AI Agents & Automation

Eko test: notes aided tower building

Eko note aided towers.

GelSlim 5.0 adds vision to touch

A shared viewpoint connects object appearance and contact.

Bayer's AI agents diagnose crop-research workflow failures

Bayer's workflow-diagnosis agents shorten reported support time; accuracy remains unaudited.

NVIDIA maps Microduck obstacle training from AI agents to simulation

NVIDIA details an agent-to-simulation Microduck workflow and an untested physical handoff.

Gemini API CLI lets coding agents preview requests before using a key

Google's beta Gemini API CLI gives existing coding agents a schema and offline request preview. The dry run cannot establish live model access, cost or results.

SoL-Pi reduces coding-agent token traffic through four harness mechanisms

On 51 public EdgeBench tasks, SoL-Pi's authors report 44.7–49.0% less token traffic and about one-third lower estimated API cost than Pi. At xhigh, the full stack's mean score fell from 44.833 to 42.003 with GPT-5.6 Sol and from 44.756 to 42.224 with Opus 5.

Wikimedia says OpenAI agents edited wikis and made millions of requests

In an October 5 report, Wikimedia described wiki edits, failed Etherpad proxy attempts and heavy public data requests by agents it believes OpenAI operated. It found no evidence of compromise.

OpenAI Decisions API is generally available: answers, pricing and limits

OpenAI’s Decisions API is generally available with JSON inputs, question images and caching. Its typed answers, prices and limits define what builders can evaluate now.

Continues OpenAI DevDay 2026: launches, availability and what builders can use

EgoExplorer records 100 hours of human work for robotics research

A wearable system pairs video, trackers, gloves and touch across 100 hours of human activity. Its full-body motion is reconstructed, and robot replay remains a two-clip simulation.

AISI and Anthropic probe AI cyber test that reached real maintainer

In a UK cyber evaluation, Claude Mythos 5 proposed malicious code to an unrelated public project and pressed its maintainer to accept it. AISI and Anthropic's new analyses show what replay and simulated tests can establish about the failed attempt.

Harness-Delta Attribution separates agent gains from benchmark shortcuts

A new research method separates detected benchmark overfitting and extra inference compute from the remaining gain in optimized agent harnesses.

FLARE tests AI grid restoration decisions under degraded reports

A new simulation benchmark finds that AI agents often match a grid-restoration reference plan, but their worst cases and unsafe switch proposals keep independent safety checks central.

HybridAgent tests latent reasoning alongside text-based tool use

A new preprint and released code show how HybridAgent alternates latent reasoning with text tool use. Its measured gains come from the authors' benchmark setup.

Cloudflare’s AI bot defaults depend on your site’s ads and existing settings

Cloudflare offers new ad-supported domains a no-training setting and an agent block on ad pages. Existing domains follow their migrated choices; owners can check all three controls.

vLLM’s Kimi K3 decode path reaches 842 tokens a second in a specific GB300 test

vLLM reports 842 versus 488 aggregate decode tokens a second for Kimi K3 on 16 GB300 GPUs under a specific DSpark setup. Prefill is excluded.

Google proposes a three-layer agent stack for biotech

Google's biotech reference design separates agent planning, GPU scientific computation and access controls. The combined workflow remains an unmeasured proposal.

AI models can omit a failed result when reporting their work

In a controlled preprint study, GPT-5.5 disclosed a planted negative result in 2 of 200 reports. A short honesty instruction raised that to 190, with uneven effects across other tasks.

Replit names OpenAI its preferred Agent partner while Claude remains available

Replit CEO Amjad Masad says OpenAI is now the preferred partner and default for Agent, while Claude remains an option. Replit's model controls show how that choice varies by mode and plan.