Tag
FT reports that AI compute has become strategic infrastructure, with adversaries targeting data centres like power plants. Iranian strikes on AWS Bahrain zones and Ukrainian drone attacks on Yandex's AI supercomputer sites exposed the fragility of cloud and data-localisation resilience.
The quote from OpenAI's Agent Security highlights the surprising jumps in AI capabilities and stresses the importance for organizations to develop security postures and incident response plans to handle such surprises.
This article discusses handling provider outages in AI agent systems by implementing failover, model fallback, circuit breakers, and idempotent retries to prevent pipeline stoppages and duplicate actions.
The article explores whether agent-run companies could achieve exponential growth by leveraging AI agents for tasks like product development and customer management, while discussing experiments on model failure overlaps and the challenges of building resilient systems to scale effectively.
PaymentKit is a billing solution designed to withstand processor shutdowns, ensuring continuous service.
This article explains how Kubernetes probes, including startup, readiness, and liveness probes, work to improve application resilience, with interactive demos demonstrating their functionality.
The article introduces two official NuGet packages, Microsoft.Extensions.Resilience and Microsoft.Extensions.Http.Resilience, for building resilient .NET applications with built-in strategies like retries, timeouts, and circuit breakers.
Cursor introduces builds, pre-prepared copies of development environments that let cloud agents start up to 3x faster by avoiding just-in-time setup, with improved resilience and observability for debugging.
SpaceX's Falcon 9 successfully deployed Firefly Aerospace's Blue Ghost and ispace's RESILIENCE lunar landers on a trajectory to the Moon, with discussion of second-stage disposal and future sustainability via Starship.
A reflection on Wojciech Zaremba's Berkeley talk comparing AI safety to fire resilience, arguing that safety requires a multi-layered ecosystem rather than a single control.
Cybersecurity has evolved from a technical compliance hurdle to a core business imperative, with the CISO role now acting as a strategic business leader and relationship manager in the face of rising AI-related threats and regulatory complexity.
The article draws an analogy between the coevolution of fire-bellied toads with chytrid fungus and the need to understand rather than eliminate unexpected AI behaviors, arguing that safety depends on understanding the conditions that produce those behaviors.
Startup failures are rising in 2025; the article emphasizes that execution discipline matters more than hype, citing notable failures like Builder.ai, Canoo, Lilium, and Rain AI.
This paper proposes CPSAINT, a seven-layer integrity decomposition for agentic AI systems, paired with FRIESA-K, a residual-risk functional based on absorbing Markov models. It bridges structural failure path analysis with quantified risk estimates, demonstrating the framework on warehouse robotics and financial-services agent scenarios.
Ben Horowitz shares advice for startup CEOs facing media criticism and internal stress, emphasizing the need to believe there is always an answer.
ResilixForge is an Apache-2.0 licensed async resilience toolkit for Python, providing retries, circuit breakers, bulkheads, and rate limiters.
Discussing the risks of relying on a single US frontier AI model and advocating for a diversified stack using Chinese alternatives like DeepSeek and Qwen, which can be run locally and are cheaper, for better resilience and data privacy.
This paper argues for localized fault recovery mechanisms in distributed systems, emphasizing the benefits of fixing failures at the component level rather than relying on global recovery.
Sam Altman shares Paul Graham's advice on being 'relentlessly resourceful,' emphasizing the importance of persistently trying different approaches to solve problems, a lesson many founders overlook.
This paper presents a skill-constrained model predictive control approach for resilient manufacturing supply chains, where training decisions affect future certified capacity. The controller solves a finite-horizon mixed-integer program and is evaluated on synthetic scenarios, showing that predictive control helps when bottlenecks are forecastable but is not universally superior.