Your LLM Is Only as Good as What It Retrieves
A Researcher's Perspective on Retrieval Quality in RAG Systems
Latest AI/ML News
42628 matching items
White House working group on AI – Statement from FLI’s Anthony Aguirre
"This move by the White House is also a recognition that Big Tech cannot self-regulate, and that the status quo is unacceptable."
Securing the Future of Trusted Semiconductor Supply Chains
Many countries view artificial intelligence (AI) as critical to economic competitiveness and national security. As a result, sovereign AI—the idea that national governments should develop, control, and govern AI in order to boost economic growth, guarantee security, and ensure strategic autonomy—has become a key strategic consideration in the global AI buildout. The post Securing the Future of Trusted Semiconductor Supply Chains appeared first on Center for Security and Emerging Technology .
Microsoft, Google and xAI will let the government test their AI models before launch
CSET’s Jessica Ji shared her expert perspective in an article published by CNN. The article examines new agreements between Microsoft, Google, and xAI to allow the U.S. government to evaluate unreleased AI models for cybersecurity and national security risks before launch. The post Microsoft, Google and xAI will let the government test their AI models before launch appeared first on Center for Security and Emerging Technology .
Government-Funded Research Seeds Entire Industries. What Would Be Lost Without It.
CSET’s Steph Batalis, Katherine Quinn, and Rebecca Gelles shared their expert analysis in an op-ed published by Barron's. Their piece examines the economic and scientific impact of proposed funding cuts to the National Institutes of Health (NIH), arguing that NIH-backed research plays a foundational role in driving medical innovation, biotechnology growth, and U.S. competitiveness. The post Government-Funded Research Seeds Entire Industries. What Would Be Lost Without It. appeared first on Center for Security and Emerging Technology .
Kubernetes v1.36: Declarative Validation Graduates to GA
In Kubernetes v1.36, Declarative Validation for Kubernetes native types has reached General Availability (GA). For users, this means more reliable, predictable, and better-documented APIs. By moving to a declarative model, the project also unlocks the future ability to publish validation rules via OpenAPI and integrate with ecosystem tools like Kubebuilder. For contributors and ecosystem developers, this replaces thousands of lines of handwritten validation code with a unified, maintainable framework. This post covers why this migration was necessary, how the declarative validation framework works, and what new capabilities come with this GA release. The Motivation: Escaping the "Handwritten" Technical Debt For years, the validation of Kubernetes native APIs relied almost entirely on handwritten Go code. If a field needed to be bounded by a minimum value, or if two fields needed to be mutually exclusive, developers had to write explicit Go functions to enforce those constraints. As the Kubernetes API surface expanded, this approach led to several systemic issues: Technical Debt: The project accumulated roughly 18,000 lines of boilerplate validation code. This code was difficult to maintain, error-prone, and required intense scrutiny during code reviews. Inconsistency: Without a centralized framework, validation rules were sometimes applied inconsistently across different resources. Opaque APIs: Handwritten validation logic was difficult to discover or analyze programmaticall…
DeepSeek-V3: A Large-Scale MoE Pretraining Benchmark for MLPerf Training v6.0
How MLCommons is bringing large-scale Mixture-of-Experts (MoE) pretraining to the MLPerf Training v6.0 suite. The post DeepSeek-V3: A Large-Scale MoE Pretraining Benchmark for MLPerf Training v6.0 appeared first on MLCommons .
Stop Sending IDE-Catchable AI Code Errors to Review
AI coding tools might have handed your developers a productivity gain, but they’ve created a problem for your code review process. Pull request volume is up significantly, and the code arriving for review carries error patterns that weren’t common before generative AI. Yet it’s the same people with the same working hours who are in […]
Codex is gaining steam
but I wish it had this
Last Week in AI #340 - OpenAI vs Musk + Microsoft, DeepSeek v4, Vision Banana
First week of Musk v. Altman, OpenAI ends Microsoft legal peril over its $50B Amazon deal, DeepSeek previews new AI model that ‘closes the gap’ with frontier models, and more!
MolmoAct 2: An open foundation for robots that work in the real world
MolmoAct 2 is a fully open robotics foundation model that brings faster, stronger 3D action reasoning to real-world robot tasks, alongside a major new bimanual manipulation dataset for researchers to study, reproduce, and build on.
Introducing Pinecone Marketplace: Getting to Production in Minutes
Stop answering the same questions. Turn docs into a "system of knowledge" with Marketplace. No-code RAG for support, legal, and onboarding with cited answers.
Join us at Big Technology's AI Summit in San Francisco on 6-18
OpenAI President Greg Brockman, Perplexity CEO Aravind Srinivas, Box CEO Aaron Levie, and more join us for a day of newsmaking conversations, live at San Francisco's Commonwealth Club.
