Live intelligence

Latest AI updates

93 updates

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.

the-decoder1h agoScore 70

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.

arxiv13h agoScore 72

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.

arxiv13h agoScore 85

hydrological modelingdeep learning comparisonphysical laws

Research Editor selected

SumOne: Task-Agnostic Summarizer for Agent Memory Compression

SumOne is a standalone summarizer that compresses long-horizon agent histories via masked reconstruction, using a frozen reconstructor and GRPO to optimize without task-specific rewards. It reduces inference cost by ~78% on DeepSearchQA and ~75% on SWE-bench Verified with minimal performance loss, and transfers across models and harnesses. The method enables plug-and-play context compression for agents.

Why it matters: Provides a concrete, novel AI development with measurable efficiency gains and cross-task transfer, valuable for agent deployment.

arxiv13h agoScore 82

AI agentscontext compactionself-supervised learning

Industry

Uber discloses AI agent ride and delivery volumes, showing minimal early adoption

Uber executives revealed that only four rides were arranged by third-party AI agents in Q3 2026, with about 500 Uber Eats orders fulfilled by AI agents during Halloween week, out of 320 million total rides and deliveries. This is the first time Uber has disclosed such figures, highlighting the nascent state of AI agent integration in ride-hailing and delivery. The low numbers contrast with predictions of AI agents like Meta's Muse or OpenAI's Dots disrupting the platform.

The Verge AI1d agoScore 72

AI agentsUber platformdisintermediation

Industry

The Security Companies Trying to Get AI Agents Under Control

The article highlights the emerging challenge of securing AI agents, as these autonomous systems become more prevalent. It notes that security companies are developing new tools and strategies to manage and control AI agents, addressing risks like unauthorized actions and data exposure. The piece also mentions that OpenAI may release new models as soon as Friday, adding urgency to the security discussion.

AIHUB1d agoScore 78

AI securityAI agentsOpenAI models

Products Editor selected

openJiuwen Agent-Memory: First Audio-Visual Dual-Stream Memory for Agents

openJiuwen introduces Agent-Memory, a system that gives AI agents a dual-stream memory for both audio and visual content, enabling them to understand long videos and recall key events accurately. The system reportedly boosts long-video memory precision and cuts question-answering latency by up to 97%. This marks a significant step in agent memory capabilities for multimodal understanding.

Why it matters: Provides a concrete, recent AI development in agent memory for long-video understanding, with measurable performance gains.

qbitai1d agoScore 82

AI agentsmultimodal learningvideo understanding

Products Editor selected

Agentic cheminformatics with Google Antigravity

Google DeepMind introduces agentic cheminformatics capabilities in Google Antigravity, designed to accelerate drug discovery workflows. The system can search compounds and propose novel rigidified macrocycles using specialized Cheminformatics Skills. This integration aims to streamline research tasks in medicinal chemistry.

Why it matters: Provides concrete details on a new AI-driven tool for drug discovery, offering practical value for researchers and industry professionals.

Google DeepMind2d agoScore 85

Google Antigravitycheminformaticsdrug discovery

Tutorial Editor selected

ServiceNow on-behalf-of token exchange with Bedrock AgentCore Gateway

This post demonstrates how to configure Amazon Bedrock AgentCore Gateway for on-behalf-of (OBO) token exchange with Microsoft Entra ID, enabling AI agents to call ServiceNow REST APIs while preserving the user's identity. It covers setting up Entra application registrations, ServiceNow inbound OAuth integration, and AgentCore Gateway configuration with native OBO support. The pattern eliminates custom Lambda interceptors and applies to other OAuth-protected REST APIs.

Why it matters: Provides a concrete, recent technical guide for implementing identity propagation in AI agents, a critical enterprise requirement, with step-by-step configuration details.

AWS Machine Learning Blog2d agoScore 85

Amazon Bedrock AgentCoreOAuth 2.0AI agents

Products Editor selected

Same Model, Different Harness: CUGA, Hermes, and Deep Agents — When to Pick CUGA?

The blog post introduces CUGA, a new agent harness from IBM Research, and compares it with existing harnesses like Hermes and Deep Agents. It explains the design philosophy and use cases where CUGA excels, such as complex multi-step tasks requiring structured reasoning. The post provides practical guidance on when to choose CUGA over alternatives.

Why it matters: Provides a concrete comparison of a new AI agent harness, offering actionable selection criteria for practitioners.

