GM! Welcome to Get Into AI (formerly The Status Code).

We're your AI compass, pointing to what matters in this fast-moving field.

Here’s what we have for today:

1. Meta’s new AI coder

2. Another safety expert leaves OpenAI

3. What Anthropic CEO is saying about DeepSeek

Not subscribed yet? Stay in the loop with AI weekly by reading Get Into AI for five minutes.

1/  Meta's Big Bet: Llama 4 Aims to Lead the AI Race

Mark Zuckerberg just dropped some big news about Meta's Llama 4, and it's turning heads in the AI world.

First up: Llama 4 mini has finished its initial training phase. Think of it as the younger, nimbler sibling in the Llama family. The bigger versions are still in training, but early results look promising.

What makes this special is that Llama 4 isn't just another language model. It's what Meta calls an "omni-model" capable of simultaneously handling text, images, and other types of data.

It also has built-in agent capabilities, which means it can actively assist with tasks rather than simply responding to prompts.

But here's the real kicker: Zuckerberg predicts that by the end of 2025, we'll see AI coding assistants that match mid-level human engineers in skill. That's not just impressive – it could reshape how software gets built.

2/ Copyright Office Draws the Line on AI Creativity

The U.S. Copyright Office released guidelines that clear up a lot of confusion about AI-created work. Here's what you need to know:

Using AI as a creative tool? Your work can still be protected by copyright. It's like using Photoshop – the tool doesn't take away your creative ownership.

But if you're just writing prompts and picking from AI outputs? That's not enough for copyright protection. The key word here is "human authorship" – they want to see your creative fingerprints, not just your prompt engineering skills.

Think of it this way: directing a photoshoot gets you copyright protection. Picking your favorite photos from someone else's shoot doesn't.

3/ The DeepSeek Debate Continues: Anthropic's CEO Weighs In

The buzz around DeepSeek's AI breakthrough just got more interesting. Enter Dario Amodei, CEO of Anthropic (the company behind Claude), who offers a perspective that cuts through the hype.

First, the headline-grabber: DeepSeek claims they can match top AI performance at bargain prices. Big enough news to knock 17% off NVIDIA's stock price. But Amodei suggests we take a breath.

In his analysis, DeepSeek's cost reduction isn't breaking any laws of physics. It's more like the steady price drop we see in smartphones - impressive but expected. Think of it as AI following its own version of Moore's Law.

Amodei says DeepSeek really shines in its clever engineering. They've found a way to run their AI more efficiently by splitting tasks between specialists rather than using one do-it-all system.

This perspective matters, coming from Anthropic's CEO. His company focuses on building safe, interpretable AI systems, so he knows how to balance innovation with practical implementation.

That’s it for this week, folks! If you want more, be sure to follow my Twitter (@BarunBuilds)

🤝 Share Get Into AI with your friends!