A Permanent AI-Driven Underclass? Maybe, But Likely Not.
Don't buy the prophesies of economic doom. They're impossibly pessimistic about human nature.
Transforming skewed data to identify outliers
I have a number of right-skewed (sometimes highly skewed) datasets. It represents count data, where each point represents a single business with a single count for an activity. There is only one data point with one variable per business. For the purposes of fraud detection I am trying to identify which points are outliers. The purpose is to use the outliers as jumping-off points for investigation--i.e. if a count associated with a business is an outlier, we would like to investigate the activities of that business further. I am not interested in removing these outliers from the dataset. Currently I am using an adjusted boxplot ( Hubert & Vandervieran) with outliers outside [Q 1 - he -4(medcouple) IQR; Q 3 +he 3(medcouple) IQR], where h = 1.5 . However, I have thousands, sometimes hundreds of thousands, of points, and the above method takes a lot of memory and generates hundreds of outliers (i.e. hundreds of businesses to potentially investigate). Would it be statistically valid to instead perform some monotonic transformation on the skewed data (removing no points beforehand), and then use p-value to identify the outlying points? I do not know the underlying distribution of the dataset, so I would be trying different transformations until some test for normality turned up positive. EDIT: This question is different from the last one because the last one was closed for being too broad. I tried to focus this one more narrowly. I also attempted to clarify this post in order to a…
Empowering Carrot Ads with Domain Adaptive Learning
Authors: Trey Zhong, Xiyu Wang Contributors: Joseph Haraldson, Sharad Gupta, Sarah Lamacchia Introduction Carrot Ads is Instacart’s omnichannel retail media solution that allows retailer partners to build and scale their own advertising businesses on either their owned-and-operated (O&O) websites and apps or their whitelabel Storefront hosted by Instacart. Carrot Ads empowers retailers and CPG brands to accelerate revenue, while improving the customer experience, engagement and Ads return on investment. It features enterprise-grade infrastructure, AI-powered optimization, years of proprietary first-party data and flexibility to choose from retailer-sourced Ads demand, Instacart-sourced demand from 7,500+ CPG brands, or both. However, onboarding a new partner onto Carrot Ads introduces a key challenge: the ‘cold start’ problem, where limited historical interactions make it difficult to predict user behavior accurately. To serve performant ads, our systems rely on predicting a user’s Click-Through Rate (CTR) to generate a ranking score. On the Instacart Marketplace, we have billions of historical signals to train a model to do so. But when a partner launches a new ads experience on their O&O e-commerce site, there is often little to no interaction history for that property, so training an accurate model becomes challenging. User behavior can vary dramatically between websites — for example, browsing patterns on a grocery site differ from those on a pet supply or electronics si…
Kubernetes v1.36: Admission Policies That Can't Be Deleted
If you've ever tried to enforce a security policy across a fleet of Kubernetes clusters, you've probably run into a frustrating chicken-and-egg problem. Your admission policies are API objects, which means they don't exist until someone creates them, and they can be deleted by anyone with the right permissions. There's always a window during cluster bootstrap where your policies aren't active yet, and there's no way to prevent a privileged user from removing them. Kubernetes v1.36 introduces an alpha feature that addresses this: manifest-based admission control . It lets you define admission webhooks and CEL -based policies as files on disk, loaded by the API server at startup, before it serves any requests. The gap we're closing Most Kubernetes policy enforcement today works through the API. You create a ValidatingAdmissionPolicy or a webhook configuration as an API object, and the admission controller picks it up. This works well in steady state, but it has some fundamental limitations. During cluster bootstrap, there's a gap between when the API server starts serving requests and when your policies are created and active. If you're restoring from a backup or recovering from an etcd failure, that gap can be significant. There's also a self-protection problem. Admission webhooks and policies can't intercept operations on their own configuration resources. Kubernetes skips invoking webhooks on types like ValidatingWebhookConfiguration to avoid circular dependencies. That mea…
Meet the Finalists: JetBrains x Codex Hackathon
Put a capable coding model inside a developer’s primary workspace, and the IDE stops being a place where you write code. It becomes a place where you direct an agent, watch how it reasons, manage what it pays attention to, and decide when its output is worth shipping. That was the defining theme of the […]
The distillation panic
‘Distillation attacks’ is a horrible term for what is happening right now.
Import AI 455: AI systems are about to start building themselves.
