https://forck.live/items/24728-germany-wants-to-lead-the-world-in-industrial-ai-how-nvidia-will-help forck.live en 2026-10-19T19:27:08+00:00 Germany Wants to Lead the World in Industrial AI: How NVIDIA Will Help https://forck.live/items/24719-cutting-our-inference-bill-by-68-by-being-obsessive-about-optimization forck.live en 2026-10-19T19:12:53+00:00 Cutting our inference bill by 68% by being obsessive about optimization https://forck.live/items/24706-introducing-flux-3-max forck.live en 2026-10-19T18:51:34+00:00 Introducing FLUX 3 Max. https://forck.live/items/24702-runway-mcp-is-now-in-grok-bot forck.live en 2026-10-19T18:21:31+00:00 Runway MCP is now in Grok Bot. https://forck.live/items/24693-introducing-clairic-a-new-system-one-class-of-extraction-models forck.live en 2026-10-19T17:48:26+00:00 Introducing Clairic, a new System One class of extraction models https://forck.live/items/24672-low-medium-high-and-xhigh-presets-use-fast-search forck.live en 2026-10-19T17:25:02+00:00 Low, medium, high, and xhigh presets use Fast Search https://forck.live/items/24671-xhigh-preset-output-budget-raised-to-512-000-tokens forck.live en 2026-10-19T17:25:02+00:00 xhigh preset output budget raised to 512,000 tokens https://forck.live/items/24670-tool-approvals-for-mcp-and-connector-calls forck.live en 2026-10-19T17:21:52+00:00 Tool approvals for MCP and connector calls https://forck.live/items/24663-pegasus-1-5-from-twelvelabs-is-now-available-in-amazon-bedrock forck.live en 2026-10-19T16:49:19+00:00 Pegasus 1.5 from TwelveLabs is now available in Amazon Bedrock https://forck.live/items/24664-how-deepgram-optimized-voice-ai-at-scale-with-aws-trainium forck.live en 2026-10-19T16:23:22+00:00 How Deepgram optimized voice AI at scale with AWS Trainium https://forck.live/items/24640-push-based-kv-transfer-to-improve-ttft-and-tpot-for-disaggregated-inference-in forck.live en 2026-10-19T16:05:17+00:00 Push-based KV transfer to improve TTFT and TPOT for disaggregated inference in vLLM https://forck.live/items/24641-improve-ttft-for-multi-turn-agentic-workloads-with-bi-directional-kv-transfer forck.live en 2026-10-19T16:03:23+00:00 Improve TTFT for multi-turn agentic workloads with bi-directional KV transfer in vLLM https://forck.live/items/24629-genrecs-tokenization-transforming-ads-recommendation-at-meta-with-stable-model forck.live en 2026-10-19T16:00:47+00:00 GenRecs Tokenization: Transforming Ads Recommendation at Meta with Stable Model Vocabulary https://forck.live/items/24654-improving-agent-accuracy-with-harvey-graph forck.live en 2026-10-19T16:00:00+00:00 Improving Agent Accuracy With Harvey Graph https://forck.live/items/24647-who-judges-the-judge-three-helpful-ways-to-audit-your-llm-evaluators forck.live en 2026-10-19T16:00:00+00:00 Who judges the judge? Three helpful ways to audit your LLM evaluators https://forck.live/items/24622-the-thin-orchestrator-how-we-made-our-ai-agents-more-reliable-by-giving-the forck.live en 2026-10-19T15:30:37+00:00 The Thin Orchestrator: How We Made Our AI Agents More Reliable by Giving the Router Less to Do https://forck.live/items/24623-nvidia-advances-agentic-security-for-media-from-european-public-service forck.live en 2026-10-19T15:00:31+00:00 NVIDIA Advances Agentic Security for Media, From European Public Service Broadcasting to Global Content Production https://forck.live/items/24608-introducing-reason forck.live en 2026-10-19T13:34:35+00:00 Introducing: Reason https://forck.live/items/24574-announcing-torchneuron-pytorch-code-now-runs-natively-on-aws-trainium-chips forck.live en 2026-10-19T13:02:20+00:00 Announcing TorchNeuron: @PyTorch code now runs natively on AWS Trainium chips with a single configuration change. No rewrite needed. https://forck.live/items/24570-inside-prometheus-meta-s-journey-to-a-gigawatt-scale-supercluster forck.live en 2026-10-19T13:00:10+00:00 Inside Prometheus: Meta’s Journey to a Gigawatt-Scale Supercluster https://forck.live/items/24700-announcing-the-chroma-awards-season-2 forck.live en 2026-10-19T12:00:00+00:00 Announcing the Chroma Awards Season 2 https://forck.live/items/24521-your-team-s-ai-can-now-live-on-more-than-one-machine-and-every-person-gets forck.live en 2026-10-19T09:37:17+00:00 Your team's AI can now live on more than one machine, and every person gets their own. https://forck.live/items/24648-google-named-a-leader-in-the-2026-idc-marketscape-for-agentic-ai-platforms forck.live en 2026-10-19T07:00:00+00:00 Google named a Leader in the 2026 IDC MarketScape for Agentic AI Platforms https://forck.live/items/24751-assign-it-to-amp forck.live en 2026-10-19 Assign It To Amp https://forck.live/items/24704-torchneuron-running-hf-models-on-aws-trainium-is-easier-than-ever forck.live en 2026-10-19 TorchNeuron: running HF models on AWS Trainium is easier than ever https://forck.live/items/24698-together-post-training-frontier-grade-model-shaping-the-way-that-fits-you forck.live en 2026-10-19 Together Post-Training: frontier-grade model shaping, the way that fits you https://forck.live/items/24653-diffusion-at-scale-introducing-mercury-3 forck.live en 2026-10-19 Diffusion at Scale: Introducing Mercury 3 https://forck.live/items/24610-factory-and-databricks-building-software-powered-by-data forck.live en 2026-10-19 Factory and Databricks: Building Software Powered by Data https://forck.live/items/24535-gradium-is-now-available-on-eden-ai forck.live en 2026-10-19 Gradium is now available on Eden AI https://forck.live/items/24534-minimax-m3-1-flash-redefining-agent-performance-and-cost forck.live en 2026-10-19 MiniMax M3.1 Flash: Redefining Agent Performance and Cost https://forck.live/items/24404-gpt-6-astra-sensenova-roborsi forck.live en 2026-10-18T16:00:52+00:00 商汤实测 GPT-6-Astra 并开源 SenseNova-RoboRSI：具身智能下一步怎么走？ https://forck.live/items/24204-investing-in-the-public-benefit forck.live en 2026-10-18 Investing in the public benefit