Newsletter Archive

Saturday, April 11, 2026

CoreWeave stock pops 11% on deal to power Anthropic's Claude · ChatGPT voice mode is a weaker model · Trump’s AI Chip Export Push Stymied by Bureaucratic Bottleneck                                                                                
AI Pulse AI Pulse Daily Saturday, April 11, 2026
CoreWeave stock pops 11% on deal to power Anthropic's Claude

CoreWeave signed a deal to power Anthropic's Claude inference, making Anthropic its ninth major AI model provider customer. The announcement follows Meta's $21 billion infrastructure commitment to CoreWeave, underscoring the GPU shortage and infrastructure bottleneck driving AI deployment.

CNBC Tech · 1 min read Industry

ChatGPT voice mode is a weaker model

ChatGPT's voice mode runs on GPT-4o (April 2024 cutoff), not the latest frontier model, despite users expecting the smartest available AI. Willison highlights the perception gap when different interfaces access different capability tiers.

Simon Willison · 2 min read Community

Trump’s AI Chip Export Push Stymied by Bureaucratic Bottleneck

Trump's push to expand US AI chip sales globally is hitting bureaucratic roadblocks at the export licensing agency: staffing shortages, licensing delays, and unclear policy direction are hampering multi-billion-dollar export approvals.

Bloomberg Tech · 1 min read Policy

DeepSeek Looks for Data Center Engineers in Inner Mongolia

DeepSeek is hiring data center engineers in Inner Mongolia as it scales operations using Nvidia Blackwell chips—hardware officially banned from export to China. The move underscores how Chinese AI labs are circumventing U.S. sanctions to build frontier compute capacity.

Bloomberg Tech · 1 min read Industry

NVIDIA Releases AITune: An Open-Source Inference Toolkit That Automatically Finds the Fastest Inference Backend for Any PyTorch Model

NVIDIA open-sourced AITune, a Python toolkit that automatically benchmarks and selects the fastest inference backend (TensorRT, Torch Inductor, TorchAO) for any PyTorch model without manual rewriting. Handles Computer Vision, NLP, and generative AI workloads on NVIDIA GPUs.

MarkTechPost · 6 min read Tools

Harvard’s Kreiman Seeks $100 Million to Build AI Memory Tech

Harvard neuroscientist Gabriel Kreiman's new AI lab is raising ~$100M to develop technology enabling perfect human memory recall. The venture, spun from Harvard, targets a sci-fi premise: offloading memory constraints.

Bloomberg Tech · 1 min read Market

A Coding Guide to Markerless 3D Human Kinematics with Pose2Sim, RTMPose, and OpenSim

Step-by-step Colab tutorial for building a complete markerless 3D human kinematics pipeline: 2D pose estimation via RTMPose, multi-view triangulation, filtering, and biomechanical analysis with OpenSim. Covers environment setup, calibration, synchronization, and visualization of motion data from multi-camera video.

MarkTechPost · 11 min read Tools

ChatGPT rolls out new $100 Pro subscription to challenge Claude

OpenAI launched a $100/month Pro tier to match Claude's pricing, escalating the subscription war between frontier model providers.

BleepingComputer · 2 min read Industry

Quick Hits
Amazon takes a jab at Nvidia over chips shift - Semafor Semafor Hardware
AI demand is so high, AWS customers are trying to buy out its entire capacity - Network World Network World Industry
Japan Bets $16 Billion to Propel Startup Rapidus Into AI Chips Bloomberg Tech Hardware
How Knowledge Distillation Compresses Ensemble Intelligence into a Single Deployable AI Model MarkTechPost Tools
Health AI and the law: Could your chatbot doc testify against you? Mashable Policy
Tests show $30,000 AI GPUs are terrible password crackers — RTX 5090 gaming GPU outperforms Nvidia H200 and AMD MI300X Tom's Hardware Hardware
Chinese Nvidia Cloud Partner procured 300 servers with banned AI GPUs worth $92 million — shares of data center supplier Sharetronic plummet following Super Micro smuggling arrest Tom's Hardware Policy

Share AI Pulse Daily

Post on X Share on LinkedIn

Got feedback? Just hit reply — we read every response.

You're receiving this because you subscribed to AI Pulse Daily.

Visit AI Pulse  ·  Manage preferences  ·  Unsubscribe

← Back to AI Pulse