Hey folks!

Remember our last newsletter where we talked about Rabbit's new AI thing, the R1? Well, we've got a special treat for you! We made a video to dive deep into what this Rabbit R1 is all about.

This Rabbit R1 is like a brainy phone that understands you and does tasks by just talking to it. It's like having a smart helper with you all the time. Let's check it out!

WATCH THE VIDEO ON YOUTUBE

Today:

OpenAI Signs Up 260 Businesses for Corporate Version of ChatGPT

In just four months, OpenAI's got 260 companies on board for its business-focused ChatGPT. This move, showing big demand, marks their main shot at cashing in on their famous AI tool. They've roped in over 150,000 employees from these firms to use ChatGPT. 

OpenAI's keeping the details like how much they charge under wraps, saying it varies by company. Launched last August, ChatGPT Enterprise promises extra features and tighter security, like data encryption. They've even got a new cheaper option for smaller teams at $25 a month per person, plus a $20 monthly deal for solo users.

THE BUSINESS TIMES

OpenAI eyes partnership with CNN, Fox and Time to secure news content licensing

OpenAI's chatting with big news players like CNN, Fox, and Time to license their stuff and train their AI without stepping on copyright toes. They're eyeing news articles, videos, and photos for their chatbots to be smarter and more up-to-date. Fox Corp's even launching a blockchain thing to track AI use of its content. Time's CEO's hopeful about striking a fair deal. 

This move, especially after OpenAI got sued by The New York Times and some authors for using their work without paying, is a big deal to dodge future legal headaches. OpenAI's free ChatGPT-3.5 runs on old data, but their fancier versions can now surf the net, not tied to old training limits.

COINTELEGRAPH

1X, robotic startup backed by OpenAI, receives $100M in funding

1X Technologies, a Norwegian robotics startup, just bagged a cool $100 million in funding. This cash injection, led by EQT Ventures with Samsung and others chipping in, is meant to boost their game-changing bipedal android, NEO, and support their existing business clients. 

These guys, originally Halodi Robotics, have been at it for a decade, crafting robots for heavy-duty jobs. Their big leap was teaming up with OpenAI in 2022, blending AI smarts with their robots. This means their robots, like the wheeled Eve, can do complex tasks and learn on the job in real-world settings, guided by VR tech. 

Now, they're eyeing the consumer market with NEO, designed for everyday home help. Despite stiff competition in the robotic world, 1X is pushing boundaries, with a total funding pot of $137 million. They're all about making robots more human-friendly and useful in the real world.

VENTURE BEAT

Microsoft unveils new generative AI and data solutions across the shopper journey, offering copilot experiences through Microsoft Cloud for Retail

Ahead of NRF 2024, Microsoft revealed new generative AI and data solutions for retailers, enhancing the shopping experience from start to finish. They're rolling out copilot templates on Azure OpenAI Service, allowing personalized shopping experiences and better store operations support. Microsoft Fabric introduces industry-specific retail data solutions for deeper insights. Additionally, Microsoft Dynamics 365 Customer Insights and Microsoft Retail Media Platform are getting new features.

Microsoft Fabric offers data solutions to unify retail data and apply AI, extracting actionable insights. Retailers can now quickly implement specific data models and transform data into predictive insights for better customer engagement.

Finally, Microsoft Dynamics 365 Customer Insights introduces new Copilot features for marketers, allowing AI-driven project creation and campaign management. The Microsoft Retail Media Platform, powered by PromoteIQ, introduces the Retail Media Creative Studio, using generative AI for ad creation.

MICROSOFT

India data centre firm Yotta's Nvidia AI chip orders to reach $1 bln -CEO

Yotta, an Indian data center firm, is set to hit a $1 billion milestone in AI chip orders from Nvidia. This includes a fresh $500 million order for nearly 16,000 Nvidia AI chips. These purchases are part of Yotta's push to enhance its AI cloud services. 

Nvidia's AI deals in India are significant, especially with U.S. export restrictions affecting chip sales to China. Yotta, backed by Indian billionaire Niranjan Hiranandani, operates three data centers across India and collaborates with Nvidia. This move is in line with India's booming demand for cloud and data center services.

REUTERS

Google Cloud launches new generative AI tools for retailers

Google Cloud just rolled out a bunch of new AI tools for retailers, aiming to jazz up online shopping and other retail operations. These include an AI-powered chatbot that can chat with customers on websites and mobile apps, offering personalized product suggestions. This tech isn't just online; physical stores are getting a piece of the action too with Google Distributed Cloud Edge, blending hardware and software.

