English text is being prepared.

By relevance · Total 1633 results Clear search

Oct 282026Wednesday

01:19
Unite.AIOct 28 01:19
Oracle Releases AI Security Mesh Powered by OpenAI Cybersecurity Models
At Oracle AI World, Oracle released Oracle AI Security Mesh, a managed platform powered by OpenAI's frontier cybersecurity models that helps enterprises discover and remediate security vulnerabilities faster, covering applications and infrastructure on-premises, on OCI, and on other clouds. The platform first uses a discovery framework to map customer systems, then dedicated agents investigate weaknesses, verify reproducibility, and help prepare code fixes; all fixes must be approved by the customer before being applied, and each agent runs with an independent identity and just-in-time permissions. Oracle also mentioned that OCI Identity Assurance can use biometrics to confirm the identity of approvers for sensitive operations, while OCI Zero Trust Packet Routing provides underlying network access control for AI workloads.
At Oracle AI World, Oracle released Oracle AI Security Mesh, a managed platform powered by OpenAI's frontier cybersecurity models that helps enterprises discover and remediate security vulnerabilities faster, covering applications and infrastructure on-premises, on OCI, and on other clouds. The platform first uses a discovery framework to map customer systems, then dedicated agents investigate weaknesses, verify reproducibility, and help prepare code fixes; all fixes must be approved by the cust
Matched fieldsBody
#hot-signal #Product updates #Safety / alignment #Agents
01:33
Unite.AIOct 28 01:33
Oracle Launches Media Fabric AI Platform, Integrating Content Production and Streaming Distribution
At Oracle AI World, Oracle launched Oracle Media Fabric, integrating AI content production, enterprise media models, and distributed video distribution into a single platform. The platform consists of Oracle Creative Studio, Oracle AI Media Models, and Oracle Video Edge; Creative Studio is model-agnostic and can select a model for each creative task based on security, provenance, cost, and IP protection requirements. Oracle says Media Fabric is planned to be available within the next 12 months.
Matched fieldsBody
#hot-signal #Product updates #Video #Multimodal
06:40
Unite.AIOct 28 06:40
GSA Expands OneGov Agreement with Anthropic, Raising Claude Procurement Discount to 40%
The U.S. General Services Administration (GSA) and Anthropic announced on October 27, 2026 the expansion of the OneGov agreement, under which government buyers receive a 40% usage-based discount on Claude Enterprise, Claude for Government, and the Claude API, with the scope extended from the federal government to state, local, tribal, and territorial governments. The agreement sets no minimum purchase volume; agencies pay based on actual usage and can place orders through direct sales, resellers, and major cloud platforms. Anthropic says about 500 U.S. federal, state, and local government agencies are using Claude, and weekly active users of Claude for Government have grown about eightfold since February 2026.
The U.S. General Services Administration (GSA) and Anthropic announced on October 27, 2026 the expansion of the OneGov agreement, under which government buyers receive a 40% usage-based discount on Claude Enterprise, Claude for Government, and the Claude API, with the scope extended from the federal government to state, local, tribal, and territorial governments. The agreement sets no minimum purchase volume; agencies pay based on actual usage and can place orders through direct sales, resellers
Matched fieldsBody
#hot-signal #Industry news #English text is being prepared.#Anthropic
05:16
The New StackOct 28 05:16
Replit CEO Says 1937 Coase Theory Will Make Organizational Charts Obsolete
Replit CEO Amjad Masad has proposed a "self-driving company" vision, arguing that AI agents will compress internal coordination costs within enterprises, and predicts that by the end of 2030 all Fortune 500 companies will become self-driving companies. To this end, Replit has launched Replit for Business and the "Agent Habitat" concept; Cloud Agents are already available to Pro and Enterprise customers, while skills remain in early access. Replit's internal testing shows token consumption dropped 51%, and a data task that originally took 16 minutes was shortened to 1 minute.
