Understand the signal before you apply.
AI releases, research, funding shifts and application guidance—checked against cited sources and connected to current global opportunities.
352 published insights · page 16 of 20
OpenAI Disrupts Russia-Origin Influence Operation Using ChatGPT for Promotion
OpenAI says it banned a cluster of ChatGPT accounts that very likely originated in Russia and used AI-generated social posts to promote a fabricated think-tank-style influence operation with...
Figure Launches Index, a Global Human-Video Pipeline for Humanoid Robot Training
Figure has taken Index out of stealth, revealing a global human-video collection pipeline for training its Helix humanoid AI system, with more than 16 million uploaded videos and 44,000 week...
CoreWeave and Anyscale Caption 600 TB of Video in 95 Minutes
CoreWeave and Anyscale report a 1,600-GPU multimodal data pipeline that captioned 600 TB of video in 95 minutes while sustaining 12,666 captions per second.
IBM Releases Granite 4.2 Reasoning Models and Granite Speech 5.0
IBM released Apache-2.0 Granite 4.2 reasoning LLMs in 3B, 8B and 30B sizes plus compact Granite Speech 5.0 Turbo CTC models for high-throughput ASR.
Argonne STREAMLINE Scales AI Nuclear Modeling to 100 Particles
Argonne says STREAMLINE uses neural-network quantum wave functions and supercomputing to scale nuclear many-body modeling from roughly a dozen particles toward about 100.
ChatGPT Site Tools Bring WebMCP Actions to the Desktop Browser
OpenAI's ChatGPT desktop app can discover and invoke WebMCP site tools in its built-in browser, letting supported websites expose structured actions to AI agents.
Apple PROOF-Gen Turns Failed Agent Traces Into Better Distillation Data
Apple researchers propose PROOF-Gen, a method that uses reflection and per-scenario prompt optimization to recover successful tool-calling trajectories from teacher-model failures.
Apple PROOF-Gen Recovers Failed Agent Trajectories for Better Distillation
Apple researchers introduce PROOF-Gen, a method that turns failed tool-calling runs into verified training trajectories and reports gains on benchmarks and a production agent pipeline.
AWS Backs Agentic Resource Discovery, an Open Specification for AI Agent Discovery
AWS detailed the Agentic Resource Discovery specification, an Apache-2.0 open standard intended to help agents discover tools, skills, MCP servers and other agentic resources across registri...
NVIDIA Says Vera Rubin Raises Agentic AI Throughput per Megawatt
NVIDIA published AgentX benchmark results claiming Vera Rubin NVL72 delivers major throughput-per-megawatt gains for production-style agentic AI workloads compared with prior generations.
NVIDIA Puts Groq 3 LPX Into Full Production for Agentic AI Inference
NVIDIA says Groq 3 LPX is now in full production as an interactive inference accelerator for Vera Rubin, targeting extremely fast token generation for long-context and multi-step AI agents.
AWS Adds Managed Ray Capabilities to SageMaker HyperPod
AWS added managed Ray support to SageMaker HyperPod on EKS, combining KubeRay-compatible distributed Python workloads with HyperPod health monitoring, recovery and SageMaker Studio integrati...
NVIDIA Details Spectrum-X Multiplane Ethernet for Giga-Scale AI Clusters
NVIDIA detailed Spectrum-X Multiplane networking for very large AI clusters, combining adaptive routing, congestion control and NIC-level plane balancing.
NVIDIA Details BlueField-4 Scale-In Infrastructure for Agentic AI Factories
NVIDIA introduced Scale-In as a fifth AI-networking pillar, using BlueField-4, DOCA and Spectrum-X to offload security, storage, telemetry and access services from host CPUs.
OpenAI Brings GPT-5.6 Sol, Terra and Luna to Kiro for Agentic Development
OpenAI and AWS added GPT-5.6 Sol, Terra and Luna to Kiro, pairing the model family with structured requirements, design and long-running software-development workflows.
NVIDIA Vera CPU Targets the CPU Bottlenecks Behind Agentic AI Fleets
NVIDIA argues that agentic AI infrastructure needs strong single-thread CPU performance as well as concurrency because orchestration, tools and sandboxed code create highly variable workload...
Liquid AI and Artificial Analysis Launch Pipette for On-Device AI Benchmarking
Pipette is an open-source benchmarking platform for edge AI deployments, launching with 1,000+ model, quantization, runtime, device and context configurations plus reproducibility protocols.
Intel Details Diamond Rapids, Crescent Island and Wildcat Lake for Agentic AI
Intel used Hot Chips 2026 to detail a 256-core Xeon architecture, a 480GB air-cooled inference GPU and an 18A client SoC aimed at agentic AI from datacenters to edge devices.
Turn intelligence into applications.
Create a free alert for the topics, roles or countries that matter. We email only new verified matches.
More ways to save
Discover deals, coupons and free courses on our sister site.
