Elon Musk's xAI open released Grok-1

PLUS: Get your Haiku on

Sponsored by


Good morning, human brains, and welcome back to your daily munch of AI news.

Here’s what’s on the menu today:

  • What the actual Grok? 😤 🤘

    xAI open-sourced its Grok-1 AI model. You can download it on GitHub.

  • Mercedes-Benz tests car-building robots 🚗 😴

    It partnered with Apptronik to automate low-skill, repetitive tasks.

  • Get your Haiku on 🏃‍♂️ 💨

    Anthropic launched its fast, powerful, and cost-effective model.


Grok it like it’s hot 😤 🤘

On Sunday, Elon Musk’s xAI open released Grok-1. You can download it now, it’s about 300 GB and contains 773 files.

What the Grok?

Grok-1 is xAI’s open-source model. Its architecture and weights are available under the Apache 2.0 license, which enables commercial use.

Why do I care?

xAI trained Grok-1 within just 3 months of the company’s founding. It boasts 314 billion parameters and is the largest open-source mixture-of-experts model.

What is a mixture-of-experts model?

It’s a model that combines several, smaller models with different specialties. For reference, GPT-4 is a mixture-of-experts model that consists of 8 different models.

Grok-1 is completely open-source?

Well, not entirely. xAI didn’t reveal its training datasets/code.

Is it any good?

It hasn’t been fine-tuned for any application or optimized for safety yet. Check it out on GitHub, here.

Didn’t Elon sue OpenAI?

Why yes. Two weeks ago, we reported on Musk’s lawsuit against OpenAI. He claims it illegally transitioned from a non-profit to a Microsoft subsidiary.

What else has xAI done?

Back in July, we covered xAI’s live news updates. Over 1.6 million people listened to its Tesla collaboration, use of Twitter data for AI training, and more.

In November, we reported on xAI’s AI model, Grok. It uses real-time world knowledge to wittily answer questions, suggest questions to ask, and more.

The next day, we covered xAI’s coding tool, PromptIDE. It allows you to create and test prompts while using Grok-1.

Hire a world class AI team for 80% less

Trusted by leading startups and Fortune 500 companies.

Building an AI product is hard. Engineers who understand AI are expensive and hard to find. And there's no way of telling who's legit and who's not.

That's why companies around the world trust AE Studio. We help you craft and implement the optimal AI solution for your business with our team of world class AI experts from Harvard, Princeton, and Stanford.

Our development, design, and data science teams work closely with founders and executives to create custom software and AI solutions that get the job done for a fraction of the cost.

P.S. Wondering how OpenAI DevDay impacts your business? Let’s talk!


Replace pesky humans with bots 🚗 😴

On Friday, Mercedes-Benz and Apptronik announced their commercial agreement. The goal is to integrate robots in the automobile manufacturing process.

What’s the agreement?

Mercedes-Benz will test humanoid robots from Apptronik to automate repetitive tasks in its manufacturing facilities.

Humanoid robots?

The robot, Apollo, weighs 160 lb, and can lift objects up to 55 lb. It stands at 5’ 8” and will deliver parts to the production line, inspect components, and more.

Why robots?

Mercedes-Benz claims it’s to address its labor shortages and support its workforce.

"This is a new frontier, and we want to understand the potential both for robotics and automotive manufacturing to fill labor gaps in areas such as low-skill, repetitive and physically demanding work and to free up our highly skilled team members on the line to build the world's most desirable cars."

Jörg Burzer — Mercedes-Benz

What else has Mercedes been up to?

Back in June, we reported on Mercedes integrating ChatGPT into its cars. It enabled voice commands, contextual conversation, and integration with third-party apps.


Join over 110,000 people learning actionable tips and growing personally and professionally with Amanda.

Each week, subscribers receive a deep dive that covers topics ranging from personal growth, productivity, career growth, mental health, money and relationships.


Anthropic launched Haiku 🏃‍♂️ 💨

Last week, Anthropic launched Claude 3 Haiku. It’s designed to offer fast, affordable AI solutions for enterprises.

What is it?

Claude 3 Haiku is the latest addition to its Claude 3 model family. It claims to offer unmatched speed, cost-efficiency, and performance for businesses.

What’s so great about it?

It excels at customer support, chat interactions, and more. It analyzes documents, processes images, and is cheaper to use than GPT-3.5 and Gemini 1.0 Pro.

Is it safe?

Anthropic claims it’s been thoroughly tested and contains multiple layers of enterprise-grade security measures.

What about the other Claude 3 models?

Two weeks ago, we reported on Opus, Sonnet, and Haiku. You can access them now via Anthropic’s API, Amazon Bedrock, and more.



Learn No-Code and AI skills by committing to the 100DaysOfNoCode or 100DaysOfAI challenge. Receive a 30-minute bite-sized lesson straight to your inbox every day. It’s free, fun and effective! Both challenges start April 1st, 2024 ⚡️

Fine — a network of AI agents that provide development and coding assistance.

Tailor — creates and sends AI-generated news article summaries.

Flowsage — quickly and intuitively create flowcharts.

Think Pieces

Looking for a job in the AI robotics field? Here is a list of 61 robotics companies that are hiring.

Can Blockchain technology uncover deepfakes? Fox and Polygon claim it can verify data and authenticate content.

5 of 45 of this year’s Pulitzer Prize winners used AI in their research. This is the first time that AI usage is required to be disclosed.

Startup News

Microsoft expanded Copilot Pro to 22 more countries. It also now comes with a 1-month free trial.

Cerebras Systems introduced the world’s largest AI chip. The WSE-3 can handle LLMs with tens of trillions of parameters.

Apple is reportedly negotiating with Google. The goal is to integrate Google Gemini into several iPhone features.


VLOGGER — Google’s image-to-video tool with controllable face/body representations, variable-length videos, and more.

VisionGPT-3D — a 3D vision comprehension framework that leverages LLMs and CV (Computer Vision) models to enhance multimodal functionality.

Follow-Your-Click — enhances the controllability of object movement, background animation, and more in image-to-video models.



Here’s how different LLMs react when you tell them you pushed the wrong button on a nuclear reactor.

Tag us on Twitter @BotEatBrain for a chance to be featured here tomorrow.


Until next time 🤖😋🧠 

What'd you think of today's newsletter?

Login or Subscribe to participate in polls.