Replit CEO Amjad Masad has proposed a "self-driving company" vision, arguing that AI agents will compress internal coordination costs within enterprises, and predicts that by the end of 2030 all Fortune 500 companies will become self-driving companies. To this end, Replit has launched Replit for Business and the "Agent Habitat" concept; Cloud Agents are already available to Pro and Enterprise customers, while skills remain in early access. Replit's internal testing shows token consumption droppe
Matched fieldsBody
#hot-signal #Industry news #Agents #English text is being prepared.
04:11
Codex：GitHub ReleasesOct 28 04:11
OpenAI Codex 0.165.0 Released: Adds Google Cloud Vertex AI Provider and…
OpenAI Codex 0.165.0 adds a built-in Google Cloud Vertex AI provider (configured via model_provider = "gcp") and adds GPT-6.1 Terra to the model catalog. This version also enables memory v2 by default and supports importing external agent memory, while allowing reasoning effort to be adjusted during conversations by default. The session selector adds repository filtering, and the agent command center supports grouping by directory.
OpenAI Codex 0.165.0 adds a built-in Google Cloud Vertex AI provider (configured via model provider = "gcp") and adds GPT-6.1 Terra to the model catalog. This version also enables memory v2 by default and supports importing external agent memory, while allowing reasoning effort to be adjusted during conversations by default. The session selector adds repository filtering, and the agent command center supports grouping by directory.
Matched fieldsBody
#engineering #Product updates #Agents #English text is being prepared.
00:00
NVIDIA Developer · Generative AIOct 28 00:00
NVIDIA Explains How to Fine-Tune Nemotron 3.5 Lightning with the Nemotron-Personas Dataset
NVIDIA has released the Nemotron-Personas France and Belgium datasets and provided a complete post-training workflow for fine-tuning Nemotron 3.5 Lightning with them. The workflow includes defining behavior contracts, generating synthetic conversations with personas, fine-tuning with LoRA on NVIDIA NeMo, merging adapters, and then comparing and evaluating item by item against the base model on a frozen test set. The France dataset was built in collaboration with Pleias, and the Belgium dataset involved Pleias and KU Leuven; the Nemotron repository provides reusable steps and agent skills for data generation, training, and evaluation.
NVIDIA has released the Nemotron-Personas France and Belgium datasets and provided a complete post-training workflow for fine-tuning Nemotron 3.5 Lightning with them. The workflow includes defining behavior contracts, generating synthetic conversations with personas, fine-tuning with LoRA on NVIDIA NeMo, merging adapters, and then comparing and evaluating item by item against the base model on a frozen test set. The France dataset was built in collaboration with Pleias, and the Belgium dataset i
Matched fieldsBody
#lab #Tutorials / practice #English text is being prepared.#Open-source ecosystem
01:16
Hugging FaceOct 28 01:16Selected
Liquid AI Releases the LFM3 Series of Models, Including 1.7B Text, Vision, Audio, and Omni Versions
Liquid AI has released the LFM3 series of open-weight models, all branched from LFM3-1.7B-Base, covering text, vision, audio, Omni, and the d1 decision model. LFM3-1.7B-Base adopts a three-hybrid-layer architecture, is pretrained on 22T tokens, and in the mid-training stage over the final 3.5T tokens incorporates more agent trajectories and tool-calling data, with the context window extended to 256K tokens. The company says LFM3-1.7B is about 25% faster than LFM2-1.2B at an 8K context, and scores 38.42 versus 3.02 on WebArena-Verified; LFM3-VL-2B's average score across 62 vision benchmarks rises from 57.6 to 71.4. All models are available for download on Hugging Face, along with a single-GPU fine-tuning notebook.
