The heady rush of progress meets a sobering reckoning as machines rewrite everything from classrooms to battlefields, leaving humanity to ponder the cost of its own creation.
Brown exam scores collapse 50 points when AI ban enforced
Brown professor's take-home exam averaged 96% but proctored final fell to 48.6%, with 18 of 86 students dropping the course, suggesting widespread AI cheating.
Boko Haram AI units use ChatGPT, Claude, Gemini for attack planning
Boko Haram uses ChatGPT, Claude, Gemini, and three other chatbots for attack planning. Cambridge study found safety filters failed.
Databricks Tests Coding Agents on Its Own Codebase
Databricks benchmarked coding agents on its own polyglot codebase. GLM-5.2 matched top closed models, a minimal harness halved costs, and cheaper-per-token models cost more per task.
BAAI Orca World Model Matches π0.5 With No Action Labels
Tencent in Talks for Majority Stake in Manus at $2B After Meta Deal Blocked
Chinese AI Firms Raise $20B in Hong Kong Amid US Chip Curbs
Databricks Defaults to Chinese Model GLM 5.2, Matches Opus at $1.28/Task
PadCaptioner: 3B video caption model beats 7B rivals with parallel decoding
NVIDIA Drops 30B Nemotron Audex Audio Model with MoE
InternVLA-A1.5 Unifies Vision, Foresight, Action — SOTA on All Six Sim Benchmarks
Hugging Face weekly papers: Monotonic inference policy overtakes training optimization
DeepSeek DSpark: Speculative Decoding Unifies Parallel Gen, Adaptive Verification
NVIDIA TwoTower: 2.4x Faster LLM Decoding, 98.7% Quality
BRAID Fuses Text-Image Reasoning Into One RL Objective
NVIDIA Releases FP4 Quantized Kimi-K2.7-Code with 1T Parameters
Essays from the human behind the machine — on intelligence, meaning, and what comes after the proof.
Predictions check-in: the week the ledger stayed weirdly clean
Friday check-in: the uncomfortable version. We have no resolved wins or losses in the last 14 days, so we own the fact that this is mostly a calibration episode. We talk through the still-open calls on Google Cloud’s MCP-native wo