Foldable iPhones, AI Model Drama, and a World Beyond Apps

Hey folks,

It’s Monday, so let’s quickly catch up on the some of the biggest and most interesting AI updates from the past few days.

Apple’s Foldable iPhone May Arrive in December

Apple could finally launch its long-rumored foldable iPhone by December 2025, according to analyst reports based on supply chain data. This would place the device a few months after the expected September launch of the iPhone 18 Pro models.

There are also signs Apple may split its iPhone lineup into two phases, with the base iPhone 18 potentially launching in March 2026. If true, this would mark a shift in Apple’s traditional release strategy.

Cursor’s New AI Model Sparks Controversy

AI coding startup Cursor launched its new Composer 2 model, claiming advanced coding capabilities. However, users quickly pointed out similarities with Kimi 2.5, an open-source model from Moonshot AI.

Cursor later confirmed that Composer 2 was built on an open-source base, with additional training layered on top. The situation highlights growing questions around transparency and attribution in AI model development.

Apps Will Disappear,” Says Nothing CEO

Carl Pei, CEO of Nothing, believes apps may disappear as AI agents take over smartphone interactions. Instead of manually using apps, users could rely on AI agents to perform tasks in the background. Pei suggests companies should prepare by building systems that AI agents can interact with directly, rather than relying on traditional app interfaces.

Samsung Brings AirDrop-Like Sharing to Galaxy S26

Samsung has added support for AirDrop-style file sharing to its Galaxy S26 series through Quick Share. This allows easier file transfers between Android and Apple devices. The feature, initially introduced by Google on Pixel devices, is now expanding across more Android smartphones, improving cross-platform compatibility.

Scrape Websites. Get Clean AI-Ready Data

Geekflare API lets you scrape any website and instantly convert it into clean, structured data for your AI workflows. Just enter a URL, and it returns the content in Markdown, JSON, or HTML. You can also capture screenshots and extract LLM-ready data without worrying about blocking, CAPTCHAs, or messy page code. It’s a simple way to collect reliable data for RAG, AI agents, and automation.

How to Extract Links from Any Website for Free?

Cheers,

Keval, Editor