Is Google finally leading the AI race?

In partnership with

GM! Welcome to Get Into AI.

I’m like an espresso shot.

Quick. Potent. Guaranteed to wake you up to what's happening in the world of AI.

Here’s what I have for today:

1. Google’s Stealing the Show

2. Gemini Robotics Makes Machines Move Smarter

3. Is Claude Ready To Replace Junior Devs?

Before we begin, a word from our sponsors:

Learn AI in 5 minutes a day

This is the easiest way for a busy person wanting to learn AI in as little time as possible:

1. Sign up for The Rundown AI newsletter

2. They send you 5-minute email updates on the latest AI news and how to use it

3. You learn how to become 2x more productive by leveraging AI

Sign up to start learning.

Alright, let’s dive in!

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

Three main stories for the day.

1/ Gemma 3: Google's Open Model Gets Supercharged

Google just dropped Gemma 3, and it's turning heads faster than a cat video on social media.

The standout feature? A massive 128k context window that lets this model remember more than my grandma remembers embarrassing stories about me.

Available in various sizes (1B, 4B, 12B, 27B), this multilingual powerhouse speaks 140+ languages and handles both text and visuals with ease.

The 27B version is even matching Gemini-1.5-Pro on benchmarks and ranked 9th on LLMArena, outperforming some serious competition.

What's the big deal? Gemma 3 represents a new Pareto frontier for open models in its weight class – it's like someone figured out how to get sports car performance out of a family sedan.

Plus, it completely absorbs PaliGemma's vision capabilities while maintaining the ShieldGemma safety features.

2/ Gemini Robotics Makes Machines Move Smarter

Google DeepMind is teaching robots to think with their new Gemini Robotics Models.

These AI systems, built on Gemini 2.0, aim to create robots with better reasoning, interactivity, dexterity, and generalization skills.

They've partnered with Apptronik and opened their Gemini Robotics-ER model to trusted testers including Boston Dynamics.

The goal? AI that works for any robot shape or size – from humanoids to industrial arms.

In demonstrations, these models showed off by wrapping timing belts around gears and folding origami. You know, the tasks that require the kind of fine motor skills I wish I had when trying to wrap Christmas presents.

3/ Claude Sonnet 3.7 Is Getting Insane at Coding!

Claude 3.7 Sonnet is flexing its coding muscles so hard it's making a group of redditors nervous.

In a popular reddit thread, a user converted a complex 4,269-line JavaScript application into a fully-featured Vue 3 app in a single session, complete with proper component structure, Pinia stores, and Vue Router.

Not everyone agrees with them though. Some praise its ability to create bug-free applications while others note it occasionally hallucinates features like an overeager intern trying to impress the boss.

One user shared how Claude transformed Mixpanel CSV data into a comprehensive dashboard in minutes, potentially saving thousands in BI tool costs.

It's like having a coding genie without the three-wish limit. But like any genie, you better be careful what you wish for. Do you want to be replaced?

🐤Tweet of the day

Not AI-related, but something to think about…

Catch you tomorrow! ✌️

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

🤝 Share Get Into AI with your friends!