Data, algorithms and models

Will AI Make Cyber Swords or Shields?

Andrew Lohn Krystal Jackson
| August 2022

Funding and priorities for technology development today determine the terrain for digital battles tomorrow, and they provide the arsenals for both attackers and defenders. Unfortunately, researchers and strategists disagree on which technologies will ultimately be most beneficial and which cause more harm than good. This report provides three examples showing that, while the future of technology is impossible to predict with certainty, there is enough empirical data and mathematical theory to have these debates with more rigor.

China’s Advanced AI Research

William Hannas Huey-Meei Chang Daniel Chou Brian Fleeger
| July 2022

China is following a national strategy to lead the world in artificial intelligence by 2030, including by pursuing “general AI” that can act autonomously in novel circumstances. Open-source research identifies 30 Chinese institutions engaged in one or more of this project‘s aspects, including machine learning, brain-inspired AI, and brain-computer interfaces. This report previews a CSET pilot program that will track China’s progress and provide timely alerts.

China’s Industrial Clusters

Anna Puglisi Daniel Chou
| June 2022

China is banking on applying AI to biotechnology research in order to transform itself into a “biotech superpower.” In pursuit of that goal, it has emphasized bringing together different aspects of the development cycle to foster multidisciplinary research. This data brief examines the emerging trend of co-location of AI and biotechnology researchers and explores the potential impact it will have on this growing field.

CSET’s Map of Science reveals that Germany leads the world in robotics for automotive engineering.

Exploring Clusters of Research in Three Areas of AI Safety

Helen Toner Ashwin Acharya
| February 2022

Problems of AI safety are the subject of increasing interest for engineers and policymakers alike. This brief uses the CSET Map of Science to investigate how research into three areas of AI safety — robustness, interpretability and reward learning — is progressing. It identifies eight research clusters that contain a significant amount of research relating to these three areas and describes trends and key papers for each of them.

AI and Compute

Andrew Lohn Micah Musser
| January 2022

Between 2012 and 2018, the amount of computing power used by record-breaking artificial intelligence models doubled every 3.4 months. Even with money pouring into the AI field, this trendline is unsustainable. Because of cost, hardware availability and engineering difficulties, the next decade of AI can't rely exclusively on applying more and more computing power to drive further progress.

It is common for observers to compare machine intelligence with individual human intelligence, but this tendency can narrow and distort understanding. Rather, this paper suggests that machines, bureaucracies and markets can usefully be regarded as a set of artificial intelligences that have been invented to complement the limited abilities of individual human minds to discern patterns in large amounts of data. This approach opens an array of possibilities for insight and future investigation.

Trends in AI Research for the Visual Surveillance of Populations

Ashwin Acharya Max Langenkamp James Dunham
| January 2022

Progress in artificial intelligence has led to growing concern about the capabilities of AI-powered surveillance systems. This data brief uses bibliometric analysis to chart recent trends in visual surveillance research — what share of overall computer vision research it comprises, which countries are leading the way, and how things have varied over time.

In an opinion piece for The Diplomat, Ellen Lu and CSET's Ryan Fedasiuk examine whether China's new data regulations will hamper AI ambitions.

Measuring AI Development

Jack Clark Kyle Miller Rebecca Gelles
| December 2021

By combining a versatile and frequently updated bibliometrics tool — the CSET Map of Science — with more hands-on analyses of technical developments, this brief outlines a methodology for measuring the publication growth of AI-related topics, where that growth is occurring, what organizations and individuals are involved, and when technical improvements in performance occur.