All AI news

AIAnthropic says dynamic workflows are powerful but can use many tokens, so users should start with a scoped task and increase complexity gradually. It points to the /claude-api managed-agents-onboard bug-hunter command in Claude Code and to templates at

AIPerplexity Developers says its new cookbook builds a browser agent that sends a screenshot and questions to pplx-decider-v1.1-27b through the Decisions API, which accepts text and image inputs. The developer's code converts the returned probabilities into clicks, scrolls, and stops.

AIMike Knoop says ARC-AGI-2 is far harder than ARC-AGI-1, even amid rapid progress on math. He adds that the final open solutions will be useful artifacts to study, and notes that the Kaggle Grand Prize bonus threshold of 85% has been reached this year, the final year for ARC-AGI-2 on Kaggle.

AIOwain Evans's team reports that models can transfer more complex traits than owl preferences through training data, including novel skills, agentic hacking, and backdoors. In the new paper, a backdoor transfers even though neither the trigger nor the backdoor behavior appears in the training data.

AIDex Horthy says the share of tasks that can be one-shot without strict process has grown, but alignment, grilling, and planning workflows still matter. He argues that heavy planning on small tasks makes developers feel slower, and predicts tools will add escape hatches so humans or models can decide to ship directly. He adds that as model capabilities improve, the "smart zone" has grown to roughly 200k–400k tokens, and HumanLayer is prototyping research-to-implement and research-to-short-design-to-implement workflows.

AILangChain says its Interrupt archive is available, with every session watchable on-demand at Context from the quoted post: LangChain simplified agent authentication, memory, and channels, and added web search as a pre-built tool for Managed Deep Agents.

AIReplit's X account shares a showcase of TikTok Ads MCP, which runs TikTok Ads from Replit. The post is a broadcast link with no further details on features, pricing, or availability.

AIFactory says it is acquiring Mentlio (YC S26), a startup building in-harness model routing, context compression, and AI usage measurement to optimize cost and quality. The founders, Ashank Shah and Ahmet Demirbas, will join Factory to bring this work into its autonomous software engineering platform for large enterprises.

AIOpenAI told investors it had $18 billion less revenue than the $68 billion it reported last month, and AI-linked stocks including Nvidia, CoreWeave, Oracle and Nebius fell Thursday. Google debuted a Gemini assistant that can act autonomously, generate code and complete work across web, mobile and desktop. Anthropic released Claude Haiku 5.5 and halved Sonnet 5.5 cache read prices.

AIOpen-source AI models, once downloaded, make illegal activity almost impossible to trace. The source gives no figures, model names, or further details on the growth it reports.

AIJared Gross, head of institutional portfolio strategy at JPMorgan Asset Management, says the AI buildout is a boom, not a bubble. He discusses AI asset allocation in a Bloomberg interview as credit markets grow jittery over more mega debt deals financing the AI arms race.

AIThe Wall Street Journal's tech newsletter covers how to prepare homes for extreme weather, how AI could help short-staffed fire departments, the electronic battlefield in warfare, and a startup seeking to double the world's compute. The source provides only a teaser, so no further figures or product details are available.

AIWPP has opened a flagship AI-enabled production facility in east London as part of a £600m WPP Production business, with CEO Cindy Rose saying AI is ushering in a "golden age of marketing". Some AI-led ads can be made up to 60% cheaper, one industry source says, while critics including Luke Scott warn that a lost generation of directors may miss the hands-on training that built careers like Ridley Scott's.

AIThe poster introduces Rei-v1, the first model they have trained, describing it as a tiny model that makes decisions for research agents. The post says it performs best on anime, programming, and academic research. Model weights are linked on Hugging Face.

AI404 Media's Behind the Blog column discusses AI and spirituality, with Jason saying the outlet writes about AI's current capabilities and harms rather than dismissing it outright. He says reporters sometimes test AI tools while working on stories to write from an informed perspective. The excerpt does not say more about the spirituality discussion.

AIAndreessen Horowitz partner Olivia Moore released a report on the top 100 consumer AI apps, finding ChatGPT still leads by a wide margin while smaller players like Suno and ElevenLabs show staying power. Moore says almost all AI revenue comes from subscriptions and token usage, and that most consumer AI is prosumer AI. The report finds no top-100 entrants in social, dating, marketplace, retail, travel, finance, or health categories.

