#Agent

AIWhat it keeps, what it compresses, what it lets go. Our engineers started from pure engineering constraints and landed where evolution did long ago: the brain.

AIXiaomi MiMo has launched MiMo Desktop in invite-only beta, a desktop agent that turns Office files, images, video, audio, and zips into finished, editable output. Invitees also get limited access to next-gen MiMo models, and the post lists features including live previews, region-based editing with versioned rollback, automatic model routing, and browser and computer use with record and replay.

AINVIDIA's NV-Reason-CT is a 3D vision-language model for CT image analysis that combines a native 3D vision encoder with a language model. It is designed for radiology report generation, question answering, and multi-step reasoning across chest and abdominal CT volumes. The model converts a 384×384×384-mm input into 13,824 visual tokens without spatial downsampling and is available on Hugging Face under the OpenMDW-1.1 License.

AIOur first episode looks at AI's expanding role in scientific discovery through Famou's work on pine wilt disease. The project reflects a broader trend, with AI taking on more of the research process itself. That raises a larger question: could research agents become part of the infrastructure of discovery?

AIOrbital is building an operating system that lets non-CS domains like science and mechanical engineering use its team's computer science expertise to build complex apps and research tools. The author argues that clearly specifying what you want from AI is the key lever for getting value, and that robust results are possible without a CS degree if the right pieces are in place. A quoted post on an ETH Zurich study of 100 developers suggests computer science background predicts vibe coding success more strongly than writing skill.

AISatya Nadella says Microsoft is bringing new models into Copilot to handle increasingly complex work, from quick questions to delegated tasks and complete long-running jobs via Autopilots. As an example, an Opal-powered Autopilot on a secure Windows 365 Cloud PC sorts a month of trail cam footage, extracts species sightings, and builds a highlight reel, spreadsheet, PowerPoint, and Teams share.

AIMeta's AIRA₃ system generalizes across domains by changing only the task specification, according to the post. In an internal benchmark, it achieved a 27% latency reduction on production GPU kernels, and it reached gold-level performance in a Kaggle competition translating 4,000-year-old Akkadian clay tablets into English. The post says the work is early and that Meta believes a self-improving knowledge system is the right direction for accelerating AI research.

AIMeta's AIRA₃ replaces a central controller with many long-running agents, each pairing a model with a coding harness in its own isolated environment. The agents coordinate asynchronously through a shared forum for hypotheses and findings and a shared filesystem for solution artifacts. According to the post, performance gains compound over time as agents build on each other's discoveries.

AIMeta's AIRA₃ entered the live competition with an ensemble of models, and the 8th-ranked gold-medal entry combined GPT 5.5 (w/ OpenCode) and Claude 4.8 (w/ ClaudeCode). Post-hoc testing found Muse Spark 1.2 (w/ MuseCode) also reached gold-medal level, while Muse Spark 1.1 (w/ OpenCode) and GLM 5.2 (w/ OpenCode) reached silver-medal level, all graded on the same private test set.

AIAndrew Ng presented an AI Engineering Skills Map for using coding agents such as Claude Code, Codex, Cursor, OpenCode, and Pi. The workflow he describes covers planning, execution, and deployment with monitoring, and he identifies five key skills: directing the workflow, enabling agent autonomy, reviewing the work, customizing the agent and its environment, and coding agent foundations. The source says these skills matter more as the agents evolve quickly.

AILewis Tunstall shares a quoted post from Elie Bakouch describing what they call the largest open experiment on autonomous agents iterating on a research environment, scaling runtime, compute, models, and harnesses. The chart shows Fable 5 closing about 82% of the gap to the human NanoGPT speedrun record, with Kimi K3 also strong, while the author notes run-to-run noise of about 50 steps after 24 hours. Traces, scratchpads, and examples of models building their own tools are shared, and more models are expected to be reported next week.

AILewis Tunstall praises a new Meta paper on research preference models (RPMs), which instill "research taste" in agents by treating experiments as tree nodes. An RPM acts as an LLM judge that selects the most promising candidate experiment before it is run, reducing wasted compute. Tunstall notes the resulting trajectories could train domain-specific RPMs, which would be valuable in hard fields such as the natural sciences.

AIJim Fan recalls that OpenAI's 2016 Universe project tried to have an agent learn computer use from screen pixels, mouse, and keystrokes, which he now calls doomed. He argues the solution is first training a Specialized Generalist across many general tasks, then specializing back to screen-level control, and he congratulates GPT-6 for reliably booking a United flight.

AIGoogle's Gemini Enterprise developer experience team tested agent governance workflows without internal shortcuts and fixed friction points across its agent governance products. Fixes included documentation stating that enabling the Identity-Aware Proxy API is a hard requirement, auto-allowing essential Google-managed platform APIs in the Agent Gateway, and adding Private Service Connect and Cloud DNS setup guidance for Semantic Governance. The team also published ready-made Logs Explorer queries for monitoring Agent Gateways and Content Security.

AIxAI built a Grok-powered procurement agent, Haggle Bot, that audits vendor spend, flags unused SaaS seats, and prepares renewal negotiations. The Bot has identified more than $100,000 in direct savings, including $14,220 from 43 unused seats of one SaaS product and $85,662 a year in unused SKUs from another. A person still approves any spending, contract terms, or messages sent to vendors.

