Machine Learning

commentary

Progress or Peril? The Brave New World of Self-Driving Science Labs

Self-driving laboratories (SDLs) promise to reshape our very understanding of research. But, as with all groundbreaking innovations, SDLs bring their own set of intriguing questions and potential challenges.

Joshua Steier, Rushil Bakhshi

blog

The Four-Day School Week, AI and Social Media Manipulation, School Safety: RAND Weekly Recap

This weekly recap focuses on the costs and benefits of a four-day school week, how artificial intelligence is bringing a new era of social media manipulation, the effects of placing police officers in schools, and more.

commentary

Money, Markets, and Machine Learning: Unpacking the Risks of Adversarial AI

The consequences of ignoring the problem of adversarial attacks in algorithmic trading are potentially catastrophic. In a world increasingly reliant on machine learning models, the financial sector needs to shift from being reactive to proactive to ensure the security and integrity of our financial system.

Joshua Steier, Sai Prathyush Katragadda

blog

Immigration Policy, the Black-White Wealth Gap, ChatGPT: RAND Weekly Recap

This weekly recap focuses on the challenges facing U.S. immigration policy, what it would take to close America's Black-white wealth gap, risks and opportunities associated with artificial intelligence, more.

q&a

AI, ChatGPT, and Language as Technology: Q&A with William Marcellino

William Marcellino, a senior behavioral and social scientist at RAND and professor at the Pardee RAND Graduate School, discusses the rapidly expanding reach of artificial intelligence, the challenges it could pose for both society and policymakers, and how the research community is poised to help.

blog

Putin's Gambits, Xi Jinping, Machine Learning: RAND Weekly Recap

This weekly recap focuses on Putin's latest “desperate measures” in Ukraine, support for Xi Jinping, whether machine-learning tools can tell if you're lying, and more.

commentary

Hack Post-Quantum Cryptography Now So That Bad Actors Don't Do It Later

The U.S. government should consider offering a public cash bounty to anyone who can crack the new forms of encryption that are being rolled out to defend against quantum computers. If a bounty helps catch a vulnerability before it's deployed, then the modest cost of the bounty could prevent much higher costs down the line.

Edward Parker, Michael J. D. Vermeer @MJDVermeer

blog

Truth Decay, School Lunch Nutrition, Regulating Space: RAND Weekly Recap

This weekly recap focuses on tackling Truth Decay, how tech experts view the military's use of artificial intelligence, a new approach to regulating outer space, and more.

blog

Helping Ukrainian Refugees, Truth Decay, Algorithmic Inequity: RAND Weekly Recap

This weekly recap focuses on how to help Ukrainian refugees, the link between cognitive processes and Truth Decay, tracking wastewater to understand the spread of COVID-19, and more.

commentary

Algorithms Teach Us Why Equity Is Hard

If all the shortcomings of humanity were stripped away, equity would still be an elusive goal for algorithms for reasons that have more to do with mathematical impossibilities than backward ideologies. But even if attaining equity is fundamentally difficult, seeking it is not futile.

Irineo Cabreros

commentary

Reining in COVID-19 Disinformation from China, Russia, and Elsewhere

During the pandemic, misinformation and conspiracy theories have spread more virulently than ever before. The vast scale of the problem means scalable solutions like machine learning could be needed to rein in the bots, trolls, and conspiracy theories being spread by bad-faith actors.

Christian Johnson, William Marcellino

blog

The Uyghurs, Getting the World Vaccinated, Policing in America: RAND Weekly Recap

This weekly recap focuses on China's disappeared Uyghurs, how the United States can help get the rest of the world vaccinated, what's next for policing in America, and more.

commentary

Towards an AI-Based Counter-Disinformation Framework

Artificial intelligence is being used to develop sophisticated malign information on social media. But AI also provides opportunities to strengthen responses to these threats and can foster wider resilience to disinformation.

Linda Slapakova

blog

Keeping COVID-19 Vaccines Moving, the Capitol Attack, Media Literacy: RAND Weekly Recap

This weekly recap focuses on keeping COVID-19 vaccines moving to save more lives; why we need a national commission to investigate the U.S. Capitol attack; media literacy as a tool to counter "Truth Decay," and more.

commentary

Artificial Intelligence and Machine Learning Are Important Tools to Improve Cost Estimation for Natural Disasters in Electric Utilities

Natural disasters in the United States cause billions of dollars of damage to electric infrastructure every year. Applying artificial intelligence and machine learning in a disaster-recovery context for electrical utilities might significantly improve cost estimating capability and responsiveness.

Ismael Arciniegas Rueda, Parousia Rockstroh

blog

COVID Learning Loss, Russian Trolls, Artificial Intelligence: RAND Weekly Recap

This weekly recap focuses on helping students recover learning losses, fighting Russian trolls, racial disparity in unemployment benefits, the race for AI leadership, and more.

commentary

A Machine Learning Approach Could Help Counter Disinformation

Disinformation has become a central feature of the COVID-19 crisis. This type of malign information and high-tech “deepfake” imagery poses a risk to democratic societies worldwide by increasing public mistrust in governments and public authorities. New research highlights new ways to detect and dispel disinformation online.

Kate Cox, Linda Slapakova, et al.

commentary

How to Train Your AI Soldier Robots (and the Humans Who Command Them)

While AI-enabled robots will have human-like characteristics, they will likely develop distinct personalities of their own. The military will need an extensive training program to inform new doctrines and concepts to manage this powerful, but unprecedented, capability.

Thomas Hamilton

commentary

First, Manage Security Threats to Machine Learning

Deception is as old as warfare itself. Until now, the targets of deception operations have been humans. But the introduction of machine learning and artificial intelligence opens up a whole new world of opportunities to deceive by targeting machines.

Rand Waltzman, Thomas S. Szayna