Live intelligence

Latest AI updates

127 updates
Research

EngineAgent: Agentizing On-Device LLM Inference Engines with Inference Stack Hierarchy Awareness and Hardware Abilities

EngineAgent is an agentic system that automates enabling new LLM models in on-device inference engines by separating reusable engine and hardware knowledge from model-specific implementation. It builds a hierarchy-aware engine index, pre-validates backend workflows, and uses a builder-auditor loop to coordinate cross-layer changes and device-side debugging. Across 12 tasks covering three mobile LLMs and four engine-backend pairs on CPUs, GPUs, and NPUs, it achieved 75% end-to-end success versus 33.3% for the strongest coding-agent baseline, while reducing average per-task time and model-API cost.

on-device LLM inferenceAI agentsmobile hardware
Products

ChatGPT introduces verified AI Business Agents for e-commerce

OpenAI is adding verified "Business Agents" to ChatGPT, letting shoppers chat with branded agents that represent a company. The feature is launching in beta first for eligible e-commerce and online service businesses in the US. Business Agents appear when a participating business shows up in a ChatGPT response, and users can ask about offerings or compare products. Businesses can configure their agent's behavior.

OpenAIChatGPTAI agents
Models

Meta Releases Muse Ember, a 1M-Context Model for High-Volume Agent Work

Meta Superintelligence Labs released Muse Ember, a hosted multimodal reasoning model aimed at high-volume agent subtasks such as tool calls, routing, extraction and long document parsing. It offers a 1M-token context window with up to 512K output tokens and accepts text, images, audio and video, but weights are closed and it is available only through the Meta Model API and Muse Code. Meta reports a score of 36 on the Artificial Analysis Intelligence Index, tying Claude Haiku 5.5, at about $0.136 per task versus $0.411 for Haiku, and 37.8 on AutomationBench-AA. Meta's reported results also show weaker performance on FrontierCode 2.0-AA at 27.2 versus Haiku's 34.9; benchmarks are self-reported.

MetaAI modelsAI agents
Research

Framework Evaluates AI Agents on Industrial Chip Floorplanning

The paper presents an evaluation framework for AI agents doing chip floorplanning that connects submitted layouts to a prescribed downstream implementation flow, checks geometric validity, measures physical quality and reports resource use. On one industrial design whose 99 macros occupy roughly 71% of a fixed L-shaped region, the authors compare their specialized Nova-SI system with general-purpose engineering agents, obtaining complete physical measurements for 39 layout outputs. Nova-SI reaches a mean of 57.97 on a 0-100 composite quality score without online feedback from the power-grid and placement evaluation, versus 31.17 for the strongest measured general-purpose configuration. Layout-level analysis shows that similar global utilization can conceal substantially different local congestion.

AI agentschip designEDA
Research

Agency: An Interposition Layer Cuts Agent Harness Costs by Up to 35%

The paper analyzes cost accumulation patterns across agent traces and argues that harness-side choices, while having minimal impact on task accuracy, strongly affect execution cost, with prior work reporting reductions up to 2.6x on the same model and task. It introduces Agency, a harness interposition platform offering a unified abstraction for portable optimizations on off-the-shelf harnesses such as Claude Code and OpenAI Codex. Proof-of-concept techniques built on this analysis reduce cost by up to 35% across real-world benchmarks without compromising accuracy or modifying the model.

AI agentsLLM inference costagent harnesses
Products

DeepSeek's Harness agent framework reaches 1.54 million daily active users

DeepSeek said its Harness AI agent framework reached 1.54 million daily active users less than three months after launch. The rapid adoption points to AI agents as a growing competition area between leading Chinese and American AI companies.

DeepSeekAI agentsHarness
Research

Nudging Decision Models: Do Digital Nudges Steer Fast, Generation-Free AI Agents?

This arXiv study adapts an online shopping experiment to three non-generative decision models (Jev, Decision-1, and d1), with 600 runs per model. It reports that the models chose a preselected default in 97.5–99.9% of cases and a product with social influence information in 75–96%, versus 47–49% in the control. The models responded more strongly to automatic nudges than reflective ones, contrary to reported behavior for reasoning-capable LLM agents. The authors present this as first evidence that non-generative decision models are susceptible to digital nudges in a System 1 pattern.

AI agentsdigital nudgesdecision models
Industry

Microsoft details AI agent system for vulnerability patching

Microsoft reported closing 1,413 vulnerabilities in October 2026, attributing the surge to MDASH, an agentic scanning system using over 100 AI agents to analyze code and classify risks. The system represents a concrete application of AI agents in cybersecurity, potentially improving patching efficiency.

MicrosoftAI agentscybersecurity
Tutorial

Beyond code generation: A developer’s guide to agentic engineering

GitHub's blog post outlines a workflow for agentic engineering, covering planning, building, verifying, and observing software development with AI agents. It highlights features like Copilot in Slack/Teams, session continuity, cloud sandboxes, stacked pull requests, and AI-assisted code review. The post aims to help developers manage increased code volume and review bottlenecks.