Liquid AI has released the LFM3 series of open-weight models, all branched from LFM3-1.7B-Base, covering text, vision, audio, Omni, and the d1 decision model. LFM3-1.7B-Base adopts a three-hybrid-layer architecture, is pretrained on 22T tokens, and in the mid-training stage over the final 3.5T tokens incorporates more agent trajectories and tool-calling data, with the context window extended to 256K tokens. The company says LFM3-1.7B is about 25% faster than LFM2-1.2B at an 8K context, and score
Matched fieldsBody
#lab #Model releases #Open-source ecosystem #English text is being prepared.
02:53
Latent SpaceOct 28 02:53Selected
Anthropic's Engineering and Product Leads Discuss the Opus 5.5 Release and Claude's Agent Architecture
In a Latent Space podcast interview, Anthropic engineering lead Katelyn Lesse and product lead Angela Jiang discussed the Opus 5.5 release, covering adjustments to the model's writing style, pricing based on intelligence per dollar, and API changes such as requiring thinking blocks to be retained across turns to prevent distillation. The interview also described Claude Managed Agents' approach of separating sessions and the harness from the sandbox that executes code, with credentials mounted separately through a proxy so that a sandbox failure does not take down the entire session; Dreaming, meanwhile, is implemented as one managed agent inspecting another agent's history and writing back to memory. The team said that as safety work for model releases has grown heavier, recent efforts have shifted more toward safety, with progress on features such as skills and memory slowing relatively.
In a Latent Space podcast interview, Anthropic engineering lead Katelyn Lesse and product lead Angela Jiang discussed the Opus 5.5 release, covering adjustments to the model's writing style, pricing based on intelligence per dollar, and API changes such as requiring thinking blocks to be retained across turns to prevent distillation. The interview also described Claude Managed Agents' approach of separating sessions and the harness from the sandbox that executes code, with credentials mounted se
Matched fieldsBody
#newsletter #Industry news #Agents #English text is being prepared.
02:00
Databricks BlogOct 28 02:00
From local patching to quantifiable improvement: Using ai_extract and MLflow Prompt Registry to improve document extraction accuracy
Databricks shared its approach to improving document extraction accuracy using ai_extract and MLflow Prompt Registry: in a project that extracted about 40 structured fields from dozens of PDF layouts, merely modifying the schema halved the errors in the revision field, but the total extraction errors across all fields actually increased by 9%. The article recommends describing field-level rules with the schema, using global instructions for document-level logic, and using precision mode when the structure is complex; it also recommends storing prompts and schemas in MLflow Prompt Registry for version management and alias-based rollback, then evaluating accuracy, coverage, and run-to-run variance based on human-reviewed ground truth. Genie Code can automatically perform per-field scoring, propose reviewable diffs, and rerun evaluations.
Databricks shared its approach to improving document extraction accuracy using ai extract and MLflow Prompt Registry: in a project that extracted about 40 structured fields from dozens of PDF layouts, merely modifying the schema halved the errors in the revision field, but the total extraction errors across all fields actually increased by 9%. The article recommends describing field-level rules with the schema, using global instructions for document-level logic, and using precision mode when the
Matched fieldsBody
#company #Tutorials / practice #English text is being prepared.#English text is being prepared.
00:45
LangChain BlogOct 28 00:45Selected
How LangChain Built the Design Agent Content Studio
LangChain's GTM engineering team, working with the design team, built Content Studio, a design agent inside Slack, on Managed Deep Agents. It can generate videos, social graphics, UI mockups, and slide decks, and has already produced more than 3,000 assets and over 80 slide decks. The article breaks down its four-layer knowledge structure (skills, wiki, code validation, memory), the dedicated LangSmith Sandbox for each Slack thread, the two rendering engines HyperFrames and the in-house motion-studio, and how the main agent orchestrates asynchronous sub-agents. The team says that after migrating to Managed Deep Agents, its own code shrank by about 80%, the median model cost per request fell from about $8 in mid-September to about 50 cents, and the entire harness has been open-sourced.
