DeepSeek released a packaged desktop version of DeepSeek Harness for macOS and Windows. Linux users can get it from the @deepseek-ai/dsh package on npm. Details: deepseek.com/harness.
Earendil released Pi 1.0 and Pi Durable, both of which reached the Hacker News front page. Pi 1.0 adds native MCP, Jev, and image-model support in Codemode; virtual-model extensions; lazy tool loading; Anthropic model-cache warming; system messages in the middle of conversations; a new TUI theme; and full-screen mode by default.
ServiceNow CoreAI built AutoSynthData to find capability gaps from a target model's failures and a stronger teacher's successful trajectories, then generate and validate new tasks for post-training. As the model improves, the curriculum shifts to tasks that remain difficult. The process was demonstrated on the EnterpriseOps Gym dataset. Generated tasks must meet feasibility, realism, and difficulty requirements; validators must provide consistency, reliability, and completeness.
GPT-6 Astra Ultrafast is now available in the OpenAI API and to eligible ChatGPT Work and Codex users. It runs on NVIDIA Blackwell GPUs.
Why it matters: The source gives Astra Ultrafast's speedup over standard mode and access points, helping readers assess its practical effect on coding-agent loops.
AWS Professional Services uses a four-agent pattern to reduce infrastructure-as-code development per application from three to four weeks to minutes. Built with the Strands Agents SDK and running on Amazon Bedrock AgentCore, the pattern connects source and target systems through MCP tools exposed by AgentCore Gateway.
Amazon Payments used a multi-objective contextual multi-armed bandit (LinUCB) on Amazon SageMaker AI to personalize content in its product-acquisition funnel. In a seven-week online A/B test, one audience segment achieved a high-single-digit relative increase in final funnel conversion, while another showed no improvement. The team ran one LinUCB model for each stage—application start, submission, and approval—and linearly combined their UCB scores to optimize the full funnel. It open-sourced a repository that can be tried with synthetic data.
NVIDIA NeMo Agent Toolkit (NAT) can use a custom memory provider to connect to Amazon S3 Vectors as a persistent memory layer for multi-agent systems deployed on Amazon EKS.
A complete implementation for Claude Platform on AWS (CPonAWS) lets production AWS workloads, developer laptops, and external CI/CD systems share one subscription while isolating production and development traffic through workspaces. It uses three access paths: a dedicated AI Services account with cross-account SigV4, workspace-level API keys, and OIDC federation. External environments use short-lived credentials and require no persistent keys.
Amazon Bedrock AgentCore supports building 'ambient agents' that listen to event streams such as Amazon S3 event notifications, EventBridge rules, and Lambda triggers and perform tasks automatically. When needed, they pause for human approval through a single ask_human tool and resume from where they stopped.
AI2 released Olmo-core 3, a major upgrade to its large language model development framework. It redesigns the open MoE training system to scale MoE training to trillions of parameters while maintaining compute efficiency.
Why it matters: The source provides throughput and memory comparisons for the MoE training stack, helping readers assess changes in the cost of large-scale MoE training.
uniopen used supervised fine-tuning on Amazon SageMaker AI and prompt-level output optimization to adapt Amazon Nova 2 Lite to its retail review policies.
Photon raised a $4.5 million seed round co-led by Gradient and A*, with participation from Vercel, Sequoia (HongShan), and others. The startup helps developers build agents for iMessage and WhatsApp and says it has attracted more than 40,000 registered developers. Revenue grew 10-fold in four months, and message volume grew fivefold last month. Its managed platform offers 99.95% availability and is certified for SOC 2 Type II and HIPAA compliance.
NVIDIA says AI factory return on investment depends on three factors: output capacity, useful life, and demand. Its platform uses full-stack co-design to maximize throughput per megawatt. SemiAnalysis AgentX data show Vera Rubin NVL72 delivers more than 30 times the throughput per megawatt of GB300 NVL72 and, on DeepSeek V4 Pro, can reduce cost per million tokens by up to 45 times.
Satlyt, founded by former Google Cloud and SpaceX Starlink engineer Rama Afullo, raised an $8 million seed round led by Houston-based Non Sibi Ventures. The funding will support software for running AI models in orbit on satellites.
After OpenAI DevDay, Latent Space interviewed Ari Weinstein, who leads computer-use product and engineering, and Nikunj Handa, head of API product.
Flow Engineering, a startup building AI tools for hardware design, raised a $50 million Series B at a $750 million valuation. The round was co-led by Antonio Gracias of Valar Equity Partners and Gavin Baker of Atreides Management, with Sequoia Capital participating.
OpenAI and EDA company Synopsys signed a multiyear strategic partnership to jointly develop GPT-Synopsys, a specialized chip-design model combining OpenAI AI technology with Synopsys EDA tools. OpenAI has been licensed to use Synopsys EDA tools.
An AWS blog post provides a full tutorial for deploying a multi-agent music-production pipeline on Amazon Bedrock AgentCore Runtime Instances. Three agents handle composition, delivery, and compliance, sharing a session through the same runtimeSessionId and colocating on one GPU instance.
Microsoft Research developed an end-to-end machine-learning pipeline that uses L1 solar-wind observations and AE and Dst indices to generate localized space-weather risk estimates 30 to 60 minutes ahead for 66,935 substations in the contiguous United States.
Amazon Bedrock Knowledge Bases added the AgenticRetrieveStream API, which breaks claims questions into subqueries, retrieves information over multiple rounds, and returns cited answers.
Destro emerged from stealth with an $8 million seed round. Its AI intelligence layer lets robots work in logistics settings while directing human workers, instead of building its own robot hardware.
DeepSeek and Huawei partnered to build programming tools for Huawei Ascend chips and open-sourced them. The core is TileLang, a programming language originating with Peking University researchers. DeepSeek says its programming model is simpler than CUDA and that it has used TileLang for about a year as a main tool in its AGI work.
