🧑‍⚖️ Elon vs OpenAI: When Lawsuits Become the New Clickbait

Hello AI Enthusiast,

In this week's spotlight, Elon Musk is suing OpenAI, saying they've changed their mission for the worse. OpenAI is hitting back, saying Musk was well aware of their direction. It just keeps feeling like a movie with all the drama, and when email conversations are shared things get interesting.

Now, let's check out what you all thought in last week’s polls.

Last time we delved into your thoughts on two hot topics: Apple's venture into generative AI and Klarna's AI-driven customer service. A whopping 87.5% of you think Apple entering the generative AI scene could really shake things up. However, one voter raised a thoughtful point: “Self-driving vehicles are an important way to raise productivity in our economies. The innovators at Apple were a great hope that this could be done.” Is it a perspective worth pondering?

And with 82.5% eager for a closer look at Klarna's AI and its impact on service quality, a deep dive is on the way. Stay tuned. 🤓

Now, on with the news. 👇

News Bytes 🗞️

• Elon Musk is suing OpenAI, alleging the organization has strayed from its original mission of developing AGI for the benefit of humanity. The lawsuit accuses OpenAI of becoming a closed-source subsidiary of Microsoft, prioritizing profit over ethics.

  • 💡 Our take: Elon Musk is suing OpenAI, but not only for justice. He's making people talk about it, using controversy to grab attention. OpenAI changed its goal and joined Microsoft, but only after Musk left. Sounds fair enough. Yet Musk might just be stirring the drama for publicity reasons. He knows how to stir the pot, turning tech scenes into dramatic movie-like twists.

• OpenAI is defending against Elon Musk's lawsuit by stating he only contributed $45 million, not as much as initially promised, and wanted to either connect OpenAI with Tesla or take over. OpenAI revealed past emails showing Musk's agreement that, while advancing in AI, it was fine to diminish openness and share the resulting benefits of AI instead.

  • 💡 Our take: The exposed email conversations indicate Musk was in the loop about the organization's priorities, further suggesting the lawsuit is more about gaining publicity. His failed bid to align OpenAI with Tesla kindled his own AI ambitions with Grok. What we find truly intriguing from their email exchanges is how OpenAI once speculated Google as its top-notch competitor and Musk's knack for demonizing competitors.

• AI robotics company Figure declared a substantial raise of $675M in a Series B funding round, achieving a $2.6B valuation. Figure and OpenAI are joining forces to develop advanced AI models for humanoid robots, while also leveraging Microsoft Azure for improved AI infrastructure and training. This injection of funds aims to accelerate Figure's commercial deployment timeline of humanoid robots, potentially transforming operations in sectors facing labor shortages or performing unsafe tasks.

  • 💡 Our take: Figure's latest funding and collaboration with OpenAI and Microsoft Azure might seem like it's gearing up to compete with Tesla's robots. But, creating smart robots isn't easy. Past trials by OpenAI and Apple's recent pull-out from self-driving cars (more on this in our previous newsletter) serve as reminders. On a side note, Microsoft has made another smart move, using its strengths in computing power, necessary to build AI products, to position itself as a key player in the AI landscape.

• Anthropic announces Claude 3 model family with three models offering impressive cognitive capabilities - Opus, Sonnet, and Haiku. Opus, the most advanced model, claims to outperform OpenAI's GPT-4 on specific benchmarks. It boasts upgraded capabilities in analysis and forecasting, the model can examine both text and images, with enhanced proficiency in following multi-step instructions, responding in multiple languages, delivering more expressive responses, and exhibiting a more nuanced comprehension of requests.

  • 💡 Our take: The spotlight often lands on the intelligence of these models, but we think their speed and efficiency deserve equal attention. Particularly, the lightweight models, Sonnet and Haiku, stand out for their speed and practicality in product development. Plus, the pricing model, favoring extensive data input, makes it particularly appealing for RAG applications, where the ability to feed these models with extensive context is crucial. Claude's large context window certainly facilitates this, making it easier to draw relevant answers from a broad base of knowledge.

• The "Institute for Intelligent Computing" at Alibaba has developed an AI video generator, called EMO, which can transform still images of faces into expressive, singing actors. It can pair audio with matching facial expressions and movements, even mastering non-English phonetics, showcasing its potential in a demo of the Sora lady, created by OpenAI, singing "Don't Start Now" by Dua Lipa. Here are some demo videos on GitHub.

  • 💡 Our take: Alibaba's unveiling of EMO revealed that different companies work on similar breakthroughs simultaneously, keeping their developments under wraps. Once one makes the first move and reveals a new development, in this case OpenAI with Sora, it urges the others to step forward with their own advances.

• Google's DeepMind has developed an AI model named Genie, which has learned to convert images into interactive, playable environments by watching 30K hours of video games. Despite current limitations like slow frame rate and the tendency to 'hallucinate' unrealistic scenarios, Genie shows great promise, with researchers seeing it as a stepping stone towards robust world modeling in AI and a potential tool to simulate diverse real and imaginary environments.

  • 💡 Our take: Genie’s AI-driven environment simulation might just be the unconventional playground future robotics train in. This data generation mechanism could reimagine complex training techniques. A valuable breakthrough in a data-hungry sector.

• Adobe Research has introduced Project Music GenAI Control, a generative AI music generation and editing tool that allows users to create and fine-tune music based on text prompts. This tool aims to revolutionize how creators craft audio for their projects, offering deep control and customization similar to Photoshop for images. The collaboration with researchers from UC San Diego and Carnegie Mellon University highlights the advancement in AI technology for creative applications.

• Abu Dhabi's International Holding Co. is adding an AI-powered observer, Aiden Insight, to its board to enhance strategic decision-making by processing vast amounts of data. This move underscores the UAE's commitment to leveraging AI technology for business innovation and growth.

• Google Cloud and Stack Overflow have partnered to integrate AI-powered features into the developer knowledge platform, aiming to provide developers with better coding assistance capabilities and access to validated technical knowledge. This collaboration signifies a joint commitment to leveraging AI responsibly and empowering developer creativity through the integration of Google Cloud's AI technology with Stack Overflow's community-vetted knowledge.

Educational Pill 💊

RAG explained

RAG, or Retrieval-Augmented Generation, is like a clever robot that combines the smarts of a search engine with the creativity of AI to provide answers. Before it responds to a question, it quickly scans through a vast library of documents to find relevant information. Then, using this info, it crafts an answer that's not just accurate but also rich in detail. This makes RAG incredibly useful for creating responses that need up-to-date knowledge or specialized expertise, making it perfect for applications like chatbots, virtual assistants, and content creation tools.

LOLgorithms 😂

When TV shows meet reality.

From our community 🤝

Yesterday, we officially launched our new initiative, the AI Coach Program. It's designed for past students from our Master in Prompt Engineering program who have made significant contributions to our community since they joined. They will assist in supporting new students as they create their AI prototypes during our course. We're happy to have people on board who are eager to share their knowledge and help others achieve their goals. 🫶

That is the end of our newsletter.

Remember, if your company is looking to implement AI technologies, we also offer customized corporate training.

See you next week. 👋