Among these new goodies, there's a large language model feature to make product searches smarter. It's currently in trial mode with some stores, but it's set to be widely available later in the year.

CNBC

Rabbit sells out two batches of 10,000 R1 pocket AI companions over two days

Rabbit's pocket-sized AI gadget, the R1, has proven immensely popular, selling out of its first and second batches of 10,000 units within a day each. The first batch is set to ship in March, while the second batch is expected in May to June 2024. 

The R1 features a 2.88-inch touchscreen and runs on Rabbit's own OS, acting as a universal controller for various apps, allowing tasks like music playback, grocery shopping, and messaging without using a phone. It can also be trained to interact with specific apps. A third batch is now available for preorder.

THE VERGE

🧠RESEARCH

PIXART-δ: FAST AND CONTROLLABLE IMAGE GENERATION WITH LATENT CONSISTENCY MODELS

"PIXART-δ, a new tech for turning text into images, is a game-changer. It's way faster than its predecessor, creating high-quality, large images in half a second—7 times quicker! Plus, it's easy to train and runs on common GPUs. It even lets users tweak image details, making it a top pick for image generation, rivaling other big names in the field."

InseRF: Text-Driven Generative Object Insertion in Neural 3D Scenes

"InseRF is a new tool that spices up 3D scenes by adding objects based on what you describe and where you mark in a 2D view. Unlike older methods that just tweak looks or take things out, InseRF can create new stuff in 3D, which is a big deal. It uses a cool trick: starting with a 2D edit and then popping it into 3D. It's smart enough to make these new objects fit just right in the scene without needing any fancy 3D input. Plus, their tests show it's better than what's already out there."

URHand: Universal Relightable Hands

"URHand is a groundbreaking universal hand model that can handle all kinds of lighting, angles, and even different people's hands. It's perfect for creating photorealistic hand images under any conditions. This model is so smart that it can even personalize itself with just a few phone pics. By combining physics-inspired shading and neural magic, it makes hand modeling look easy and realistic, outperforming everything else out there."

🛠️TOP TOOLS

MultiOn - AI solutions that streamline digital experiences. Their technology focuses on automating tasks in the background of digital platforms, allowing users to minimize hands-on work and concentrate on more fulfilling activities. 

Luma AI - multimodal AI technologies that aim to augment human imagination and capabilities. They specialize in transforming text into 3D representations, offering interactive scenes and an API for video to 3D conversion. 

Sendspark - AI-powered video prospecting platform that allows users to create and personalize sales videos at scale. The key features include AI voice and video cloning, dynamic backgrounds, and personalized video pages. 

Hypotenus AI - AI writing assistant designed for generating diverse types of content quickly and efficiently. It's capable of creating blog articles, product descriptions, social media posts, and more.

Motion - productivity platform that combines various organizational tools into one system. It offers an intelligent calendar, automatic task and project management, and meeting scheduling.

📲SOCIAL MEDIA

🗞️MORE NEWS

The young people sifting through the internet’s worst horrors

San Francisco AI company, Sama, is under fire for paying workers a mere $2.20 per hour to view disturbing content, leading content moderators in Nairobi, Kenya, to sue both Sama and Meta. The job involves endless hours of watching violent and horrifying material, including sexual abuse and gruesome murders. This issue highlights the exploitation of workers in Africa, Asia, and South America, supporting AI initiatives at tech giants like OpenAI, Microsoft, Google, and Meta. THE VERGE

Victoria's Secret & Co. to Create New AI-Powered Shopping Experiences with Google Cloud

Victoria's Secret & Co. has entered a multi-year partnership with Google Cloud to enhance online shopping experiences and operational efficiency. The collaboration will leverage Google Cloud's AI and generative AI technologies to create personalized customer experiences at scale, including a generative AI-powered conversational assistant. The goal is to offer tailored product recommendations and advice based on customer preferences and experiences. Google Cloud's Vertex AI Search will also enhance website and mobile app search capabilities. PR NEWSWIRE

Ask-AI raises $11 million for its Generative AI assistant for enterprise workflows

