Výsledky vyhledávání - "Tetlock, Philip E"

Report

Identifying good forecasters via adaptive cognitive tests

Autor: Merkle, Edgar C., Petrov, Nikolay, Zhu, Sophie Ma, Karger, Ezra, Tetlock, Philip E., Himmelstein, Mark

Assessing forecasting proficiency is a time-intensive activity, often requiring us to wait months or years before we know whether or not the reported forecasts were good. In this study, we develop adaptive cognitive tests that predict forecasting pro

Externí odkaz: http://arxiv.org/abs/2411.11126

Zobrazit plný text záznamu

Report

ForecastBench: A Dynamic Benchmark of AI Forecasting Capabilities

Autor: Karger, Ezra, Bastani, Houtan, Yueh-Han, Chen, Jacobs, Zachary, Halawi, Danny, Zhang, Fred, Tetlock, Philip E.

Forecasts of future events are essential inputs into informed decision-making. Machine learning (ML) systems have the potential to deliver forecasts at scale, but there is no framework for evaluating the accuracy of ML systems on a standardized set o

Externí odkaz: http://arxiv.org/abs/2409.19839

Zobrazit plný text záznamu

Report

Wisdom of the Silicon Crowd: LLM Ensemble Prediction Capabilities Rival Human Crowd Accuracy

Autor: Schoenegger, Philipp, Tuminauskaite, Indre, Park, Peter S., Tetlock, Philip E.

Human forecasting accuracy in practice relies on the 'wisdom of the crowd' effect, in which predictions about future events are significantly improved by aggregating across a crowd of individual forecasters. Past work on the forecasting ability of la

Externí odkaz: http://arxiv.org/abs/2402.19379

Zobrazit plný text záznamu

Report

AI-Augmented Predictions: LLM Assistants Improve Human Forecasting Accuracy

Autor: Schoenegger, Philipp, Park, Peter S., Karger, Ezra, Trott, Sean, Tetlock, Philip E.

Large language models (LLMs) match and sometimes exceeding human performance in many domains. This study explores the potential of LLMs to augment human judgement in a forecasting task. We evaluate the effect on human forecasters of two LLM assistant

Externí odkaz: http://arxiv.org/abs/2402.07862

Zobrazit plný text záznamu