LangChain's GTM engineering team, working with the design team, built Content Studio, a design agent inside Slack, on Managed Deep Agents. It can generate videos, social graphics, UI mockups, and slide decks, and has already produced more than 3,000 assets and over 80 slide decks. The article breaks down its four-layer knowledge structure (skills, wiki, code validation, memory), the dedicated LangSmith Sandbox for each Slack thread, the two rendering engines HyperFrames and the in-house motion-s
Matched fieldsBody
#company #Tutorials / practice #Agents #English text is being prepared.
08:00
Cartesia Products and ResearchOct 28 08:00
Cartesia Research: From Tokens to Bytes, How Tokenization Priors Affect Language Model Downstream Transfer
Cartesia has released the second article in its series, using two types of scaling laws, compute-to-loss and loss-to-loss, to compare byte-level Transformers, BPE, BPE-NP with pre-tokenization removed, BLT-like, and H-Net. The conclusion is that BPE tokenization is nearly 5 times more compute-efficient than byte-level models in pretraining, but at the same validation bpb, OLMES downstream loss is significantly worse, equivalent to a compute gap of about 1.9 times; H-Net simultaneously obtains the compute efficiency brought by compression and the transfer capability of byte-level models, with a direct compute-to-OLMES fit of about 2.06 times. The article also points out that the manual regularization step of pre-tokenization has a significant impact on both metrics, and discusses the prompt boundary issue of BPE.
Cartesia has released the second article in its series, using two types of scaling laws, compute-to-loss and loss-to-loss, to compare byte-level Transformers, BPE, BPE-NP with pre-tokenization removed, BLT-like, and H-Net. The conclusion is that BPE tokenization is nearly 5 times more compute-efficient than byte-level models in pretraining, but at the same validation bpb, OLMES downstream loss is significantly worse, equivalent to a compute gap of about 1.9 times; H-Net simultaneously obtains th
Matched fieldsBody
#company #Papers/Research #English text is being prepared.#English text is being prepared.
08:00
Google Developers BlogOct 28 08:00
LiteRT in Google Play services gets its biggest upgrade…
LiteRT in Google Play services receives its biggest upgrade: GPU inference based on ML Drift is about 1.4x faster than the old GPU delegate and is provided through the new CompiledModel API. After switching to this runtime, app size shrinks by about 8 MB, because the runtime is distributed uniformly with Play services and is no longer packaged into the APK; currently more than 100,000 Android apps already bundle LiteRT or TensorFlow Lite. This solution supports Android 7 and above, Wear OS 4+, and Android Automotive OS (Android 11+); NPU support is still in development.
LiteRT in Google Play services receives its biggest upgrade: GPU inference based on ML Drift is about 1.4x faster than the old GPU delegate and is provided through the new CompiledModel API. After switching to this runtime, app size shrinks by about 8 MB, because the runtime is distributed uniformly with Play services and is no longer packaged into the APK; currently more than 100,000 Android apps already bundle LiteRT or TensorFlow Lite. This solution supports Android 7 and above, Wear OS 4+, a
Matched fieldsBody
#engineering #Product updates #English text is being prepared.#English text is being prepared.
08:00
Apple NewsroomOct 28 08:00
Apple unveils MacBook Pro with PROLED XDR display and touch
Apple has unveiled the all-new MacBook Pro, bringing the PROLED XDR display and touch operation to the Mac for the first time, available in 14-inch and 16-inch models. The new display uses tandem OLED technology, reaching peak HDR and SDR brightness of 2000 nits, a contrast ratio of 10000000:1, and support for a 120Hz variable refresh rate. The new machines are powered by M5 Pro and M5 Max chips, with up to 128GB of unified memory and 16TB of storage, battery life of up to 24 hours, pre-orders opening October 28, and availability on November 6.
