Robots > chips? 🤖

The AI copyright debate intensifies

June 27, 2025

Welcome back! Meta’s AI strategy is zeroing in on voice. According to Bloomberg, Meta is reportedly in talks to acquire Play AI, a startup that lets anyone clone voices for AI-powered apps. Think chatbots with voices, personalized narrators, or automated customer service that sounds human.

If Meta gets its way, your next conversation with AI might sound a lot more like…you.

Nvidia’s Next Frontier? Robots

CNBC

Fresh off a trillion-dollar ascent, Nvidia is eyeing its next act: robotics. Nvidia’s Isaac platform is already powering robots in warehouses and factories. Now, the company wants to scale it, with CEO Jensen Huang calling robotics Nvidia’s “next big business.”

A glimpse into the future: Imagine real-world GPTs with camera eyes and arms. Huang envisions a future where AI agents learn in digital simulations and then operate physical robots in sectors from logistics to eldercare.

Why it matters: The smartest tech companies are building more than the software. They’re building the machines that run it.

Google Launches Gemini CLI

Google is rolling out a new way to code with AI, and it doesn’t live in your browser. Gemini CLI is a command-line interface tool that brings Gemini 1.5’s AI coding capabilities into your local terminal environment.

A new kind of coding assistant: Unlike ChatGPT’s native IDE or GitHub Copilot, Gemini CLI is open source and customizable.

  • Developers can generate functions, ask for shell commands, debug errors, and get explanations—all without leaving the terminal.

  • Also: It’s another sign that Google wants to build trust by putting more control in the hands of users.

Why it matters: From Gemini CLI to Microsoft’s Copilot stack, the battle for developer mindshare is heating up. Google’s pitch? You don’t need to change your workflow, just upgrade it.

Courts Send Mixed Signals on AI and Copyright

Is training AI on copyrighted material fair use? Two high-profile lawsuits are starting to draw the lines, and the answer isn’t so simple.

Anthropic wins, Stability AI holds the line. In a major win for AI companies, a judge recently ruled that Anthropic did not breach copyright law by training its models on books scraped from the web. The decision backed the argument that large-scale AI training can count as fair use under US law.

Also: Getty has dropped key copyright claims against Stability AI, narrowing its UK suit to just one legal question—whether Stability’s scraped image data is protected by database rights.

Why it matters: The tension between creators and AI developers is growing. These cases can shape the playbook, and so far, the courts are leaning more permissive than protective.

Prompt your way to production

Pheonix.new

Phoenix.new lets developers build full-stack web apps through simple text prompts. Built by Fly.io, it interprets your instructions, writes the code, and runs it in a live environment with real-time previews.

How you can use it:

  • Generate CRUD apps from natural language

  • Modify features with prompt-based edits

  • Preview functionality in real time

  • Go from launch to production directly in your browser

Pricing: Paid

Clone your voice, or create a new one

11.ai

11.ai is a powerful voice synthesis tool that creates realistic audio using cloned voices or text-to-speech. Great for content creators, game devs, or anyone in need of high-quality narration without a mic.

How you can use it:

  • Generate dialogue for games or animations

  • Create multilingual voiceovers

  • Produce audio for podcasts or courses

  • Clone your voice for digital avatars

Pricing: Free

Jobs, announcements, and big ideas

  • Songscription launches to become the Shazam for sheet music.

  • YouTube rolls out AI-powered discovery tools to help users find content faster.

  • Xiaomi unveils AI glasses with 2K video recording and built-in voice assistant—for just 1,999 yuan.

  • There’s a breezy, 5-minute morning read that keeps you sharp on everything business and tech. The Hustle covers stories that matter to consumers and summarizes the fascinating rabbit holes you don’t have time to sift through yourself. 2M+ people get the daily in their inbox to stay informed and have some fun—join them!*

  • Google expands “Ask Photos” to more users, turning your camera roll into a searchable memory bank.

  • Google's Doppl app now lets you try on outfits virtually with AI.

  • Suno acquires WavTool to supercharge AI music editing for pros.

  • Gemma 3n is here: Google DeepMind’s new model boosts multimodal AI on edge devices.

*This is sponsored advertising content.

Let’s talk about AI slop. A comment on my recent YouTube video said people don't worry about "AI slop" until it's something that affects their jobs. Am I overreacting?

That’s a wrap! See you next week.

—Matt (FutureTools.io)

P.S. This newsletter is 100% written by a human. Okay, maybe 96%.