Latest Artificial Intelligence News & Technical Dispatches

Open-weight models now match proprietary frontier on 87% of benchmarks. But the remaining 13%—complex multi-step reasoning, long-horizon tool calling, and adversarial robustness—still separates a $0.00 model from a $15.00 model. This benchmark audit across 22 models reveals exactly where open weights win, where they fail, and the hybrid strategy that gets you the best of both.

Groq closed a $650M Series D at $4.2B valuation as AI agent token consumption surges 340% year-over-year. The funding will scale LPU inference capacity to meet demand from agent builders needing deterministic, sub-50ms latency at $0.05/M tokens.

Benchmark Groq LPU SRAM against Cerebras Wafer-Scale Engine CS-4 on tokens per second, memory bandwidth, thermal budgets, and agent inference economics.

Analyze the Cerebras CS-4 and Nvidia Rubin inference speed war, comparing wafer-scale SRAM against modular HBM4 clusters for autonomous agent swarms.

Inspect the Cerebras CS-4 architecture revealed at Hot Chips 2026, featuring WSE-4 silicon specs, on-chip SRAM memory density, and low-latency agent inference.

Anthropic has fully restored Claude Mythos 5 and Fable 5 access after the US Commerce Department lifted export controls imposed on June 12, 2026. The 7-week disruption affected 340,000+ enterprise users and forced emergency migrations to alternative models.

The MCP 2026-07-28 specification eliminated session state, introduced Multi Round-Trip Requests (MRTR), and added header-based routing. Six months in, this analysis examines what actually changed — and what didn't — for production agent builders.

Nvidia posted $96.2B Q2 revenue (beating $92.2B estimate), $2.22 EPS, and guided $106-110B for Q3. Stock surged 8.7% in one day, adding $441.5B in market cap to reach $5.49 trillion.

Nvidia posted $96.2B Q2 revenue (beating $92.2B estimates) and warned hyperscalers of 15%+ price hikes on Vera Rubin and Blackwell systems in 2027. This analysis examines what rising GPU costs mean for AI agent builders and inference economics.

XPENG Robotics raised over $900 million at a $6.3 billion valuation — China's largest embodied AI funding round — backed by IDG, Tencent, and Alibaba. IRON humanoid robot enters mass production by end of 2026.

Moonshot AI's Kimi K3 2.8T open-weight model matches Claude Opus 5 on coding benchmarks at a fraction of the cost. This deep dive analyzes the head-to-head comparison across Terminal-Bench, SWE-bench, token economics, and production deployment trade-offs.

Skild AI released S1 on August 25 — a robotics foundation model that learns complex 10-minute tasks from a single human video demonstration with no fine-tuning, achieving 66% success on unseen tasks.

SemiAnalysis revealed OpenAI's Jalapeño chip — taped out with Broadcom in 16 months on TSMC N3P — hits 13.4 PFLOPs at 700W, beating Nvidia Rubin's 900-1,150W on perf-per-watt. This analysis examines the custom silicon race and its impact on inference economics.

The Netherlands fined Uber €825 million for automated driver deactivations without human review. This analysis examines what the ruling means for AI agent governance, automated decision-making regulation, and the compliance requirements for autonomous systems.

Z.ai's GLM-5.3-Flash became the most-discussed AI model on X and Reddit within 48 hours of its August 26 release. The 320B-parameter natively multimodal MoE runs at 18B active parameters under MIT license at $0.075/M input — 3x cheaper than DeepSeek V4-Flash.