
Microsoft's Phi-4 Vision Matches Models 10x Its Size
Microsoft releases Phi-4-reasoning-vision-15B - a 15B open-weight multimodal model trained on 240 GPUs in 4 days that competes with 100B+ parameter models on math, science, and GUI understanding.
They summarize our coverage. We write it.
Newsletters like this one rebroadcast our headlines - often without the full review, the source reading, or the analysis underneath. Our weekly briefing sends the work they paraphrase, straight from the desk, before they get to it.
Free, weekly, no spam. One email every Tuesday. Unsubscribe anytime.

Microsoft releases Phi-4-reasoning-vision-15B - a 15B open-weight multimodal model trained on 240 GPUs in 4 days that competes with 100B+ parameter models on math, science, and GUI understanding.

OpenAI is developing an internal code repository to replace GitHub, putting the company on a collision course with its biggest backer.

VS Code 1.110 ships native browser control for AI agents, installable agent plugins with MCP support, persistent session memory, and a new Agent Debug panel.

OpenAI's Codex desktop app hits the Microsoft Store with native Windows sandboxing, PowerShell agents, WinUI skills, and multi-agent parallel execution - no WSL required.

A 439,000-worker construction shortage is delaying AI data centers while electricians command $200K salaries and Big Tech scrambles to fill the gap.

A hands-on review of GitHub Copilot in 2026 - from agent mode and the new CLI to multi-model support and the five-tier pricing system, testing whether the market leader still deserves its crown.

Amazon, Google, Meta, Microsoft, xAI, Oracle, and OpenAI will meet at the White House on March 4 to sign a voluntary pledge to build, bring, or buy their own electricity supply for AI data centers - but the agreement has no enforcement mechanism.

A detailed comparison of Moonshot AI's 1T-parameter Kimi K2.5 against Microsoft's 14B Phi-4 - the most extreme size gap in frontier AI, with 71x the parameters but vastly different use cases.

Microsoft's 14B dense transformer that consistently beats models 5x its size on MATH and GPQA, available under the MIT license for unrestricted commercial use.

A data-driven comparison of Alibaba's Qwen3.5-27B and Microsoft's Phi-4 - a 27B hybrid architecture versus a 14B STEM specialist, testing whether raw parameter count or training efficiency wins in practice.

Vercel releases the Chat SDK, a TypeScript library installable via 'npm i chat' that lets developers write chatbot logic once and deploy to Slack, Microsoft Teams, Google Chat, Discord, GitHub, and Linear. MIT licensed, AI-provider agnostic, now in public beta.

Bridgewater Associates warns AI capex has entered a 'dangerous phase' as Alphabet, Amazon, Meta, and Microsoft commit $650 billion to infrastructure in 2026, up 67% from last year.