Monday, April 13, 2026
Meta ditches open source (wait, what?)
Meta just dropped Muse Spark—their first proprietary model after years of open-source evangelism (bold move), while Anthropic got officially blacklisted by the Pentagon after losing their appeal. Meanwhile, OpenAI's enterprise business is printing money at 40% of revenue, and PyTorch is turning supercomputers into programmable APIs with Monarch. So Meta's going closed—would you?
Top Stories
PyTorch
Monarch is PyTorch's distributed programming framework that makes massive cluster training feel like local development through a simple API, now with native Kubernetes support, improved RDMA networking, and agent-optimized features like distributed SQL telemetry. Recent updates include 100x smaller install size, AWS EFA and AMD ROCm support, and integrations with SkyPilot and VERL for easier deployment across cloud providers.
CNBC
Meta debuts Muse Spark, its first proprietary AI model developed under newly acquired executive Alexandr Wang, marking a strategic pivot from open-source to compete with OpenAI and Google while introducing paid API access for developers.
OpenAI's enterprise revenue has grown to over 40% of total revenue and is projected to match consumer revenue by late 2026, as companies move beyond AI experimentation to deploying production-ready agent systems across their organizations using OpenAI Frontier infrastructure and partnerships with major consulting firms and cloud providers.
Bridgy Fed now bridges Mastodon's ActivityPub and Bluesky's AT Protocol, enabling cross-network interaction between previously siloed decentralized social platforms. This interoperability could help decentralized networks compete more effectively against centralized platforms by allowing users across different protocols to communicate.
CNBC
An appeals court upheld the Pentagon's blacklisting of Anthropic from defense contracts while litigation continues, though a separate ruling allows the AI company to work with other federal agencies. The unprecedented designation of an American AI company as a supply chain risk stems from disagreements over autonomous weapons use and surveillance safeguards.
Keep Reading
Industry Voices
Enjoyed this issue?
Get daily AI intel delivered to your inbox. No fluff, just the stories that matter.