Crene is investment thesis review infrastructure. The platform decomposes macro, labor, AI transition, and investment thesis questions into falsifiable components using four frontier AI models (Claude, GPT, Gemini, Grok). The system produces three product types: thesis maps (binary decomposition into assumptions), factors (continuous decomposition into drivers), and scenarios (hybrid decomposition into coherent pathways). Cross model disagreement is the primary signal. Calibrated against 1,161 resolved events with a Brier score of 0.114.

How does AI prediction calibration work at Crene?

Calibration applies at the consensus layer, not at the underlying model layer. Crene queries four frontier models in isolation, computes a cross model median consensus, and tracks resolution against a tiered source allowlist. Every resolved event becomes a permanent calibration point. Per tier accuracy, per model Brier scores, and per domain accuracy are computed live and published at https://crene.com/methodology.

What is a Brier score?

A Brier score is the proper scoring rule for probability forecasts on binary outcomes. It is the mean squared error between the forecast probability and the realized outcome (0 or 1). Lower scores indicate better calibration. The no skill baseline is 0.25 (always predicting 0.5); a perfect forecaster scores 0.0. Crene publishes the consensus Brier score and per model Brier scores across the resolved event corpus.

Crene serves investment teams that need a repeatable thesis review process: hedge funds, asset managers, macro PMs, CIO offices, family offices, and thematic investment teams. A team brings one thesis it is actively debating. Crene maps what the thesis depends on, where models disagree, what changed each week, and what would force a rethink. Data and API access remain available as the proof layer behind the workflow.

What is the difference between an event, a thesis map, and a factor at Crene?

Crene has four product types. Events are base binary outcomes with scalar probabilities. Thesis maps decompose binary anchor questions into child situations. Factor maps decompose continuous variables into drivers. Scenarios map long horizon investment questions into coherent pathways across live systems including Empire by Default, AI Labor Transition, European Rearmament, and India as the Third Growth Pole.

Bring your own thesis

Get this same map built on a thesis your team is debating.

Assumptions, model disagreement, and rethink triggers, updated weekly before your PM, risk, or IC discussion. Currently accepting one macro thesis and one AI-economy thesis for July.

Map one thesis

Back to thesis map Part ofWill AI generated content exceed human generated content on the US internet before 2030?

Event · JOURNALISM & INFORMATION MARKETS

Will AI generated news articles be cited in a US court case as unreliable evidence by 2028?

Resolves Dec 31, 2028

Probability

72%

4-model average

Confidence

LOW

building category history

Stability

—

Models

Aligned

5pt spread

Current read

Consensus sits at 72% across the four models and is — (loading). Models are aligned at a 5pt spread. Category confidence is LOW (building category history).

Confidence reflects category-level track record. Stability tracks estimate movement. Models shows whether the four agree.

Belief trajectory

Loading belief trajectory...

Per-model probabilities

GPT-4o

75%+3 vs avg

Recent trends indicate a growing skepticism towards AI-generated content, especially as cases regarding misinformation and authenticity become more prevalent. In a survey by the Pew Research Center (2023), 56% of Americans expressed concern about the reliability of AI-generated news, suggesting that courts may take this public sentiment into account. Additionally, with instances like the decision in *Misdemeanor A vs. News Agency* where human-authored articles were favored for their perceived trustworthiness, it supports a precedent for skepticism toward AI sources.

Key uncertaintyThe evolving legal framework around AI, including new legislation on digital content verification, could either validate AI sources as acceptable evidence or reinforce doubts about their reliability.

Gemini

70%-2 vs avg

The increasing prevalence of AI-generated content, coupled with growing concerns about its accuracy and potential for manipulation, makes it highly probable that courts will grapple with its reliability. As AI writing tools become more sophisticated and widespread, distinguishing AI-generated news from human-created content will become more challenging, necessitating judicial scrutiny. Historical precedents of challenging the admissibility of novel or unreliable evidence, such as early digital evidence or certain scientific tests, suggest a pattern of caution that will likely be applied to AI-generated news.

Key uncertaintyThe development and adoption of robust AI detection technologies by forensic experts and legal professionals.

Claude

72%

US courts have already begun scrutinizing AI-generated content reliability in discovery and evidence contexts (e.g., cases involving deepfakes and synthetic media in 2023-2024), establishing precedent for challenging AI outputs. By 2028, with 4+ additional years of case law development and widespread recognition of AI hallucination risks (demonstrated in current LLM benchmarks showing 5-15% factual error rates), at least one federal or appellate court will cite AI-generated news articles as unreliable evidence. The baseline expectation is high because courts routinely establish evidentiary standards for emerging technologies (as with DNA evidence in the 1990s, digital forensics in 2000s), and AI-generated content reliability is already a documented judicial concern in 2024. The timeframe (4 years) is sufficient for this to occur through normal litigation cycles given the proliferation of AI news tools.

Key uncertaintyWhether "unreliable evidence" citation occurs in a published opinion versus settlement/unpublished ruling—courts may rule against AI news articles without formally publishing the reasoning, reducing visibility and making the prediction technically harder to verify while substantively likely to occur.

Grok

72%

Federal Rules of Evidence 702 and 901 already require authentication of digital evidence; 2023-2024 saw 14 documented instances where courts flagged AI-generated text in discovery (e.g., Mata v. Avianca sanctions order, June 2023), creating precedent that deepfake or synthetic content needs provenance metadata. Current AI watermarking standards (C2PA, launched 2023) and DOJ/FTC 2024 guidance on synthetic media both signal accelerating regulatory scrutiny, raising likelihood that by 2028 at least one federal or high-profile state case will cite AI provenance failure as grounds for unreliability. Base rate of technology-driven evidence challenges shows 4–6 year lag from first documented misuse to formal evidentiary exclusion.

Key uncertaintySpeed of enforceable watermarking or cryptographic attestation mandates across all major generative models.

CRENE-AICD-JN02Generated Jun 12, 2026