Many decades ago, U.S. Air Force Colonel John Boyd introduced the concept of the “OODA loop,” for Observe, Orient, Decide, and Act. These are the four steps of real-time continuous decision-making. Boyd developed it for fighter pilots, but it’s long been applied in artificial intelligence (AI) and robotics. An AI agent, like a pilot, executes the loop over and over, accomplishing its goals iteratively within an ever-changing environment. This is Anthropic’s definition: “Agents are models using tools in a loop.”...

In recent years, a movement has been steadily growing across the global tech industry to address a ubiquitous and insidious type of bugs known as memory-safety vulnerabilities. A computer’s memory is a shared resource among all programs, and memory safety issues crop up when software can pull data that should be off limits from a computer’s memory or manipulate data in memory that shouldn’t be accessible to the program. When developers—­even experienced and security-conscious developers—­write software in ubiquitous, historic programming languages, like C and C++, it’s easy to make mistakes that lead to memory safety vulnerabilities. That’s why proactive tools like ...

Web 3.0—the distributed, decentralized Web of tomorrow—is finally poised to change the Internet’s dynamic by returning ownership to data creators. This will change many things about what’s often described as the “CIA triad” of ...

AIOps refers to the use of LLM-based agents to gather and analyze application telemetry, including system logs, performance metrics, traces, and alerts, to detect problems and then suggest or carry out corrective actions. The likes of Cisco have deployed AIops in a conversational interface that admins can use to prompt for information about system performance. Some AIOps tools can respond to such queries by automatically implementing fixes, or suggesting scripts that can address issues...

This is an integrity failure. Specifically, it’s a failure of processing integrity. And while we can think of particular patches that alleviate this exact failure, the larger problem is much harder to solve.

Davi Ottenheimer ...

We study subliminal learning, a surprising phenomenon where language models learn traits from model-generated data that is semantically unrelated to those traits. For example, a “student” model learns to prefer owls when trained on sequences of numbers generated by a “teacher” model that prefers owls. This same phenomenon can transmit misalignment through data that appears completely benign. This effect only occurs when the teacher and student share the same base model.

Interesting security implications.

I am more convinced than ever that we need serious research into ...