AILauren Tan (@poteto) posts a short recommendation to try coding with Grok Bot. Jonathan Wilke (@jonathan_wilke) says he now handles 90% of his coding work through Grok and calls it amazing.

AIMike Knoop says the 85% Grand Prize bonus threshold has been reached on Kaggle. The ARC Prize 2026 leaderboard lists Tufa Labs first at 88.06%, followed by Rabbithole at 80.56% and Yi-Chia Chen at 77.22%. Knoop says this will be the final year for ARC-AGI-2 on Kaggle and expects an open-source, low-cost, offline reproducible solution and model.

AIEthan Mollick says Google's main challenge after Gemini 4 is what it does with a strong model. He argues that Anthropic and OpenAI are moving toward a single interface for many tasks using orchestrator agents. He says the fragmented products of the Gemini 3 era will not work for what comes next.

AIAi2 says its old scheduler made every scheduled workload eventually run at HIGH priority. Researchers kept idle jobs running to reserve GPUs for experiments, because the incentives rewarded holding capacity even with no active work.

AIAi2 reports that over a 30-day test of its new scheduler, teams received 98% of the GPU hours they were owed, based on actual demand. Cluster occupancy stayed at 98%, and spare capacity went to interruptible work without drawing down team budgets.

AIAi2 has published a blog post on the design decisions behind its scheduler, along with rollout lessons and next priorities. The post links to The text provided does not include further details of the scheduler's features or results.

AIMeta spent heavily on Anthropic's Claude models, with internal use reaching up to 60,000 employees and a projected $10 billion yearly spend, according to The Algorithmic Bridge. The author says Meta then released Muse Spark, which scored 52 on the Artificial Analysis intelligence benchmark, on par with Claude Opus 4.6.

AITessl Blog describes an agentic code review workflow built for teams whose coding agents produce pull requests faster than humans can review them. The workflow runs review against a written standard in the repository, applies four parallel perspectives covering correctness, security and privacy, scale and resilience, and maintainability, then records each finding, verdict, and response. Tessl Code Review, which the post says is free to start, runs these perspectives as skills, and the team's memory of past decisions is fed back into the standard.

AIAi2's AI Infrastructure team replaced its priority-based scheduler for GPU clusters with GPU time budgets, hierarchical fair-share allocation, and a time-slicing contract. The team says the change moved debates over how much GPU time each research project deserves from case-by-case operational decisions into a transparent budgeting process. The clusters range from 88 to 1024 GPUs across NVIDIA H100, B200, and B300 hardware, and serve about 150 internal researchers.

AIMatthew Green says there is a 1% chance cryptography is impossible in the Minicrypt world, and a 15% chance public-key encryption algorithms lose confidence. He warns that AI produces surprises far faster than humans can replace cryptographic standards, so recovery requires preparation in advance.

AIPostman describes the architecture behind Agent Mode, its AI agent for API testing, documentation, discovery, and implementation. The post covers limiting tools per task, using schema-based queries, building purpose-shaped context handlers, and running on Amazon Bedrock with cross-Region inference and prompt caching. Postman reports that tool-selection errors rose once the visible toolset exceeded about 40 tools.

Why it matters: The post shows concrete patterns for tool scoping, context handling, and Bedrock routing and caching, which apply to any team moving an agent past a prototype.

AIAmazon Bedrock Managed Agents, powered by OpenAI, entered public preview, and OpenAI's GPT-6 Astra, GPT-6.1 Sol, and GPT-6.1 Luna became generally available on Amazon Bedrock. AWS also released Strands Decider 2B, a 2B-parameter open source decision model that answers in about 115ms locally, and said the Strands harness uses 28 percent fewer tokens than popular harnesses while matching their accuracy.

AIAi2's AI Infrastructure team replaced its priority-based GPU cluster scheduler with a system using GPU time budgets, hierarchical fair-share allocation, and a time-slicing contract. The team says the change turns decisions about how much GPU time each research project receives into a transparent administrative budgeting process. Its clusters, which range from 88 to 1024 GPUs including H100, B200, and B300 units, serve about 150 researchers facing demand two to three times available capacity.

AIFrancois Chollet says ARC-AGI-2 scores on Kaggle are now very good. The ARC Prize 2026 high score is 88.06%, set by Tufa Labs, and teams scoring over 85% share a $150K bonus prize.