Dev Weekly: Coinbase Fires Engineers, GitHub Gets Agents Panel, and Intel Gets a Government Lifeline

This week delivered some jaw-dropping stories that had developers debating everything from AI adoption policies to the future of chip manufacturing. From a CEO’s controversial firing decisions to government bailouts, here’s what dominated our feeds:

🔥 Coinbase CEO fires engineers for not using AI - 🌐 Brian Armstrong made waves by explaining why he let go of engineers who didn’t immediately try AI tools, calling it essential for staying competitive.

🤖 GitHub launches agents panel - 🌐 You can now delegate coding tasks to Copilot from anywhere on GitHub and track progress as it works in the background.

💰 US government takes 10% stake in Intel - 🌐 After SoftBank’s $2B investment, the government stepped in with its own rescue package for the struggling chip giant.

🏢 Corporate Shakeups & Big Money Moves

The Coinbase AI Ultimatum

Brian Armstrong sparked heated debate this week when he defended firing engineers who didn’t immediately try AI tools. His logic? In a fast-moving industry, you can’t wait for people to come around to game-changing technology.

This isn’t just about one company’s policy - it reflects the growing pressure developers face to adapt to AI tools. Whether you agree with the approach or not, it’s clear that AI adoption is becoming a career requirement, not just a nice-to-have skill.

The takeaway: Companies are getting serious about AI adoption, and waiting on the sidelines might not be an option much longer.

Intel Gets Multiple Lifelines

Intel had quite the week. First, SoftBank threw them a $2B lifeline, then the US government announced plans for a 10% stake. This isn’t just about one company - it’s about national security and maintaining domestic chip manufacturing capabilities.

For developers, this matters because Intel’s struggles have ripple effects across the entire hardware ecosystem. A stronger Intel means more competition with NVIDIA and potentially better options for both traditional computing and AI workloads.

The M&A Frenzy Continues

Enterprise software companies went on a buying spree this week, reshaping the developer tools landscape:

SAP completes LeanIX acquisition - This purchase strengthens SAP’s process intelligence capabilities, giving enterprises better visibility into their business workflows and optimization opportunities.

Zoom acquires AI translation startup - The move enhances Zoom’s real-time meeting capabilities with AI-powered translation, making global collaboration more seamless.

IBM buys Apptio for $4.5B - IBM’s biggest acquisition focuses on cloud cost management, helping enterprises optimize their growing cloud spending as workloads migrate to the cloud.

Databricks completes MosaicML acquisition - This deal strengthens Databricks’ position in the generative AI space, giving them better tools for training and deploying large language models.

These acquisitions show how companies are doubling down on AI capabilities and cloud optimization - two areas where developer expertise is becoming increasingly valuable.

Apple Prepares for Enterprise AI

Apple quietly rolled out new ChatGPT configuration options for enterprise customers, signaling their serious push into business AI. This move positions them to compete with Microsoft’s dominance in enterprise AI tools.

The enterprise focus suggests Apple sees AI as more than just a consumer feature - they want a piece of the business productivity market that’s driving so much AI investment.

🛠️ Developer Tools & Platforms

GitHub’s Agent Revolution Continues

The new agents panel lets you delegate coding tasks to Copilot from anywhere on GitHub. Think of it as having a persistent AI pair programmer that can work on tasks while you focus on other things.

Key features:

Launch tasks from any GitHub page
Track progress in real-time
Get tagged for review when complete
Works across repositories and organizations

This builds on GitHub’s vision of AI agents that don’t just suggest code but actually complete entire features. It’s another step toward the “agentic workflows” that everyone’s talking about.

MCP Servers: The New Developer Superpower

Model Context Protocol servers are becoming the secret sauce for extending AI capabilities. GitHub published a detailed guide on building your first MCP server, complete with a turn-based game example.

Why this matters: MCP servers let you give AI tools custom capabilities specific to your domain. Instead of general-purpose AI, you can create specialized assistants that understand your codebase, APIs, and workflows.

The barrier to entry is surprisingly low - you can build a working MCP server in an afternoon and suddenly have AI tools that know exactly how your systems work.

⚡ Infrastructure & Cloud

Meta’s Data Center Power Play

Meta got approval for gas power plants to support their $10B data center, despite environmental concerns. This highlights the massive energy demands of AI infrastructure and the hard choices companies face.

The decision also shows how AI training and inference workloads are reshaping energy policy. Every major tech company is scrambling to secure reliable power for their AI ambitions.

Nvidia’s H20 Chip Drama

Nvidia reportedly halted production on H20 AI chips amid regulatory uncertainty. The Commerce Department backlog is creating bottlenecks for companies trying to get AI hardware approved.

This regulatory friction is pushing companies to diversify their AI infrastructure strategies and consider alternative chip providers. It’s also driving more investment in domestic chip manufacturing.

🧠 AI & Machine Learning

Anthropic Bundles Claude Code

Anthropic integrated Claude Code directly into their enterprise plans, making advanced code generation available to business customers without additional setup.

What’s included:

Multi-language code generation and debugging
Codebase-aware suggestions
Enterprise-grade security and compliance
Integration with existing development workflows

This move puts pressure on OpenAI and GitHub Copilot to improve their enterprise offerings. The AI coding assistant market is heating up as companies realize how much developer productivity these tools can unlock.

The AI Development Workflow Evolution

Between GitHub’s agent panel, Claude Code integration, and MCP servers, we’re seeing the emergence of truly agentic development workflows. AI isn’t just suggesting code anymore - it’s completing entire features, managing tasks, and integrating with your existing tools.

The question isn’t whether AI will change how we code, but how quickly we’ll adapt to these new collaborative workflows.

📊 Platform Updates

Git 2.51: Cruft-Free Repository Management

The latest Git release introduces cruft-free MIDXs (multi-pack-index), significantly improving repository management and performance. This update tackles one of Git’s long-standing issues with repository bloat.

Key improvements:

Better performance for large repositories with cruft-free MIDXs
Improved merge conflict resolution
Enhanced security features
Streamlined command-line experience

For developers working with large codebases, this update should provide noticeable performance improvements during common operations like fetch and clone.

Browser Wars: Firefox and Translation

Firefox 142 shipped with enhanced client-side translation features, pushing back against Chrome’s dominance in web browsing. The focus on privacy-first translation keeps your data local instead of sending it to cloud services.

Open Source Tool Updates

The open source ecosystem saw several significant releases:

LibreOffice 25.8 - New productivity features for office workflows
Thunderbird 142 - PDF signing capabilities and enhanced security
HandBrake 1.10.1 - Better video encoding compatibility
Ventoy 1.1.06 - Improved bootable USB creation
PowerDNS Server 5.0 - Enhanced DNS server performance
Calibre 8.9 - Better ebook management features
NetBeans 27 - IDE improvements for Java developers

Stack Overflow Chat Renewal

Stack Overflow announced a major revamp of their chat system, focusing on real-time collaboration and community building. The new chat aims to bridge the gap between Q&A and ongoing developer discussions.

This update reflects Stack Overflow’s broader effort to remain relevant in an era where developers increasingly turn to Discord, Slack, and AI tools for help.