analytic extension: the authors could recognize the population model, and a large, stable population with relatively few preference types is credible. But it remains an axiomatic characterization. Turning it into a resource-allocation optimization problem would add a computational task this paper does not study.
That is the weak point, and it is decisive under the programme’s scope: the population can be continuized, but the paper’s results do not ask computational questions about it. The closest mirror is outside scope, so there are no qualifying anchors or per-anchor prover questions to give.
The strongest case against a ChoCo mirror is that this paper supplies no computational result to mirror. Proposition 1 characterizes when utilitarian social preferences are time consistent or time invariant. It defines no computational problem, instance, algorithm, or complexity question. Its application to households or shared resources may be important, but importance of the setting does not supply the missing computational task.
The proponent’s population model is plausible: a large polity could have a distribution \(\mu\) over preference types \(\theta=(U_\theta,\beta_\theta)\). Replacing the finite sum by an integral gives an analytic extension of the paper’s framework. Under suitable conditions, the time-consistent weights would follow the same reweighting pattern, with \(w_\tau(\theta)\) proportional to \(w_0(\theta)\beta_\theta^\tau\). That makes the population continuous, but it does not turn Proposition 1 into an algorithmic result. The aggregate remains a welfare representation whose properties are being characterized.
A stronger proposal might use the paper’s shared-resource examples and ask for a feasible extraction path that maximizes collective welfare while preserving time consistency. But that requires adding a resource constraint, a representation of utilities and the type distribution, and a rule for choosing welfare weights. Those choices define a new optimal-control problem; they are not computational questions already posed or answered by this paper. A high-multiplicity population could make that new problem sensible, but it would not make it a continuous mirror of the paper’s results.
So the negative case is strong as a scope argument, though not a proof that no worthwhile follow-on could be built. The continuum of preference types is credible. What is missing is a paper-grounded computational question for ChoCo to classify.
The adversarial triple: the proponent anchors on up to three named results; the opponent sees that case and must defeat every anchor; the judge decides which case convinced it. These are the pipeline’s own outputs, generated by tools/triple_run.py — no human edited them. The paper’s own text is not reproduced here beyond the quoted statement above.