Machine Learning

139 Results

number of items to display at a time

Expert Insights

Oct 14, 2025

Pivots and Pathways on the Road to Artificial General Intelligence Futures

The authors of this paper introduce a framework for policy discussions about artificial general intelligence (AGI), showing how six key dimensions can be used to classify and compare different combinations of factors and potential AGI futures.
Tool

Sep 29, 2025

The RAND Tool for Evaluating Algorithmic Bias: A User Guide

With the RAND Tool for Evaluating Algorithmic Bias, a user can assess and correct performance differences in binary predictive algorithms across groups. This guide explains how to use the tool, even when complete group information is unavailable.
Article

Aug 8, 2025

Dr. Whiskers and a Trip to Hogwarts: Adventures with Generative AI

Aisha Nájera participated in a contest to coax bad answers from AI models—part of a federal effort to make the models safer and more trustworthy by exposing their weaknesses. She outmaneuvered different chatbots more than a dozen times, proving that you can't always trust what you get from them.
Expert Insights

Jun 24, 2025

Toward Best Practices for AI Evaluation and Governance: A Proposal for a European Union General-Purpose AI Model Evaluation Standards Task Force

The authors propose a task force in the European Union that would develop standards for general-purpose artificial intelligence evaluations, including the task force’s responsibilities, potential influence, and possible sources of failure.
Research

Jun 12, 2025

Automated Grading for Efficiently Evaluating the Dual-Use Biological Capabilities of Large Language Models

The authors used an automated grader to evaluate the ability of large language models to answer knowledge-based questions and generate protocols that explain how to perform laboratory techniques that could be used to create biological threat proxies.
Commentary

Jun 10, 2025

We'll Be Arguing for Years Whether Large Language Models Can Make New Scientific Discoveries

If AI can indeed make scientific discoveries, that would not only have practical impacts for society but would also provide evidence that we've achieved true digital intelligence. But reaching expert consensus on what counts as a scientific discovery by an AI may prove more elusive than expected.
Commentary

Jun 6, 2025

Mitigating Emerging Human Intelligence Challenges with Forecasting

The decline in human intelligence capabilities likely necessitates a shift in intelligence collection sources. Probabilistic crowdsourced forecasting could offer a way to mitigate collection gaps and complement traditional methods by combining human expertise and judgments with data-driven insights.
Research

Jun 2, 2025

One Team, One Fight: Volume I, Insights on Human-Machine Integration for the U.S. Army

The authors investigate the difficulties that the Army might encounter as it attempts to pair humans with artificial intelligence algorithms to accomplish specific warfighting tasks and how these potential obstacles can be overcome.
May 29, 2025

Using GPT-4 for Title and Abstract Screening in a Literature Review of Public Policies: A Feasibility Study

This study explores the use of a large language model (LLM) to screen titles and abstracts in a review of public policy literature to assess the percentage of articles GPT-4 recommended for exclusion that should have been included.
Expert Insights

May 1, 2025

Preliminary suggestions for rigorous GPAI model evaluations

This document presents suggestions that promote the internal validity, external validity and reproducibility of general-purpose AI evaluations, including benchmark evaluations and human uplift studies, across four evaluation life cycle stages.
Research

Apr 28, 2025

Orbitomics: Economic Ripple Effects of ISAM on Space Surveillance

Orbitomics examines the economic impact of advanced technologies, Federated Learning, Edge Computing and Quantum, in Space Situational Awareness. This paper provides a Theory of Change for enhancing space security and governance by 2035.
Video

Apr 10, 2025

Why AI Projects Fail

Artificial intelligence (AI) projects fail at a much higher rate than information technology (IT) projects that do not involve AI. What are the root causes that lead to these failures—and what can be done to minimize them?

First
Previous
Page 2 of 12
Next
Last

Machine Learning

Machine Learning

Related Topics

Breaking Down the Federal Budget: New Tools from RAND

Dr. Whiskers and a Trip to Hogwarts: Adventures with Generative AI

Explore RAND’s Work on this Topic

Error

Pivots and Pathways on the Road to Artificial General Intelligence Futures

The RAND Tool for Evaluating Algorithmic Bias: A User Guide

Dr. Whiskers and a Trip to Hogwarts: Adventures with Generative AI

Toward Best Practices for AI Evaluation and Governance: A Proposal for a European Union General-Purpose AI Model Evaluation Standards Task Force

Automated Grading for Efficiently Evaluating the Dual-Use Biological Capabilities of Large Language Models

We'll Be Arguing for Years Whether Large Language Models Can Make New Scientific Discoveries

Mitigating Emerging Human Intelligence Challenges with Forecasting

One Team, One Fight: Volume I, Insights on Human-Machine Integration for the U.S. Army

Using GPT-4 for Title and Abstract Screening in a Literature Review of Public Policies: A Feasibility Study

Preliminary suggestions for rigorous GPAI model evaluations

Orbitomics: Economic Ripple Effects of ISAM on Space Surveillance

Why AI Projects Fail

Oops! An unexpected error has occurred.

RAND Headquarters

U.S. research divisions

International research divisions

Machine Learning

Machine Learning

Related Topics

Breaking Down the Federal Budget: New Tools from RAND

Dr. Whiskers and a Trip to Hogwarts: Adventures with Generative AI

Explore RAND’s Work on this Topic

Error

Pivots and Pathways on the Road to Artificial General Intelligence Futures

The RAND Tool for Evaluating Algorithmic Bias: A User Guide

Dr. Whiskers and a Trip to Hogwarts: Adventures with Generative AI

Toward Best Practices for AI Evaluation and Governance: A Proposal for a European Union General-Purpose AI Model Evaluation Standards Task Force

Automated Grading for Efficiently Evaluating the Dual-Use Biological Capabilities of Large Language Models

We'll Be Arguing for Years Whether Large Language Models Can Make New Scientific Discoveries

Mitigating Emerging Human Intelligence Challenges with Forecasting

One Team, One Fight: Volume I, Insights on Human-Machine Integration for the U.S. Army

Using GPT-4 for Title and Abstract Screening in a Literature Review of Public Policies: A Feasibility Study

Preliminary suggestions for rigorous GPAI model evaluations

Orbitomics: Economic Ripple Effects of ISAM on Space Surveillance

Why AI Projects Fail

Oops! An unexpected error has occurred.