Thursday, November 12
Today in AI & tech
Nvidia CEO Jensen Huang faces ouster over affairs; TSMC bets $1B on AI software
- Nvidia's board prepares to remove Jensen Huang following alleged affairs and national security concerns, triggering Congressional inquiry and stock drop.
- TSMC invests $1B in Anthropic and OpenAI simultaneously, signaling hardware-software consolidation as Nvidia leadership destabilizes.
- United Airlines bans unauthorized AI agents from booking flights, first major carrier to restrict autonomous purchasing—setting industry precedent.
- Martin Fowler's tests-as-data framework makes AI-generated code correctness measurable and verifiable at scale, addressing validation gaps.
- Indie studio's quiet AI bug fix sparks broader reckoning over disclosure, craft, and what counts as shipping your own code.
Critical RCE Vulnerability in Apache HTTP Server
A newly disclosed CVE enables remote code execution on Apache HTTP Server, requiring urgent patching across web infrastructure.
Nvidia CEO Jensen Huang's Job Suddenly on the Line
Nvidia's board is reportedly preparing to oust Jensen Huang following multiple alleged affairs—including one with an executive said to have Chinese intelligence ties—triggering a Congressional inquiry and stock drop.
United Airlines Bans AI Agents From Booking Flights
United becomes the first major carrier to explicitly prohibit unauthorized AI agents from purchasing tickets on behalf of travelers, raising immediate questions about where autonomous agents are permitted to act.
TSMC to Invest $1B in Anthropic and OpenAI
Taiwan's chipmaking giant enters the AI investment arms race the same week Nvidia's leadership is in crisis, signaling deep hardware-software convergence at the frontier.
AI Agents Are Testing the Limits of Zero-Trust Security
Autonomous agents are breaking assumptions baked into zero-trust architectures, forcing security teams to rethink identity and access models designed for humans.
More stories
View allTests as Data: Ensuring AI-Generated Code Does What We Want
Martin Fowler's team presents a framework using tests-as-data to validate AI-generated code output at scale, making correctness measurable rather than assumed.
Big Walk Used AI to Fix a Bug — Now Indie Gaming Is Having a Reckoning
An indie studio's quiet use of AI for a bug fix has sparked a wider debate about disclosure, craft, and what it means to ship "your own" code.
Lock Your GitHub Actions
Practical guidance on pinning Actions to commit SHAs to prevent supply chain attacks via mutable version tags.
We Built an AI Control Plane to Reduce LLM Expenses and Prevent Data Leaks
ThoughtWorks shares architecture for a centralized routing layer that cuts LLM costs and enforces data governance across teams.
The Data Center Backlash Might Really Just Be About Electricity
MIT Tech Review argues community opposition to AI data centers is fundamentally a grid-capacity and energy-cost issue, not an AI issue per se.
Long-Horizon Agents and Durable Execution
Martin Fowler explores how durable execution frameworks (Temporal, Restate) become essential when AI agents run tasks spanning hours or days.
How GitHub Used AI to Port Copilot From TypeScript to Rust in Record Time
ByteByteGo breaks down GitHub's AI-assisted migration—a real-world data point on AI's ROI for large-scale refactoring.
How LinkedIn Uses AI to Review 40,000 Pull Requests a Week
LinkedIn's AI-assisted PR review system surfaces patterns on what automated review can and can't catch at massive scale.
Git Security Vulnerabilities Announced
GitHub discloses a new batch of Git CVEs affecting credential handling and object parsing—upgrade paths detailed.
Customer Support Is the Gnarliest AI Problem
Stack Overflow Blog examines why customer support—despite seeming like a perfect AI fit—remains resistant to full automation due to edge cases and trust requirements.
Tools & repos to watch
vercel-labs/agent-browser
A browser automation tool built for AI agents to navigate and interact with the web programmatically.
github.com · Nov 12
homeassistant-ai/ha-mcp
An MCP server for Home Assistant, letting AI agents control smart home devices through the Model Context Protocol.
github.com · Nov 12
openai/codex-security
OpenAI's repo for security tooling around Codex, focused on safe code generation and vulnerability detection in agent workflows.
github.com · Nov 12
Nod by n8n
Open-source personal agent with its own virtual computer—runs workflows autonomously on your behalf without a SaaS dependency.
producthunt.com · Nov 10
Skipbox
Cheap, fast GitHub Actions runners with unlimited minutes, targeting teams hitting cost ceilings on CI/CD.
producthunt.com · Nov 10