Ask-AI, a generative AI solution for enterprises, has secured $11 million in Series A funding, led by Leaders Fund, with participation from Seed investors Vertex Ventures, State of Mind Ventures, GTMFund, and others. The company's new sidekick, ASK, aims to streamline repetitive tasks and enhance customer support by integrating with over 50 enterprise systems like Salesforce, Slack, and Google Drive. ASK provides contextual information and concise answers, improving productivity and customer satisfaction. Ask-AI's CEO, Alon Talmor, highlights the need for user-friendly AI applications in the workforce. The company, founded by Talmor, plans to double its team size soon. CTECH

‘Set it and forget it’: automated lab uses AI and robotics to improve proteins

Researchers have developed a self-driving laboratory that uses AI and robotics to autonomously reengineer enzymes, eliminating the need for human intervention, except for occasional hardware maintenance. The system employs a simple machine-learning model to propose sequence changes for protein engineering, streamlining the process. In a study, the self-driving lab successfully improved the heat tolerance of enzymes after 20 experimental rounds. While self-driving labs hold promise for accelerating scientific discoveries, there is a need for more advanced hardware and the ability to generate knowledge interpretable by machines. Researchers plan to apply this technology to various protein-engineering challenges while emphasizing that it complements, rather than replaces, human involvement in scientific work. NATURE

AI Discovers That Not Every Fingerprint Is Unique

AI has challenged the long-held belief that all fingerprints are unique. Columbia Engineering undergraduate Gabe Guo used a deep contrastive network to analyze 60,000 fingerprint pairs, some from the same person but different fingers, and others from different people. The AI improved its accuracy in distinguishing these prints, reaching 77% for a single pair and significantly higher for multiple pairs. This could potentially enhance forensic efficiency tenfold. The findings, though initially met with skepticism, highlight AI's transformative potential in well-established fields and the possibility of AI-driven scientific discoveries by non-experts. COLUMBIA ENGINEERING

Data Suggests Growth in Enterprise Adoption of AI is Due to Widespread Deployment by Early Adopters, But Barriers Keep 40% in the Exploration and Experimentation Phases

A recent study commissioned by IBM reveals that 42% of enterprise-scale organizations with over 1,000 employees have actively deployed AI in their businesses. Additionally, 40% are exploring or experimenting with AI. Among those already exploring or deploying AI, 59% have accelerated their investments in the technology. The top barriers to AI adoption include limited AI skills and expertise (33%), data complexity (25%), and ethical concerns (23%). The study also highlights the industries and regions leading in AI adoption and the key drivers behind AI adoption in enterprises. IBM

Geographical imbalances remain in AI research funding

Geographical imbalances in AI research funding within the European Union (EU) are evident, with disparities especially pronounced in eastern member states. Germany, Italy, and France receive the majority of funding from EU programs, and they also allocate significant national funding to their AI industries. However, Europe faces a €10 billion funding deficit compared to major AI players like China and the US. While the European Commission invests €1 billion per year in AI technologies through Horizon Europe and the Digital Europe scheme, most of this funding is concentrated in western European member states. Eastern member states like Poland and Romania have lower participation rates, highlighting a gap in AI readiness across the continent. This imbalance reflects the need for more equitable distribution of AI research funding in the EU. SCIENCE BUSINESS

How AI Replaced the Metaverse as Zuckerberg’s Top Priority

Mark Zuckerberg's shift in priorities from the metaverse to AI has raised questions about Meta's future. Initially, Zuckerberg's metaverse vision led to Meta's rebranding, but it has incurred significant losses and skepticism. Employees complain about innovation challenges and the need to align with Zuckerberg's favored projects. Now, Zuckerberg is refocusing on AI, particularly Meta's AI research group, FAIR. Industry observers have praised Meta's AI efforts, emphasizing generative AI. However, Meta's history of shifting priorities keeps stakeholders cautious about what's next. BLOOMBERG

OpenAI: ‘Impossible’ to Train Models Without Copyrighted Content

OpenAI, creators of ChatGPT, is facing a lawsuit from The New York Times for allegedly using copyrighted material to train their AI models. They argue it's "impossible" to train such models without using copyrighted content, as almost every type of human expression falls under copyright. OpenAI believes that using only public domain content would limit AI effectiveness significantly. The company asserts that they follow the law, including copyright, and consider their training methods as fair use. OpenAI also mentioned that websites can block their web crawler, GPTBot, and creators can opt out of having their content used in future DALL-E datasets.  ARSTECHNICA