The first step towards recursive self improvement
The AI Progress Chart Everyone Is Misreading — Beth Barnes & David Rein
Beth Barnes and David Rein on the one graph that ate the AI timelines discourse, and why the two people who built it are the most careful about how you read it. **SPONSOR** Prolific - Quality data. From real people. For faster breakthroughs. https://www.prolific.com/?utm_source=mlst Interview: https://youtu.be/cnxZZTl1tkk --- Beth Barnes and David Rein from METR on the one graph that ate the AI timelines discourse, and why the people who built it are the most careful about how it gets read. Beth founded METR after leaving OpenAI alignment. David is first author on GPQA and co-author on HCAST and the METR Time Horizons paper. Together they built the measurement Daniel Kokotajlo called the single most important piece of evidence on AI timelines: the log-linear line of "how long a task a frontier model can complete at 50% reliability" vs release date. The conversation opens on reward hacking. Current models can articulate in chat why a behaviour is undesired and then execute it anyway as agents. From there: construct validity, Melanie Mitchell's four-problem taxonomy, and the ARC-AGI 1-to-2 collapse as a worked example of adversarially-selected benchmarks regressing once labs target them. Beth's counter: METR deliberately does not adversarially select. David's: models do not have to do the right thing for the right reasons. Methodology, then specification — David's compiler analogy, Beth on four-month tasks as expensive to evaluate rather than unspecifiable. Then the SWE-bench…
How do we move the CRASA Summit’s commitment to collaborative regulation from intent to evidence-led action?
The G5 Regulatory Collaboration Summit 2026, hosted by the Communications Regulators’ Association of Southern Africa (CRASA) and the Malawi Communications Regulatory Authority (MACRA) at the Bingu International Convention Center in […] The post How do we move the CRASA Summit’s commitment to collaborative regulation from intent to evidence-led action? appeared first on Research ICT Africa .
Better Models Won’t Save Your Agent
Most agent failures are data failures, not model failures. Pinecone Nexus is a Knowledge Engine that compiles enterprise data into structured artifacts agents can query in one step.
LWiAI Podcast #243 - GPT 5.5, DeepSeek V4, AI safety sabotage
Our 243rd episode with a summary and discussion of last week’s big AI news!
Pinecone Nexus: The Knowledge Engine for Agents
Pinecone Nexus is a knowledge engine for the agentic AI era, moving reasoning from retrieval to compilation — with KnowQL as the standard query language for agents.
⚡Vector Search at 10B Scale, 📊 Lance Format Benchmarks, 🚗 AV Pipelines at Scale
Distributed vector search at 10B scale, more efficient storage with Lance format v2.2, and production AV pipelines simplified, plus upcoming events and community updates.
Lessons on UX, security, and scale when building an enterprise-grade Slack agent
We built a Deep Search Slack agent for large companies. Here is what we learned about user experience, enterprise security, and Redis-backed rate limiting.
Lessons on UX, security, and scale when building an enterprise-grade Slack agent
We built a Deep Search Slack agent for large companies. Here is what we learned about user experience, enterprise security, and Redis-backed rate limiting.
Boosting multimodal inference performance by >10% with a single Python dictionary
If we've said it once, we've said it once per millisecond: never block the GPU.
AI Weekly Issue #489: PE built AI's new distribution layer
OpenAI took $10B from a 19-firm Wall Street consortium. Anthropic is closing $1.5B from Blackstone, Goldman, and Hellman & Friedman. Same rooms, different portfolios. The week AI's go-to-market stopped being SaaS and started being private equity.
How to Work and Compound with AI
Context as infra, taste as config, verification for autonomy, scale via delegation, closing the loop.
US military reaches deals with 7 tech companies to use their AI on classified systems
CSET’s Helen Toner shared her expert insight in an article published by the Associated Press. The article examines the Pentagon’s agreements with seven major tech companies to integrate artificial intelligence (AI) into classified military systems, expanding AI use in decision-making, logistics, and battlefield operations. The post US military reaches deals with 7 tech companies to use their AI on classified systems appeared first on Center for Security and Emerging Technology .
DOD expands its classified AI work with 8 companies — excluding Anthropic — amid ongoing dispute
CSET’s Lauren Kahn shared her expert insight in an article published by DefenseScoop. The article explores the Pentagon’s growing efforts to integrate advanced artificial intelligence (AI) capabilities into classified military operations and the broader implications of expanding AI adoption across the Department of Defense. The post DOD expands its classified AI work with 8 companies — excluding Anthropic — amid ongoing dispute appeared first on Center for Security and Emerging Technology .
Space Is Critical Infrastructure—It Needs an Alliance To Guard It
CSET’s Kathleen Curlee and former U.S. Air Force pilot Brian Golden shared their expert analysis in an op-ed published by Newsweek. The article discusses the growing importance of space infrastructure to modern life and argues that increased international coordination is needed to ensure the security and stability of the space domain. The post Space Is Critical Infrastructure—It Needs an Alliance To Guard It appeared first on Center for Security and Emerging Technology .
Are AI’s Consumer Applications Hitting a Wall?
