Luma will blow your mind

+ AI * robots = the future

Mindstream Banner

PRESENTED BY

The letter "J" was the most recent to be added to the English alphabet - in 1524. Before that, "I" did double duty. Juicy!

Love or hate this intro? Hit reply and tell us why!

What’s in store:

  • Meet Luma, A Sora competitor that will blow you away.

  • Robots that follow orders, literally.

  • Around the web: Wells Fargo fires over a dozen employees for using "mouse movers".

  • Jurny’s fundraiser closes tomorrow - secure your stake!

Read Time: 4 minutes

AI CINEMA

Luma AI, a startup from San Francisco, has introduced Dream Machine, a cutting-edge AI system that creates high-quality videos from simple text descriptions.

The tech enables creators and companies to produce original video content quickly and efficiently - much like OpenAI’s Sora.

With Dream Machine, users can input prompts like "a corgi puppy chasing a ball on a beach at sunset" and, in about two minutes, receive a realistic five-second video.

Early testers have praised its ability to render objects, characters, actions, and environments with smooth motion and coherent storytelling.

Here’s what you should know:

  • Dream Machine generates high-quality videos from text prompts in two minutes.

  • Early users praise its rendering abilities, but challenges remain.

  • Ethical and legal issues need to be addressed as technology advances.

AI takes the director’s chair

However, despite its advanced capabilities, Dream Machine and similar systems like Sora and Kling can still struggle with certain prompts, sometimes producing nonsensical or problematic outputs.

Improving their quality and consistency is vital for realising their full commercial potential.

As these tools make video content generation easier, they also bring ethical and legal challenges, such as the potential for creating deepfakes and infringing on intellectual property rights.

Regulators are working to address these issues, but technology is advancing faster than governance frameworks can keep up.

The launch of Dream Machine marks significant progress in AI-powered video generation.

As the technology evolves, it promises to unlock new forms of creativity and expression, ushering in the age of AI-generated media.

Someone use this to remake Game of Thrones’ final season, please.

Is AI the future of video production, or just a passing fad?

Login or Subscribe to participate in polls.

Vote for live results and see results + opinions from yesterday at the bottom of the email.

AI STARTUPS

Jurny is an AI platform automating hotel and short-term rental operations, targeting $355B in inefficiencies.

Partnered with industry giants like Airbnb, Vrbo, and Expedia, Jurny’s technology is fully automating operations for thousands of properties globally.

  • In 2023, Jurny’s customer base grew 5x and it processed $35M+ in bookings.

  • Featured by CNBC’s Squawk Box for its AI innovations and impact on hospitality, as well as by Forbes, Bloomberg, Skift, and many others.

  • In just six months since its rollout, Jurny's AI has autonomously managed every aspect of guest communications for over 143,000 reservations.

On This Day

June 17th 1885 - The Statue of Liberty arrives in New York Harbor after being shipped across the Atlantic Ocean in 350 individual pieces packed in more than 200 cases from France.

Trending Tools

BlueDot*: AI note-taker that records your meetings without bots joining calls. Install Chrome Extension →

Let’s Foodie: An AI-driven application that streamlines the process of creating recipes from available ingredients, providing clear instructions suitable for all cooking levels.

Style DNA: This AI-powered personal stylist provides personalised fashion recommendations based on your style, body shape, and preferred colours, making it easier to choose outfits for any occasion.

Wysa: An AI conversational care platform that offers mental health support through evidence-based exercises and programs. It creates a safe space for users to discuss their worries and manage stress, with options for human coaching too.

CF Spark: A free creative tool that allows users to generate images based on prompts. It's designed to spark creativity and is part of a vibrant community where creators share and inspire each other daily.

ROBOTS

MIT researchers have developed a new AI system that guides robots using language-based instructions, improving navigation without relying heavily on visual data.

This innovative approach translates visual observations into text captions, allowing a language model to direct the robot's movements.

While it doesn't outperform traditional vision-based systems, it excels in data-limited scenarios and works well when combined with visual inputs.

Imagine your home robot carrying dirty laundry to the washing machine in the basement.

For this, the robot needs to combine your verbal instructions with its visual observations to navigate effectively.

Traditionally, this requires multiple machine-learning models and vast amounts of visual data for training, which is often challenging to gather.

To address these issues, MIT and the MIT-IBM Watson AI Lab have created a method that converts visual data into language.

Instead of using complex visual representations, the robot’s point-of-view is described in text captions.

A large language model then uses these captions to decide the robot's next steps based on your instructions.

Why it’s great:

  • Efficiency: The language-based approach generates synthetic training data quickly and efficiently.

  • Versatility: It performs well in scenarios with limited visual data and can be combined with visual signals for better results.

  • Human-Friendly: Text-based representations make it easier to understand and troubleshoot the robot’s actions.

The new AI frontier

The researchers found that while this method doesn't capture some details that vision-based models do, combining both approaches improves the robot's navigation ability.

They aim to further explore how language can enhance spatial awareness and overall performance in navigation tasks.

By incorporating language models, this research opens up new possibilities for making robots more adaptable and easier to program for various tasks and environments.

Finally, a robot that listens better than your average teenager.

Mindstream Picks

During the Xbox Game Showcase, Bethesda revealed new footage of Indiana Jones and the Great Circle, set between Raiders of the Lost Ark and The Last Crusade.

Bitvavo's analyses highlight investor anticipation in week 23, with Bitcoin holding at a promising $69,000, near its 2021 bull market peak.

Kayhan Space unveiled Satcat, a new resource combining open-source data on spacecraft and debris with research tools, simplifying satellite information transmission.

The Nasdaq rose over 3%, the S&P 500 nearly 1.5%, both hitting record highs for four straight days, while the Dow fell over 0.7%.

Don’t Miss - Wells Fargo recently terminated over a dozen employees from its wealth- and investment-management unit for using "mouse movers" to simulate productivity, as reported by Bloomberg based on Financial Industry Regulatory Authority disclosures. These devices and apps, which automate cursor movements, became popular during the pandemic to evade remote work monitoring software.

AI Art

Our Image of the Day

Artwork submitted by Mindstream reader AZ: “Huge chewing gum”

Daily Image Prompt

The statue of Liberty on the moon

Yesterday’s Poll

“Do you think Meta should be allowed to train its AI without explicit user consent?”

No, privacy should come first. - 85%

Yes, it's essential for innovation. - 15%

Your Views:

“Absolutely not that is like robbery and can lead to major problems. Find other data to use.” - rkayew

“If you post on Facebook, it is in the public domain.” - david.harrison

Submit your opinions in our daily poll to be featured!

Collage of Authors

Happy Monday! We're here all week everybody.

❤️ We need your feedback to make our newsletter better.

📣 Refer our newsletter to your friends and make us feel all warm and fuzzy.

🚀 Advertise in our newsletter to reach 130,000+ founders, engineers, and content creators.

You There! What Did You Think?

Login or Subscribe to participate in polls.