GitHub CopilotAI agentssoftware development workflow
Research

PCBRouteAgent: LLM Agent for Automated PCB Routing

PCBRouteAgent is a multimodal PCB routing agent that combines a Global Routing Planner with a Detailed Routing Executor to automate printed circuit board routing. It uses board images and structured electrical/geometric data to guide KiCad's native routing engine, with bounded local recovery and structured feedback for revisions. Evaluated on 247 boards across four PCBWorld-Bench splits, it achieves clean-pass rates of 100%, 94.9%, 100%, and 40%, outperforming baselines on real-world subsets by 10-36 percentage points.

AI agentsprinted circuit board designmultimodal reasoning
Research

ReproBridge: Agent Workflow Boosts ML Reproduction Success to 92.6%

ReproBridge introduces a standardized artifact design and agent workflow to improve reproducibility of ML experiments. On a new benchmark of 1,029 tasks from 36 papers, refined artifacts raised mean reproduction success from 76.4% to 92.6% across four models and cut API costs by 19.7%. The approach narrows performance gaps between models, aiding lower-performing agents.

machine learning reproducibilityAI agentsbenchmarking
Research

XStack: A Benchmark for Migrating ML Repositories Between PyTorch and JAX

XStack is a new benchmark that evaluates AI agents on migrating entire machine learning repositories from PyTorch on NVIDIA GPUs to JAX on Google TPUs. The benchmark requires agents to deliver a working JAX repository that meets grading criteria based on the source repository's published results. Current agent performance is low, with the best agent (Claude Opus 5) achieving only 75% pass rate, highlighting the difficulty of cross-framework and cross-hardware migration.

XStackML repository migrationPyTorch to JAX
Products Editor selected

Airbnb announces AI travel agent launching in first half of 2027

Airbnb has announced it will launch an AI travel agent in the first half of 2027. The agent will assist travelers with trip planning, comparing options, managing bookings, and providing support throughout their stay. This marks a concrete product roadmap for AI integration in the travel industry.

Why it matters: Gives readers a specific timeline and feature set for a major platform's upcoming AI agent, useful for tracking AI adoption in travel.

AI agentstravel technology
Products

Perplexity launches its own take on AI agents for work

Perplexity has introduced 'Teammates,' always-on AI agents designed for workplace collaboration. The agents can take on multiple roles, pull from existing Perplexity projects and memory, and integrate with Perplexity Computer to complete tasks on-device. Teammates can work simultaneously on different jobs, positioning Perplexity as a competitor to OpenAI's Dots and Anthropic's Claude Duet.

PerplexityAI agentsenterprise productivity
Products Editor selected

AWS launches Web Answer on Amazon Bedrock AgentCore for cited, agentic web answers

AWS announced general availability of Web Answer on Amazon Bedrock AgentCore, a managed tool that returns synthesized, source-cited answers from web sources with a single tool call. It replaces the custom search-retrieve-synthesize-citation pipeline, using agentic search that plans queries, reads pages, and verifies facts before returning a grounded answer. The tool is available in us-east-1 at $10 per 1,000 calls and integrates via MCP with any AgentCore Gateway. Enterprise controls include IAM/JWT authentication, VPC PrivateLink, CloudWatch metrics, and domain filters.

Why it matters: Developers building AI agents can now eliminate complex web retrieval pipelines and get cited, current answers with one API call, reducing maintenance and hallucination risks.

Amazon BedrockAI agentsweb search
Research

No AI agent scores above 12 percent on SemiAnalysis' new real-world work benchmark

SemiAnalysis has introduced a new benchmark called GDPx that evaluates AI agents on simulated knowledge work tasks, spanning domains like chip design and quant trading. According to the benchmark, no model achieves more than 12 percent, with Claude Opus 5.5 leading slightly ahead of Gemini 4 Carbon. Critics note the benchmark has a small number of tasks per domain and uneven testing conditions.

AI benchmarkingAI agentsSemiAnalysis
Research

Corralling AI Risks with Grassroots Contracts

The paper proposes grassroots contracts, a digital adaptation of Proudhon's mutualism, to address three major AI risks on digital platforms: AI control of global platforms, unaccountable AI agents, and unattributed deepfakes. It presents implementations for a social graph with accountable AI agents and a social network with digital provenance, along with working smartphone apps. The approach aims to restore accountability and provenance through decentralized, user-driven agreements.

AI governancedigital platformsAI agents
Research Editor selected

Process-based catchment model matches deep-learning accuracy across three continents

A new process-based hydrological model, CaRT+, built on physical laws and refined with AI assistance, achieves accuracy comparable to LSTM neural networks on catchments in Britain, the US, and Australia. With 35 parameters per catchment and no daily neural network, it reaches median Nash-Sutcliffe efficiencies of 0.863 and 0.765 in held-out years on British and US data, and outperforms LSTMs on Australian catchments. The results suggest that catchment-scale laws exist as laws of form, with shared constitutive families and only parameters learned.

Why it matters: This is a concrete, recent research result showing a process-based model matching deep learning accuracy, with implications for hydrology and AI-assisted model refinement.

hydrological modelingdeep learning comparisonphysical laws