Apple has unveiled the all-new MacBook Pro, bringing the PROLED XDR display and touch operation to the Mac for the first time, available in 14-inch and 16-inch models. The new display uses tandem OLED technology, reaching peak HDR and SDR brightness of 2000 nits, a contrast ratio of 10000000:1, and support for a 120Hz variable refresh rate. The new machines are powered by M5 Pro and M5 Max chips, with up to 128GB of unified memory and 16TB of storage, battery life of up to 24 hours, pre-orders o
Matched fieldsBody
#company #Product updates #English text is being prepared.#OpenAI
08:00
Google Developers BlogOct 28 08:00
Continuously improving agent quality with the Agents CLI evaluation loop
The Google Developers Blog demonstrated the eval loop of Agents CLI v1.8.0: eval generate runs test cases, eval grade scores them, eval compare compares two runs, eval analyze clusters failures, and eval optimize automatically optimizes prompts. The example baseline averaged 4.0/5, and after accidentally swapping in a chef persona prompt it dropped to 2.75; the evaluation caught this regression before it reached production.
Matched fieldsBody
#engineering #Tutorials / practice #Agents #Evaluations / benchmarks
08:00
Google DeepMindOct 28 08:00
Google DeepMind Releases Gemini 4 Carbon, Focused on Cybersecurity and Long-Horizon Task Capabilities
Google DeepMind has released Gemini 4 Carbon, calling it its strongest model, ranking first with 59 points on the Artificial Analysis Intelligence Index v5.0, and achieving 88.2% on SEC-bench Pro, 54.8% on ExploitGym, and 71.0% on CWE-bench. The model is aimed at long-horizon engineering tasks; an internal team once had it run continuously for several days to investigate and fix hidden defects in a large codebase, and it rose to 4th place on WebDev Arena with 1,768 points. Starting this week, Carbon is first being made available to select cybersecurity defenders, then gradually expanding through the Fairwind Program, and only afterward to developers, enterprises, and consumers; after the introductory period, pricing is $4 per 1 million input tokens and $20 per 1 million output tokens.
Google DeepMind has released Gemini 4 Carbon, calling it its strongest model, ranking first with 59 points on the Artificial Analysis Intelligence Index v5.0, and achieving 88.2% on SEC-bench Pro, 54.8% on ExploitGym, and 71.0% on CWE-bench. The model is aimed at long-horizon engineering tasks; an internal team once had it run continuously for several days to investigate and fix hidden defects in a large codebase, and it rose to 4th place on WebDev Arena with 1,768 points. Starting this week, Ca
Matched fieldsBody
#lab #Model releases #Agents #English text is being prepared.

Oct 272026Tuesday

23:30
Replit BlogOct 27 23:30
Replit Cloud Agents are now available to Pro and Enterprise users…
Replit announced that Cloud Agents are now available to all Pro and Enterprise customers. Non-engineers can describe changes in natural language, and the agent locates the code, runs the application in a cloud sandbox to verify it, and submits a PR or MR with screenshots and video to GitHub or GitLab. Teams can configure a shared environment once, and review feedback can be turned into skill updates that require engineer approval; the feature is in early access. The Slack integration is now available to all customers, and Microsoft Teams support is coming soon.
Replit announced that Cloud Agents are now available to all Pro and Enterprise customers. Non-engineers can describe changes in natural language, and the agent locates the code, runs the application in a cloud sandbox to verify it, and submits a PR or MR with screenshots and video to GitHub or GitLab. Teams can configure a shared environment once, and review feedback can be turned into skill updates that require engineer approval; the feature is in early access. The Slack integration is now avai
Matched fieldsBody
#company #Product updates #Agents #English text is being prepared.
