https://www.beri.net/article/openai-astra-for-law-legal-research-bench-54-percent-vs-general-models-bake-off THE D[AI]LY BRIEF en 2026-09-22T20:49:13.120Z Astra for Law Scored 54% on a Set Vals Never Publishes https://www.beri.net/article/cognex-realsense-acquisition-facial-authentication-spin-out-license-renewal THE D[AI]LY BRIEF en 2026-09-22T16:47:11.122Z Cognex Buys RealSense, Leaving Face ID to a 25-Person Spin-Out https://www.beri.net/article/deepgram-vs-assemblyai-vs-whisper-self-hosting-crossover-call-center-2026 THE D[AI]LY BRIEF en 2026-09-22T12:53:40.029Z Deepgram vs AssemblyAI vs Whisper: If You Self-Host, Skip Whisper https://www.beri.net/article/grok-4-7-same-price-cost-per-task-matched-reasoning-effort THE D[AI]LY BRIEF en 2026-09-22T08:47:41.009Z Grok 4.7 Kept Its $2/$6 Price and Can Still Double Cost per Task https://www.beri.net/article/blackline-netnow-acquisition-credit-onboarding-data-export-erp-connector-reg-b-reasons THE D[AI]LY BRIEF en 2026-09-22T04:42:44.354Z BlackLine Buys NetNow, Whose Terms Give Credit Teams No Export Right https://www.beri.net/article/harvey-negative-gross-margin-open-weight-pivot-agentic-ai-seat-pricing-renewal THE D[AI]LY BRIEF en 2026-09-22T00:34:26.072Z Harvey Swapped Models After Agents Sank Its Margin to -50% https://www.beri.net/article/microchip-hailo-acquisition-closed-product-longevity-pledge-dataflow-compiler-license THE D[AI]LY BRIEF en 2026-09-21T20:25:17.943Z Microchip's No-EOL Pledge Stops Short of Hailo's Compiler https://www.beri.net/article/runai-vs-kueue-vs-slurm-kubernetes-gpu-scheduling-2026 THE D[AI]LY BRIEF en 2026-09-21T16:26:39.309Z Run:ai vs Kubernetes vs Slurm: Static Quotas Strand Your GPUs https://www.beri.net/article/amazon-blocks-meta-muse-cloud-vm-agent-ninth-circuit-perplexity-ruling-gap THE D[AI]LY BRIEF en 2026-09-21T12:22:51.156Z Amazon Blocks Meta's Muse Agent Where the Perplexity Ruling Stops https://www.beri.net/article/llm-api-token-inflation-black-box-audit-7-of-15-gpt-endpoints-flagged THE D[AI]LY BRIEF en 2026-09-21T08:13:22.916Z A 40-Call Audit Flagged 7 of 15 GPT Endpoints for Token Inflation https://www.beri.net/article/typesafe-jev-typed-decision-model-calibration-decomposition-shadow-eval THE D[AI]LY BRIEF en 2026-09-21T04:06:24.418Z TypeSafe's Jev Scores 62.6% Asked Once and 95% Split Five Ways