Loading...
On April 14, 2026, GitHub made Pull Requests optional for the first time. The 21-year workflow built for human collaboration is being made optional — because the contributors arriving are not human.
Cursor open-sourced Mixture-of-Kittens (MoK), a fused MoE execution kernel that pushed multi-node signaling latency from 103us to 18us and lifted 512-GPU training throughput 41%. Why an IDE company rewrote the GPU kernel — and what it means for the rest of the AI stack.
On August 13, 2026, DeepSeek released its Harness framework as v0.1.0-rc with no compatibility guarantee — and within 24 hours the community had shipped 600+ plugins. This is a strategy-level read on why the 'unfinished' version is the boldest move in the Agent race, what Cordis (the runtime underneath) actually enables, and how the platform-vs-product question just shifted.
Flue 2 (Cloudflare / Fred Schott) ships React-style agent hooks as the first stable harness-first framework. Why this is the React moment for agent composition.
Google shipped Gemini 3.7 Flash on Aug 13, 2026 — just 21 days after 3.6 Flash — at half the per-token price, with near-2x benchmark gains on the workloads that actually drive agent economics.
How Cloudflare built Kitesurf — an agent-first browser on Rust + Wasm + V8 isolates — in twelve weeks, with AI agents writing the WPT conformance tests. The architecture, the build story, and why it changes the cost of every agent you run.
AWS released AgentCore Browser Tool on Aug 13, 2026 — a managed Chromium + vision-driven agent hosting surface for legacy HTML web apps. Here's why this is the first credible replacement for rules-based RPA.
Floatboat's 2026-08 benchmark reveals the same DeepSeek-V4-Flash beats Claude Opus 4.8 on all 5 third-party agent benchmarks at 1/57 the cost — when wrapped in the right Harness. The Harness gain scales with task length: 1.9% on short tasks, 23.6% on long-horizon work. The conclusion is clear: for real Agent products, the bottleneck isn't the model, it's the system around it.
Microsoft Research ships MindTopo, a benchmark that pinpoints why VLMs ace static vision but fail at interactive planning — and what the engineering fix looks like.