Hugging Face Blog2d agoScore 85

AI agentsagent harnessestool selection

Research

Avatar Intelligence: A Paradigm for Persistent Personal Agents

This paper proposes Avatar Intelligence, a paradigm for persistent personal AI agents that represent users' goals, values, and constraints through sustained decision delegation. It introduces a maturity roadmap from human-directed execution to autonomous operation and an Agentic Oversight Index (AOI) for evaluating value against oversight burden. The authors identify three core research challenges: identity-continuous personalization, user-situated evolution via Rehearsal Engineering, and safety mechanisms combining technical control with accountability.

arxiv2d agoScore 78

AI agentshuman-AI interactionagentic oversight

Research

Toward Knowledge-Centric AI: Explicit Knowledge Layer Across Agents and Generations

The paper argues that AI agents should be built around an explicit, continuously evolving knowledge layer independent of any single agent, rather than trapping knowledge in model parameters or context windows. It identifies internalization as the bottleneck, proposing knowledge foundation models for shared neural capacity and a research agenda for reuse, cost efficiency, auditability, and continuity.

arxiv2d agoScore 78

AI agentsknowledge representationmachine learning research

Research

CoworkOS: A Harness for Long-Horizon Agent Coworkers in Real-World Automation

CoworkOS is a general-purpose harness for long-horizon work that separates task strategy from execution management, enabling agents to handle complex productivity tasks with human-in-the-loop participation. It features a three-layer architecture, hierarchical workspaces, and persistent semantic memory for continuity across sessions. In benchmarks, it is competitive with frontier agent applications, and a case study shows automated post-training improving model scores by +9.76 over five rounds.

arxiv2d agoScore 78

AI agentsautomationhuman-AI collaboration

Research Editor selected

ScienceCUA: Turning Documentation into Evolving Skills for Scientific Computer Use Agents

ScienceCUA is a two-stage framework that converts software documentation into executable, reusable skills for scientific computer-use agents. It first constructs skills from manuals and web pages, then uses execution feedback to evolve those skills by contrasting successful and failed trajectories. The framework also introduces two improved benchmarks, ScienceBoard-Verified and ScienceBoard-VE, to reduce noise and leakage. Experiments show up to 17.12 percentage points improvement in task success rate over the base model.

Why it matters: This paper presents a concrete method and benchmarks that improve AI agents' ability to use scientific software, with measurable gains and practical implications.

arxiv2d agoScore 82

AI agentsscientific computingskill acquisition

Research Editor selected

Top-k Retrieval Leaves Memory Value Undetermined; LEDGER Estimates Determined Values

This paper analyzes memory-augmented agents and shows that top-k retrieval leaves memory values undetermined, even with many logged episodes. The authors introduce LEDGER, which estimates determined Shapley values with confidence intervals and reports others as undetermined, improving the fraction of determined memories on benchmarks.

Why it matters: Provides a novel theoretical result and a practical method for memory valuation, directly useful for AI memory system design.

arxiv2d agoScore 85

memory valuationretrieval algorithmsShapley values

Research Editor selected

Grid Agents: An Agentic Framework for Automated Power System Studies and Grid Connection Assessment

The paper introduces Grid Agents, an AI system combining a knowledge ontology (Grid Ontology) and an engineering agent (Grid Studio) to automate power system studies and grid connection assessments. It includes a benchmark suite of 519 tasks from eight power-system familiesched, with verification via replaying tool calls. The system outperforms general coding agents and retrieval pipelines on regulatory evidence tasks and leads on several RAG leaderboards, while Grid Studio achieves top scores and faster performance on power-system tasks.

Why it matters: This is a concrete, recent AI development with a novel agentic framework, benchmark suite, and measurable results, offering practical value for power system engineering and AI evaluation.

arxiv2d agoScore 85

AI agentspower system engineeringretrieval-augmented generation

Research Editor selected

Strategist: Meta-Agent That Evolves Inference Strategies Across Tasks

Strategist is a meta-agent that automatically evolves inference strategies for AI agents, treating them as executable blueprints that can be reused and built upon. It retrieves relevant blueprints, writes candidate strategies, tests them on real samples, and revises them, achieving higher accuracy on 24 of 30 benchmarks with 39% lower inference cost. The strategies transfer to unseen tasks and other models without further search.

Why it matters: This introduces a concrete, novel method for automating the reasoning component of AI agents, with measurable improvements across many benchmarks, making it valuable for AI practitioners.

arxiv2d agoScore 85

AI agentsinference optimizationmeta-learning

Industry Editor selected

Y Combinator's Garry Tan spends up to $200K a year on AI agent tokens

Garry Tan, CEO of Y Combinator, revealed on Walter Isaacson's podcast that his heavy use of AI agents costs him up to $200,000 annually in tokens. He joked about experiencing 'cyber psychosis' and sleeplessness due to his excitement. The admission offers a rare glimpse into the personal financial commitment of a prominent tech leader to AI agent technology.

Why it matters: Provides a concrete, real-world benchmark for AI agent token costs from a highly influential figure in the startup ecosystem.

TechCrunch AI2d agoScore 82

AI agentstoken costsY Combinator