New data shows consumer AI app growth has flatlined as generative AI struggles tofind its true form outside of the enterprise.
Kubernetes v1.36: Pod-Level Resource Managers (Alpha)
Kubernetes v1.36 introduces Pod-Level Resource Managers as an alpha feature, bringing a more flexible and powerful resource management model to performance-sensitive workloads. This enhancement extends the kubelet's Topology, CPU, and Memory Managers to support pod-level resource specifications ( .spec.resources ), evolving them from a strictly per-container allocation model to a pod-centric one. Why do we need pod-level resource managers? When running performance-critical workloads such as machine learning (ML) training, high-frequency trading applications, or low-latency databases, you often need exclusive, NUMA-aligned resources for your primary application containers to ensure predictable performance. However, modern Kubernetes pods rarely consist of just one container. They frequently include sidecar containers for logging, monitoring, service meshes, or data ingestion. Before this feature, this created a trade-off, to get NUMA-aligned, exclusive resources for your main application, you had to allocate exclusive, integer-based CPU resources to every container in the pod. This might be wasteful for lightweight sidecars. If you didn't do this, you forfeited the pod's Guaranteed Quality of Service (QoS) class entirely, losing the performance benefits. Introducing pod-level resource managers Enabling pod-level resources support for the resource managers (via the PodLevelResourceManagers and PodLevelResources feature gates) allows the kubelet to create hybrid resource alloca…
Building a Natural Language Interface to the Spotify Ads API with Claude Code Plugins
Turning OpenAPI spec and Markdown files into a conversational ads management tool — no compiled code required. The post Building a Natural Language Interface to the Spotify Ads API with Claude Code Plugins appeared first on Spotify Engineering .
U.S. Power and Principle
James Story, former U.S. ambassador to Venezuela, visits Carnegie Council to discuss the new dynamic between American power and principle.
Stay informed: Get Access Now updates
Here's where you can subscribe to the Access Now Express newsletter and action alerts. The post Stay informed: Get Access Now updates appeared first on Access Now .
What’s next for Ai2: A conversation with Interim CEO Peter Clark
Interim CEO Peter Clark shares his thoughts on this moment for Ai2, our commitment to open science, and where the institute is headed next.
How to Engineer AI Inference Systems with Philip Kiely - #766
In this episode, Philip Kiely, head of AI education at Baseten, joins us to unpack the fast-evolving discipline of inference engineering. We explore why inference has become the stickiest and most critical workload in AI, how it blends GPU programming, applied research, and large-scale distributed systems, and where the line sits between inference and model serving. Philip shares how research-to-production can move in hours, not months, and why understanding “the knobs” of inference—batching, quantization, speculation, and KV cache reuse—lets teams design better products and SLAs. We trace the inference maturity journey from closed APIs to dedicated deployments and in-house platforms, discuss GPU lifecycles, and survey today’s runtime landscape, including vLLM, SGLang, and TensorRT LLM. Finally, we look ahead to agents and multimodality, making the case for specialized, workload-specific runtimes when performance and efficiency matter most. The complete show notes for this episode can be found at https://twimlai.com/go/766.
CSET Senior Fellow Andrew Lohn Testifies Before U.S.-China Economic and Security Review Commission
Washington, D.C. (April 30, 2026) — This morning, Andrew Lohn, Senior Fellow at Georgetown University’s Center for Security and Emerging Technology (CSET), testified before the U.S.-China Economic and Security Review Commission. The post CSET Senior Fellow Andrew Lohn Testifies Before U.S.-China Economic and Security Review Commission appeared first on Center for Security and Emerging Technology .
Building gets easier
My tool stack is changing
Agents can now create Cloudflare accounts, buy domains, and deploy
Starting today, agents can now be Cloudflare customers. They can create a Cloudflare account, start a paid subscription, register a domain, and get back an API token to deploy code right away. Humans can be in the loop to grant permission, but there’s no need to go to the dashboard, copy and paste API tokens, or enter credit card details.
How to actually protect against digital sexualized violence
AlgorithmWatch has put forward recommendations on how to implement a ban of deepfakes in the AI Act as part of the AI Omnibus procedure. To effectively protect victims of digital sexualized violence, AI companies, platforms, and perpetrators must consistently be held accountable.
AstaBench update: New results, plus adoption from industry
AstaBench’s latest update adds new frontier-model results, including GPT-5.5, and highlights growing adoption from groups including the UK AISI, General Reasoning, Elicit, SciSpace, Distyl AI, and EvoScientist.
LWiAI Podcast #242 - ChatGPT Images 2.0, Qwen 3.6 Max, Kimi-K2.6
ChatGPT’s new Images 2.0 model is surprisingly good at generating text , Alibaba Drops Qwen 3.6 Max Preview , SpaceX is working with Cursor