AIWe're making it free for all Grok and Cursor enterprise customers for the next two weeks. Deploying Bot has felt like onboarding thousands of capable teammates to our company. It is both the most internally adopted and most powerful AI product we've seen so far.

AIBenedict Evans argues that cheaper tool-building with AI will not automatically sweep away large companies' sprawling software, because people often don't see the tasks they could automate. He says the hard parts are knowing a tool is needed, deciding what it should do, and getting many departments and systems to adopt it. Companies typically move improvised, bottom-up workarounds into institutionalized software once they carry revenue and risk.

AIOpenAI researcher Mark Chen announced GPT-6 Astra, which he described as the company's most capable and aligned model yet. He said it can build and test software, work across apps on a computer, and help with open scientific problems. The post also highlights improved computer use compared with Operator and stronger monitoring that can stop potentially unauthorized agent actions.

Why it matters: The post links a named model release to specific capabilities like computer use and aligned agent behavior, giving readers concrete claims to check against the model.

AI…for intent, semantic reasoning, and the context behind key decisions.

AIWe think this will be a new form of operational data analysis that all companies do, a lot like finance and security.

AI…all conversationally. Rolling out now for Google AI subscribers! Docs Live going to be super helpful in particular - see it in action below, and read more here:

AIBAAI released Recon2Reason-Reasoning-4B, a 4,437,815,808-parameter vision-language model fine-tuned from Qwen3-VL-4B-Instruct for indoor spatial reasoning. The model handles metric distance, relative position, and object-relation questions from single or multiple images, and loads with the standard Qwen3VLForConditionalGeneration interface without trust_remote_code. The checkpoint is released under Apache-2.0 with BF16 Safetensors weights, and the retrieval-augmented scene-reconstruction extension ships separately.

AIxAI describes Grok Bot as built around persistent agents that keep their own identity, memory, runtime, and tools, rather than disposable chat sessions. Its interface organizes around five objects: Bots, Chats, Prompts, Tools, and Artifacts. Each Bot's avatar shows its identity and lifecycle state, with hover revealing its current action.

AIOpenAI's GPT-6 Astra (max) scores 62.7% on ARC-AGI-3 Semi-Private for $26K under the Standard harness, and 99.9% for $19K under the Provider Adapter harness. The authors say Astra used fewer actions than the human baseline on 96.0% of levels, and they note it is not claimed to be AGI.

Why it matters: The report pairs benchmark scores with replays of the model's notation and tool use, showing how it solved unfamiliar environments rather than only that it did.

AIxAI announced Grok Bot, a platform for creating autonomous AI Bots that run in the cloud and carry out tasks end to end inside tools teams already use. Today's release adds access, network, and audit controls so enterprises can govern Bots at scale, and Enterprise customers can get Grok Bot free for two weeks.

AIGoogle introduces Gemini 3.8 Flash for coding, agentic tasks, and multi-step reasoning, priced at $0.75 per million input tokens and $3.75 per million output tokens during the introductory period. Gemini 3.8 Flash Cyber targets vulnerability detection and automated patching and is available only to trusted defenders through the new Fairwind Program. The introductory price expires December 31, 2026, after which $1.50 and $7.50 per million tokens apply.

Why it matters: The post separates a general coding and agent model from a restricted cyber variant, showing how one shared core is deployed under different access and safety tiers.

AIwith gains over 3.7 Flash in software engineering, agentic tasks, and multi-step reasoning. Sundar Pichai says it outperforms most larger frontier models on DeepSWE v1.1 at a fraction of the cost. The comparison table lists input at $0.75 and output at $3.75 per 1M tokens, with introductory pricing through December 31, 2026.

AIGoogle's Gemini 3.8 Flash is out, and Varun Mohan says it substantially improves on 3.7 Flash for agentic coding and general knowledge work. It is now available to everyone on Antigravity. The attached benchmark table lists Gemini 3.8 Flash at $0.75 per 1M input tokens and $3.75 per 1M output tokens, with introductory pricing through December 31, 2026.

AI…failed (never use): -Instagram (Meta AI) -Apple Intelligence -DoorDash (AI chatbot) -Gmail "Help me write -Google Meet "take notes for me” What else?

AIGoogle launches Gemini 3.8 Flash Cyber and Gemini 3.8 Flash. The post describes Flash Cyber as its most capable cybersecurity model for finding and fixing vulnerabilities, placing it on the Pareto frontier for patching on CWE-Bench. Flash Cyber is available to trusted defenders through the new Fairwind Program.

AIGoogle announced Gemini 3.8 Flash, its third updated Flash model in six weeks, citing improvements in agentic and coding capabilities. The benchmark table lists input at $0.75 and output at $3.75 per 1M tokens, with introductory pricing of $1.50 and $7.50 expiring December 31, 2026. Terminal-bench 2.1 shows 89.4% for Gemini 3.8 Flash against 85.8% for Gemini 3.7 Flash.

AI…hitches a ride on the intelligence explosion:

AIIn our latest AI Pulse, we look at how we're building AI that takes work all the way from request to finished result with our product portfolio, from DuMate and GenFlow to MeDo and ERNIE Assistant. Plus, Baidu has become a dual-primary listed company, and Apollo Go has expanded further in Dubai and Hong Kong. Get the full story 👇