21:55
a16z：NewsOct 27 21:55
a16z Long Read: Databricks and the Enterprise in the Agent Era
a16z partners Ben Horowitz and Michael McGuiness wrote the article, using Databricks' four generations of technological evolution—from Spark and Lakehouse to Lakebase and agents—as a thread to discuss how agents are changing enterprises. The article states that Databricks has surpassed $7 billion in annualized revenue, up 80% year over year, at a valuation of $190 billion, with more than 17 million databases launched by agents every day. The authors argue that enterprises need an agent-facing system of record that captures definitions, permissions, and evaluation results, and believe that in the future more people's work will shift toward specifying tasks, delegating execution, and checking results.
a16z partners Ben Horowitz and Michael McGuiness wrote the article, using Databricks' four generations of technological evolution—from Spark and Lakehouse to Lakebase and agents—as a thread to discuss how agents are changing enterprises. The article states that Databricks has surpassed $7 billion in annualized revenue, up 80% year over year, at a valuation of $190 billion, with more than 17 million databases launched by agents every day. The authors argue that enterprises need an agent-facing sy
Matched fieldsBody
#analysis #Industry news #Agents #English text is being prepared.
17:00
Official Pydantic AI articleOct 27 17:00
Pydantic Engineer Receives AAIF Outstanding Technical Contribution Award for Maintaining MCP Python SDK
On the second day of AGNTCon + MCPCon North America, the Agentic AI Foundation presented the Outstanding Technical Contribution Award to Pydantic engineer Marcelo Trylesinski, recognizing his work maintaining the MCP Python SDK. The v2 he led was released in beta alongside the 2026-07-28 version of the MCP specification, adding the MCPServer API and multi-turn round-trip requests, and has since moved to a stable release line with OpenTelemetry tracing enabled by default.
Matched fieldsBody
#company #Industry news #English text is being prepared.#Open-source ecosystem
08:00
Google Developers BlogOct 27 08:00
Google builds a multi-agent conversational commerce prototype on GCP with LangGraph and Gemini…
The Google Developers Blog showcased HappyDelivery, a multi-agent conversational commerce prototype built on LangGraph, Gemini, FastAPI, and Google Cloud, which implements the previously proposed three-layer decoupled architecture of Ingress, Memory, and Cognition. The prototype abstracts channels such as WhatsApp, RCS, Apple Messages for Business, and SMS using platform-specific signature verification and a unified message schema; in testing, the Ingress layer returned HTTP 200 in approximately 295–302ms and forwarded requests to Pub/Sub. The article also provides code examples for the state machine, deterministic financial tools, and an automated validation suite.
The Google Developers Blog showcased HappyDelivery, a multi-agent conversational commerce prototype built on LangGraph, Gemini, FastAPI, and Google Cloud, which implements the previously proposed three-layer decoupled architecture of Ingress, Memory, and Cognition. The prototype abstracts channels such as WhatsApp, RCS, Apple Messages for Business, and SMS using platform-specific signature verification and a unified message schema; in testing, the Ingress layer returned HTTP 200 in approximately
Matched fieldsBody
#engineering #Tutorials / practice #Agents #English text is being prepared.
08:00
Claude BlogOct 27 08:00
How to manage Claude Code costs on Google Cloud and prove ROI
Based on a recent webinar with Google Cloud, Anthropic presents a three-lever framework—control, track, prove—for managing Claude Code costs on Google Cloud and measuring return on investment. The Claude apps gateway supports setting daily/weekly/monthly spending limits by user, IdP group, or organization, alerts within the CLI when usage reaches 75% and 95%, and returns 429 once the cap is hit. On the Google Cloud side, native capabilities include per-project billing, budget alerts, Spend caps (preview), BigQuery export, and quotas.
Based on a recent webinar with Google Cloud, Anthropic presents a three-lever framework—control, track, prove—for managing Claude Code costs on Google Cloud and measuring return on investment. The Claude apps gateway supports setting daily/weekly/monthly spending limits by user, IdP group, or organization, alerts within the CLI when usage reaches 75% and 95%, and returns 429 once the cap is hit. On the Google Cloud side, native capabilities include per-project billing, budget alerts, Spend caps
Matched fieldsBody
#lab #Tutorials / practice #English text is being prepared.#Agents