“Within five years of current trends, the cost to train a model…decreases roughly by a factor of 1000 or to around 0.1 percent of the original cost, making training vastly cheaper and increasing accessibility,” finds Paul Scharre in a new report. cnas.org/publications/r…

What is meaningful human control? Michael C. Horowitz & Paul Scharre: 'The information they have to make that decision is sufficient for them to ensure the lawfulness of the action they are taking, given what they know about the target, the weapon & context' files.ethz.ch/isn/189786/Eth…

The Pentagon seeks to deploy thousands of AI-enabled drones. In this Vox piece, Paul Scharre from CNAS emphasizes the necessity of autonomy to manage the vast control systems and mitigate personnel costs for future conflicts.

🔗Read more:
cnas.org/press/in-the-n…

ICYMI: Paul Scharre testified to Senate Armed Services Committee and SASC GOP on 'Preserving U.S. Military Advantage Amid Rapid Technological Change'. Read his testimony here: cnas.org/publications/c…

“Rising costs to train frontier AI models may drive an oligopoly at the frontier of research, but capabilities are likely to proliferate rapidly,” Paul Scharre finds in his new report. More: cnas.org/publications/r…

“Within five years of current trends, the cost to train a model…decreases roughly by a factor of 1000 or to around 0.1 percent of the original cost, making training vastly cheaper and increasing accessibility,” finds Paul Scharre in a new report. cnas.org/publications/r…

“If algorithmic improvements continue to be widely available, hardware-restricted actors will be able to train models with capabilities equivalent to once-frontier models only two to three years behind the frontier.” Read more from Paul Scharre’s report. cnas.org/publications/r…

🚨New CNAS Report🚨 “Future-Proofing Frontier AI Regulation: Projecting Future Compute for Frontier AI Models” from Paul Scharre. Read more: cnas.org/publications/r…

We are LIVE for a virtual event on the new CNAS report, 'Autonomy and International Stability: Confidence-Building Measures for Uncrewed Systems in the Indo-Pacific' with report author Tom Shugart, in conversation with Paul Scharre. Join now! #CNAS2024

twitter.com/i/broadcasts/1…

“The only way that thousands of drones work is if you have some measure of autonomy in the drones,” Paul Scharre told Vox. Read more: cnas.org/press/in-the-n…

“If algorithmic improvements continue to be widely available, hardware-restricted actors will be able to train models with capabilities equivalent to once-frontier models only two to three years behind the frontier.” Read more from Paul Scharre’s report. cnas.org/publications/r…

“If algorithmic improvements continue to be widely available, hardware-restricted actors will be able to train models with capabilities equivalent to once-frontier models only two to three years behind the frontier.” Read more from Paul Scharre’s report. cnas.org/publications/r…

ICYMI: Paul Scharre joined Fareed Zakaria to discuss the future of AI in warfare.

ICYMI: Paul Scharre testified to Senate Armed Services Committee and SASC GOP on 'Preserving U.S. Military Advantage Amid Rapid Technological Change'. Read his testimony here: cnas.org/publications/c…

Watch Paul Scharre’s full interview with Fareed Zakaria👇

🚨New CNAS Report🚨 “Future-Proofing Frontier AI Regulation: Projecting Future Compute for Frontier AI Models” from Paul Scharre. Read more: cnas.org/publications/r…

You may worry about AI taking your job, but think about this: Fully autonomous weapons could soon be a reality. How frightened should we be?

From today’s show, my conversation with author, AI-warfare expert and former Army Ranger Paul Scharre of @CNAS:

“Rising costs to train frontier AI models may drive an oligopoly at the frontier of research, but capabilities are likely to proliferate rapidly,” Paul Scharre finds in his new report. More: cnas.org/publications/r…

Register now for an event on a new report from Tom Shugart, Autonomy and International Stability: Confidence-Building Measures for Uncrewed Systems in the Indo-Pacific, in conversation with Paul Scharre. Register for this event on 3/26. cnas.org/events/virtual…

As Washington debates AI, our competitors have offered their own solutions for how to balance AI safety, innovation, and competition. America needs its own answer—soon. U.S. leadership helped birth the age of AI; now we must help the world harness this technology safely. (6/6)