Today’s top AI Highlights:

1. Nvidia is now the Most Valuable Company in the world

2. The open source trap: How companies claim openness without delivering

3. FinetuneDB’s new updates simplify dataset management and open source model fine-tuning

4. OpenAI co-founder Ilya Sutskever starts new company to build Safe Superintelligence

& so much more!

Read time: 3 mins

Latest Developments 🌍

AI Chipmaker Surpasses Microsoft and Apple 🥇

Nvidia has become the world’s most valuable company, surpassing Microsoft. This comes just two weeks after Nvidia overtook Apple as the second most valuable company. Nvidia’s stock price rose by about 3.5% on Tuesday, giving the chipmaker a market capitalization of over $3.33 trillion. Microsoft’s market cap, on the other hand, fell slightly, standing at nearly $3.32 trillion.

Nvidia’s stock has seen impressive growth, rising over 215% in the past year and over 3,400% in the past five years. The company has been a major driver of the S&P 500’s record highs in 2024, with its stock gains contributing about 1/3rd of the index’s year-to-date rise. Nvidia’s rapid growth in value shows the growth potential of the AI industry and it is just the beginning.

Open Weight, Not Open Source 🗝️

You might think that all “open source” AI models like Meta’s Llama and Google’s Gemma are truly open but the real scenario is different. A new study explores the openness of many generative AI models, revealing a trend of “open-washing” by companies seeking to capitalize on the benefits of open source without truly delivering on its promise. The researchers analyzed over 45 models, including both text and text-to-image models. This study reveals a disturbing lack of transparency surrounding training data, documentation, and even the ability to customize models.

Key Highlights:

1. Many open source models are actually “open weight”: They provide access to the trained model but withhold vital information about how it was built and trained. This allows companies to claim openness while maintaining control over the underlying technology and avoiding scrutiny of their models’ biases or ethical implications.

2. Lack of transparency regarding training data: Companies often avoid disclosing details about the data used to train their models, concealing the presence of biased or inappropriate content, and limiting the ability to understand model performance and build their own versions.

3. Open-washing: The EU’s AI Act, which aims to regulate AI development, could inadvertently be used to legitimize open-washing if its definition of open source is not clear and comprehensive. The authors argue that relying solely on licensing to determine openness could allow companies to claim openness while avoiding accountability and transparency.

One-Stop Shop for Open-Source LLM Development 🤖

FinetuneDB, a platform to build and deploy custom AI models, has just released a powerful new update that makes it easier than ever to work with open LLMs. This update introduces a streamlined platform for managing datasets and fine-tuning models, making it a one-stop shop for your AI development needs. The new release features a robust dataset manager with version control, automatic validation, a visual function calling editor that makes it easy to work with data, and support for fine-tuning and serving of popular open-source models.

Key Highlights:

1. Streamlined Dataset Management: The new Dataset Manager helps you keep track of your datasets, easily iterate on them, and ensure the quality of your data with automatic validation. You can also take advantage of the visual function calling editor to make data transformations without needing to write code.

2. Open-Source Fine-Tuning and Serving: FinetuneDB now offers the ability to fine-tune and serve popular open-source models like Llama 3. This makes it easy to experiment with different models and find the best fit for your needs.

3. Cost-Effective and Efficient: The new platform can help you reduce costs by allowing you to use open-source models that are often comparable to or even outperform proprietary alternatives. Plus, you can take advantage of optimized fine-tuning processes to improve the performance of your models.

OpenAI Co-founder Starts Safe Superintelligence ⛑️

OpenAI co-founder Ilya Sutskever just launched the world’s first AI lab to build Safe Superintelligence. Safe Superintelligence Inc. (SSI) is dedicated solely to developing a safe and robust superintelligence. Sutskever, alongside Daniel Gross and Daniel Levy, believes the problem of safe superintelligence is the most pressing technical challenge of our time.

SSI aims to advance capabilities alongside safety measures, ensuring safety always remains ahead of progress. The company emphasizes its singular focus on superintelligence, stating its business model and structure are designed to avoid distractions and short-term commercial pressures.

With offices in Palo Alto and Tel Aviv, SSI is recruiting the best talent to make this vision a reality. If you’re looking to do your life’s work in AI, this is a great opportunity!

Tools of the Trade ⚒️

1. Eleven Labs Voiceover Studio: Create voiceovers and virtual podcasts with multiple speakers and different sound effects within the same studio. Eleven Labs’ text-to-sound in this studio gives enhanced control and editing features to make the perfect voiceovers and podcasts.

2. MATLAB GPT: Assists with MATLAB programming, whether you need help with syntax, code optimization, project planning, or quick answers to specific questions. Share your MATLAB challenges, and it’ll provide clear guidance and practical examples to help you succeed.

3. Spirals: Automate 80% of repetitive writing, thinking, and creative tasks, incorporating your voice and style, and repurpose your content into various formats. You can train Spiral on examples to ensure outputs meet your preferences accurately.

4. Awesome LLM Apps: Build awesome LLM apps using RAG for interacting with data sources like GitHub, Gmail, PDFs, and YouTube videos through simple texts. These apps will let you retrieve information, engage in chat, and extract insights directly from content on these platforms.

Hot Takes 🔥

1. Ilya Sutskever's new company is guaranteed to succeed, because superintelligence that is never achieved is guaranteed to be safe. ~
   Pedro Domingos
   

2. Predictions About SSI - Safe Super Intelligence
   Day 1 Valuation - $10B!!
   Funding: $1B+
   Time to close round: 1 day
   Time to Revenue - Never, since money will be moot once we have ASI! 😀
   Estimate year for ASI - 2027 ~
   Bindu Reddy

Meme of the Day 🤡

That’s all for today! See you tomorrow with more such AI-filled content.

Real-time AI Updates 🚨

⚡️ Follow me on Twitter @Saboo_Shubham for lightning-fast AI updates and never miss what’s trending!

PS: We curate this AI newsletter every day for FREE, your support is what keeps me going. If you find value in what you read, share it with your friends by clicking the share button below!