481
Measuring political bias in Claude
We introduce and open-source an automated evaluation for political even-handedness, testing six models across 1,350 prompt pairs.
482
Meta announces LlamaCon, its first generative AI dev conference on April 29
We’re here with a couple save the dates to reserve some space on your calendars.
483
Critical vulnerability in AI coding platform Base44 allowing unauthorized access
New discovery underscores security implications of AI-powered development and the rise of Vibe Coding Platforms.
484
Alignment is not free: How model upgrades can silence your confidence signals
Delegate investigative work to AI agents, from L1 to L3 reviews, with consistency, auditability and traceability.
485
QVQ-Max: Think with Evidence
QWEN CHAT GITHUB HUGGING FACE MODELSCOPE DISCORD Introduction Last December, we launched QVQ-72B-Preview as an exploratory model, but it had many issues. Today, we are officially releasing the first version of QVQ-Max, our visual reasoning model. This model can not only “understand” the content in images and videos…
486
A stateful browser agent using self-healing DOM maps
Most browser agents forget what they’ve done. This one remembers, adapts, and learns, redefining how AI agents get work done.
487
LLM-D: Kubernetes-Native Distributed Inference
Introducing llm-d: Kubernetes-native distributed LLM inference with KV-cache routing, disaggregated serving, and SOTA performance per dollar. Built on vLLM.
488
Context Engineering for AI Agents: Lessons
This post shares the local optima Manus arrived at through our own "SGD". If you're building your own AI agent, we hope these principles help you converge faster.
489
MongoDB acquires Voyage AI
MongoDB to integrate Voyage AI's industry-leading embedding and reranking models, delivering highly accurate and relevant information retrieval to power sophisticated AI use cases NEW YORK , Feb. 24, 2025 /PRNewswire/ -- MongoDB, Inc. (NASDAQ: MDB), the leading database for modern applications,
490
Flock Now Using AI to Report to Police If Our Movement Patterns Are "Suspicious"
Company crosses a dangerous line by beginning to offer AI suspicion-generation functions
491
ClickHouse acquires LibreChat, open-source AI chat platform
We are excited to announce that ClickHouse has acquired LibreChat, the leading open-source AI chat platform.
492
Workhorse LLMs: Why Open Source Models Dominate Closed Source for Batch Tasks
Open source LLMs now outperform closed alternatives at 90% lower cost for workhorse tasks. We compare benchmarks, pricing, and real-world performance.
493
Qwen2.5-Max: Exploring the intelligence of large-scale MoE model
QWEN CHAT API DEMO DISCORD It is widely recognized that continuously scaling both data size and model size can lead to significant improvements in model intelligence. However, the research and industry community has limited experience in effectively scaling extremely large models, whether they are dense or…
494
Microsoft's Majorana 1 chip carves new path for quantum computing
Majorana 1, the first quantum chip powered by a new Topological Core architecture
495
Anthropic invests $50B in US AI infrastructure
Anthropic is partnering with Fluidstack to build data centers in Texas and New York, creating roughly 800 permanent and 2,400 construction jobs.
496
The Biology of a Large Language Model
We investigate the internal mechanisms used by Claude 3.5 Haiku — Anthropic's lightweight production model — in a variety of contexts, using our circuit tracing methodology.
497
MCP on AWS Lambda with MCPEngine
A virtual feature store for the modern AI stack
498
What “working” means in the era of AI apps
New revenue benchmarks for gen AI startups in year one, based on dozens of enterprise and consumer AI companies we’ve spent significant time with.
499
A Complete Guide to Meta Prompting
Check out our deep dive on the latest meta prompting methods, like DSPy and TEXTGRAD, and the best prompt generator tools out there.
500
The AI bubble today is bigger than the IT bubble in the 1990s
The difference between the IT bubble in the 1990s and the AI bubble today is that the top 10 companies...
501
NVIDIA DGX Spark In-Depth Review: A New Standard for Local AI Inference
Thanks to NVIDIA’s early access program, we are thrilled to get our hands on the NVIDIA DGX™ Spark. It’s quite an unconventional system, as NVIDIA rarely releases compact, all-in-one machines that bring supercomputing-class performance to a desktop workstation form factor.
502
503
Windows ML is generally available
The future of AI is hybrid, utilizing the respective strengths of cloud and client while harnessing every Windows device to achieve more. At Microsoft, we are reimagining what’s possible by bringing powerful AI compute directly to Windows devices,
504
The End of Moore's Law for AI? Gemini Flash Offers a Warning
Google’s Gemini 2.5 Flash price increase offers a warning about the economics of AI inference and the limits of ever-cheaper intelligence.
505
Vector database that can index 1B vectors in 48M
Discover why we built Vectroid, a serverless vector database that delivers exceptional accuracy and low latency without compromising on cost. Learn about our approach to solving the traditional tradeoffs between speed, accuracy, and cost in vector search.
506
Gmail can read your emails and attachments to train its AI, unless you opt out
Did you know that Gmail can use your emails and attachments for its smart features? Here's how to check your settings.
507
Noise cancellation improves turn-taking for AI Voice Agents
Turn-Taking is a big challenge AI Voice Agents are rapidly evolving, powering critical use-cases such as customer support automation, virtual assistants, gaming, and remote collaboration platforms. For these voice-driven interactions to feel natural and practical, the underlying audio pipeline must be resilient to…
508
GPT-5.1 for Developers
GPT-5.1 is now available in the API, bringing faster adaptive reasoning, extended prompt caching, improved coding performance, and new apply_patch and shell tools.
509
Robot hand could harvest blackberries better than humans
The soft robotic gripper developed at the U of A received a U.S. patent.
510
Voyage-code-3
TL;DR – Introducing voyage-code-3, our next-generation embedding model optimized for code retrieval. It outperforms OpenAI-v3-large and CodeSage-large by an average of 13.80% and 16.81% on a suite …