Cerebras Systems CEO and co-founder Andrew Feldman will appear at TechCrunch Disrupt 2026 to discuss whether demand for AI compute, energy, and infrastructure can continue to scale.
CoreWeave announced that NVIDIA Vera Rubin NVL72 systems and Spectrum-X 102.4T Ethernet are available on CoreWeave Cloud, making it one of the first cloud providers to deliver the platform. Cognition is the first customer to run Vera Rubin in production.
Why it matters: CoreWeave launched Vera Rubin NVL72 and published measured Devin throughput, offering a view of compute costs for agentic coding workloads.
Berlin startup Restate raised $20 million in a Series A led by Singular, with Redpoint Ventures and Capital One Ventures participating. Restate provides a durable workflow execution engine that allows multi-step processes to recover after crashes and network interruptions. Its in-house storage, replication, and redundancy layers make systems faster and lighter. Replit is among its customers. The new funding will go toward building a go-to-market team, hiring engineers, and expanding its San Francisco Bay Area office.
QbitAI's interview program invited Gu Xin, who leads sales and ecosystem success for Siemens Xcelerator in China, and Huang Yao, founder and CEO of Aqiu Tech, to review the path from industrial AI deployment to scale. Huang used semiconductor wafer cutting as an example, saying that after introducing agent technology, equipment OEE and capacity can improve by about 25%. Gu said industrial AI must be reliable at an industrial level. Siemens Xcelerator aims to serve as a connector and amplifier, filling in the delivery and ecosystem pieces needed to scale from 1 to 100.
DeepSeek added native support for Huawei Ascend 950 NPUs to the TileLang operator-development toolchain and released Ascend versions of five compute and communication libraries: DeepGEMM, DeepEP, FlashMLA, TileKernels, and DeepSelect.
Why it matters: This report outlines the scope of TileLang's native Ascend support and related operator libraries, showing current coverage in China's domestic compute toolchain.
Using three photos of the bronze rat head from Yuanmingyuan and a natural-language instruction, GPT-6 Astra used Codex to direct Aholo Lux3D to generate a 3D asset, then used Blender to inspect it, adjust materials, add a display stand, and render and export the result.
Leiphone compared free AgnesCode with Agnes 3.0 Flash against Codex with GPT-5.6 Sol on Task66, a DeepSeek-V3 operator optimization challenge from the Zhongzhi FlagOS competition.
Huawei HiSilicon's in-house Hi-ONE optical engine will be deployed using NPO in the next-generation supernode. In the Ascend 960 supernode, 5,500 units will replace 48,000 pluggable 800G optical modules. The Lingqu interconnect equipment between racks supports 176 1.6T ports, with about 280T of all-optical interconnect bandwidth per machine.
The author tested Qwen3.8-Flash and GLM-5.3-Flash, released on the same day, in Claude Code on the same task: building a personal workspace from scratch with tasks, calendar, notes, a dashboard, and global search.
The author tested WorkBuddy, Qianwen Office, and Doubao Work using the same messy sales records and prompt. All three delivered an Excel file, a presentation, and a summary, but the highest quarterly total was 57.7% above the lowest. The difference came from how each handled orders with missing department data: exclude them, keep them, or mark them for verification.
Of the 45 layers in GLM-5.3-Flash, 34 use Kimi's KDA linear-attention approach and 11 use DeepSeek's KPool-DSA sparse attention. Linear and Sparse attention appear together in one model for the first time. The combination reflects a shift away from finding a single best mechanism toward assigning different roles: Linear maintains long histories at low cost, while Sparse retrieves specific information as needed.
DeepSeek published an exclusive technical article on Zhihu, explaining for the first time the DSec sandbox infrastructure behind all DeepSeek-V4 training, evaluation, and data preprocessing. The technical report is available on arXiv, jointly authored by DeepSeek and Tsinghua University with a team of more than 130 authors.
At DevDay 2026, OpenAI launched the always-on Dots agent, GPT-6.1 Sol, Ultrafast mode, Decisions API, ChatGPT Spaces, and Marketplace, and said ChatGPT has 1.2 billion weekly active users.
Why it matters: This overview gathers OpenAI DevDay 2026 announcements and pricing changes alongside third-party evaluations and community feedback.
Stanford's Movement Lab and Caltech published HomeBody, connecting GPT Astra to a Unitree G1 so the robot can explore an unfamiliar kitchen and then carry out sequences such as finding objects, getting medicine, throwing out trash, and opening drawers.
On September 30, DeepSeek open-sourced foundational components for Huawei's Ascend compute platform, including the TileLang high-level language compiler, high-performance computing libraries, and distributed communication libraries. They correspond to components previously open-sourced for GPU platforms.
Ollama now supports Jev-like decision models all locally. Use decision models like Nimble for tasks like ticket triaging, model routing, and content moderation. ollama pull nimble Here’s Nimble playing Ollama racer through the new local /v1/systemone API by making decisions in real-time. 🏎️
NIO and Geely Holding announced deeper charging and battery-swapping cooperation on September 28. Geely will take a 30% stake in NIO Power by contributing 100% of E-Energy Connect and RMB 640 million in cash. NIO China plans to take a 10% stake in Haohan Energy. After the transactions, NIO Power will have a post-money valuation of RMB 16 billion. The companies plan to build 10,000 battery-swap stations by 2030; the network's annual electricity demand is expected to exceed 10 billion kWh. They will also integrate charging assets and build stations combining solar, storage, charging, and battery swapping.
Amazon Web Services published a solution for automating the Amazon Textract adapter lifecycle across accounts using AWS CloudFormation and Terraform templates to promote and deploy adapters across multiple accounts.
