Building Safer and Interoperable AI Systems
AI agents are now contemplated that can interact with each other and with information sources on the Internet. What could go wrong?
Building Safer and Interoperable AI Systems
AI agents are now contemplated that can interact with each other and with information sources on the Internet. What could go wrong?
Fit for People, Fit for Purpose: Designing Tech that Matters
My computing degree enabled me to create better-fitting technologies for people and what is important to them.
Strong community support and ongoing innovation keep the Go programming language at the forefront of cloud development.
LLMs could give robots the ability to process open-ended, natural language commands to solve complex tasks.
The Evolution of Computer Science at the University Level
Interdisciplinary approaches that blend computer science with other fields of study have been increasing at universities and in popularity with students.
Brain Implants Give People Back What They Lost
Researchers aim to have paralyzed patients use thoughts to control a prosthetic hand while physical sensations from the hand provide feedback.
Investigating Research Software Engineering: Toward RSE Research
It is necessary to build better domain-specific tooling to address the domain-specific challenges of research software and to establish RSE Research as a research field over RSE.
Artificial Intelligence Then and Now
Branding large language models as artificial intelligence primes customers to believe they have capabilities they lack.
It simply is not appropriate to write code that will be connected to the Internet in an unsafe language such as C.
Life Lessons from the First Half-Century of My Career
Reflection captures 16 people-focused and career-focused life lessons.
It Is Time to Standardize Principles and Practices for Software Memory Safety
Memory-safety standardization is an essential step to promoting universal strong memory safety in government and industry, and to ensure access to more secure software for all.
Questioning the Criteria for Evaluating Non-Cryptographic Hash Functions
There seems to be a gap in how cryptographic and non-cryptographic hash functions are designed.
Meta’s Hyperscale Infrastructure: Overview and Insights
A concise, high-level overview of Meta’s hyperscale infrastructure, with key insights from its development.
Initiatives created for the 2023 UbiComp/ISWC conference illustrate what can and should be done to support the needs of a diverse, increasingly global computing community.
Program Correctness through Self-Certification
Self-certification is simpler to apply than formal verification, and fully validates only the program runs that actually occur.
Technical Perspective: Shortening the Path to Designing Efficient Graph Algorithms
A new way of designing efficient graph algorithms that will go significantly beyond shortest paths.
Negative-Weight Single-Source Shortest Paths in Near-Linear Time
The first combinatorial algorithm for negative-weight SSSP to break through the classic ˜O(m√n log W) bound.
Technical Perspective: Unsafe Code Still a Hurdle Copilot Must Clear
Examining the conditions under which LLMs may generate insecure code.
Asleep at the Keyboard? Assessing the Security of GitHub Copilot’s Code Contributions
Our work attempts to characterize the tendency of Copilot to produce insecure code.