Researchllmhuman ai variancepersonalizationon device privacy
Posterum Measures Human-Likeness With HAVS Index
7.1
Posterum Software published a recent paper introducing the Human–AI Variance Score (HAVS) and a personalized Posterum AI search app on Google Play. The study ran 16 synthetic profiles through four LLMs (ChatGPT, Claude, Gemini, DeepSeek) and compared over 1,000 answers to Gallup, Pew, and YouGov survey questions; ChatGPT and Claude scored above 94, economics questions scored lowest (86–89).
Key Points
- 1Develops Human–AI Variance Score (HAVS) to compare model answers to human survey responses
- 2Finds ChatGPT and Claude most human-like, models score highest on morality and political questions
- 3Enables personalized, on-device profiles to tailor responses, offering transparent perspective-aware search
Scoring Rationale
Provides a novel, actionable human-likeness metric with clear results; limited by single-company source and non-peer-reviewed validation.
Sources
Public references used for this report.
Practice with real Logistics & Shipping data
90 SQL & Python problems · 15 industry datasets
Used by DS/ML engineers at top companies
High-Value Overnight OrdersEasyDelivered International ShipmentsMediumOn-Time Delivery Rate by CarrierHard
250 free problems · No credit card
See all Logistics & Shipping problems
