💬 Hertz-dev: AI That Talks Like a Human

Hey there!

I spent my entire Saturday diving into the wild world of AI.

To be honest, who needs a boring weekend when there’s so much mind-blowing tech happening? 🤯 

Between the latest AI breakthroughs and some seriously cool new apps, I couldn't tear myself away from the screen.

And guess what? You don't have to miss out either.

So, grab your favourite drink, kick back, and let’s dive into the world of AI together.

What’s Inside Today’s Newsletter

🪰 Buzz Around AI

😀 AI Recap

🪄 AI Creation

🔹 AI Learning

♨️ What’s Hot In AI

📰 AI News

Read time: 10 mins

🪰 Buzz Around AI

🎧💬 Hertz-dev: AI That Talks Like a Human

Standard Intelligence has launched Hertz-dev, an open-source AI model that creates realistic, natural-sounding speech for real-time conversations with low delay.

Key Features:

• 8.5 Billion Parameters: The AI model was trained on 20 million hours of high-quality audio, making it capable of having smooth, bi-directional conversations.

• Realistic Speech: It captures details of human speech like pauses and emotions by using real-world dialogue data.

• Customizable: The base model can be adjusted for tasks like live translation, audio classification, and building chatbots.

• 70B Parameter Version: A more powerful version with 70 billion parameters is also available under the Apache license, with code and checkpoints provided.

• Multi-Modal Plans: There are plans to expand the model to work with more types of data, not just audio.

Why It Matters:

Hertz-dev gives developers an open-source tool that they can use to build smarter voice applications. It can be used to improve customer service robots, create better voice assistants, and much more.

Nvidia accelerates robot learning and humanoid development using AI and simulation tools.

Nvidia just introduced a powerful set of AI and simulation tools for robotics at the 2024 Conference on Robot Learning (CoRL).

The new toolkit includes tools for humanoid robots, training systems, and a partnership with Hugging Face, an open-source platform.

Key Details:

• Isaac Lab: Nvidia’s open-source framework for training robots is now available, allowing for large-scale robot training.

• Project GR00T: A new initiative that adds specialized tools for humanoid robot development. This includes motion generation and environment perception.

• Hugging Face Partnership: Nvidia’s tools now work with LeRobot, a platform by Hugging Face, to speed up AI robotics development.

• Cosmos Tokenizer: A new technology that processes robot visual data up to 12 times faster than current solutions.

Why It Matters:

The race to create advanced humanoid robots is heating up. Nvidia is positioning itself as a key player, providing the essential tools for the entire industry. With new training tools and better AI models, robotics development is moving faster than ever.

😀AI Recap

Perplexity Sees Explosive Growth

At just two years old, Perplexity is making waves in the AI industry.

The company is closing in on its fourth funding round of the year, aiming for $500 million.

This latest round will significantly boost its valuation, positioning Perplexity among the top AI firms globally.

How Perplexity Stands Out:

While many AI companies tread cautiously around politically sensitive topics, Perplexity took a bold step.

Instead of avoiding election-related questions, it launched a voter information hub, offering transparency and engaging with the public.

Though it’s faced some pushback from major publishers like Forbes and Wired over its content practices, Perplexity has found ways to navigate these challenges.

The company has introduced revenue-sharing partnerships to address concerns and keep content creators on board.

🪄 AI Creation

Prompt:- Create a kawaii-style cartoon image of a cute girl reading a book surrounded by candlelight.8k --v 6.1 - 

🔹AI Learning

How to Create Eye-Catching Ads for Your Products in Minutes Using AI ✨

Want to make your products stand out with minimal effort?

With the power of AI tools, you can create stunning ads in just a few simple steps!

Follow this guide to get your product ads ready in no time:

1️⃣ Upload Your Product Photo to Qreates 📸

Start by uploading a high-quality image of your product to Qreates.

This will be the base of your ad, so make sure it's clear and well-lit to showcase your product at its best!

2️⃣ Add a Reference Image & Generate Your Custom Design 🎨

Next, upload a reference image that represents the look and feel you want for your ad.

It could be a style, colour palette, or another image you find inspiring.

3️⃣ Use Luma Labs to Add Movement 🎥

To add dynamicity to your ad use Luma Labs to add movement!

✨ It’s That Easy!

In just 3 simple steps, you can create professional, eye-catching ads that make your products pop all with the help of AI!

♨️ What’s Hot In AI

🎭 ByteDance Unveils X-Portrait 2: A Powerful AI Portrait Animator  

ByteDance has launched X-Portrait 2, an AI tool that turns static images into animated portraits.

It uses a reference video to map facial movements and bring still images to life with realistic expressions.

Key Features:

- Single Video Input: Just one video drives the animation, turning any image into a moving character.

- Realistic Expressions: X-Portrait 2 captures subtle facial movements like pouting, frowning, and tongue movement with smooth realism.

- Works with All Styles: The system works for both realistic portraits and cartoon characters, offering endless possibilities for animation and visual effects.

- Better than Before: X-Portrait 2 builds on the first version (released in July) and could soon be integrated into TikTok, offering a free alternative to other AI avatar tools.

Why It Matters:

X-Portrait 2 makes professional-grade animation accessible to everyone.

It could change how we create content and spark new ideas.

But as AI-generated faces look more real, will we lose our sense of what's genuine in the digital world?

📰 AI News

1. Black Forest Labs upgraded its FLUX1.1 Pro model. The new ‘Ultra’ mode offers 4x higher resolution for text-to-image generation. There’s also a ‘Raw’ mode for more realistic images.

   

2. Wendy’s is teaming up with Palantir to launch an AI-powered system. This system will predict shortages and automate inventory orders.

3. Mistral released a new content moderation API. It powers Le Chat, helping developers add safety features in apps across nine categories.

   

4. Krea AI now lets users create custom AI models. These models can learn specific characters, art styles, and product designs.

5. Nous Research launched its first public chatbot, Nous Chat. It’s powered by the Hermes 3-70B model.