Apple Siri AI & Google Gemini CLI: The Week AI Went From Touchscreen to Terminal

Two very different visions of AI's future landed this week — and both are worth paying attention to.

🍎 Apple's Siri AI — Redemption at Last

At WWDC 2026, Apple finally did what everyone's been waiting for: it gave Siri a complete personality transplant.

The new Siri AI (yes, it's now a standalone app on iOS 27) is conversational, expressive, and far more capable than its robotic predecessor. Key highlights:

• Cross-app context: Siri now understands what you're doing across multiple apps — no more "Siri, open Spotify" → "I don't know what you're talking about."
• Apple Intelligence v2: On-device processing for summaries, photo editing, and tab management in Safari.
• Siri AI as an app: You can actually talk to Siri as a standalone assistant, not just a springboard into other apps.

Bloomberg called it "Apple's AI redemption tour" — and after two years of watching OpenAI, Google, and Anthropic steamroll the assistant space, Cupertino is finally playing catch-up.

🌐 Google Gemini CLI — Open-Source Terminal Agent

While Apple polished the touchscreen, Google went the other direction — straight into the terminal.

Gemini CLI (released as open-source on GitHub) is a ReAct-loop AI agent that brings Gemini 2.5 Pro and Flash directly into your command line. It's:

• Free (just needs a Google account)
• Open-source (MIT-licensed on GitHub)
• Customizable — with subagent support, tool-calling, and file-system awareness

Think of it as Google's answer to Claude Code and GitHub Copilot — but with one twist: it's not locked to any IDE. Just gemini-cli in any terminal, and you're coding with AI.

"The most direct path from your prompt to our model." — Google

🔮 Two Philosophies, One Week

Both approaches are valid — and both signal that June 2026 is the month AI stopped being a feature and started being the platform.

🏷️ Tags

ai opensource apple google

What do you think — terminal agent or touchscreen assistant? Let me know in the comments.