Data is Gold!🌟

Hey Readers, Happy Sunday!

In today's edition, I'll cover the following:

  1. My concept of the week: Building Explainable AI

  2. AI news update: Reddit, Twitter, and StackOverflow's Data Gold Mine, Microsoft's AI Chip, and New Open Source LLMs from Stability.ai 

  3. My new Interview: A Chat with Tanmay, AI Developer, and TED Speaker

Let's dive in! 🀿

πŸ“š My AI post of the Week: Building Explainable AI

AI becomes increasingly prevalent in the business world, there is a growing need for transparency and trust in these systems. For AI to be adopted and embraced by business professionals, it must be explainable, understandable, and trustworthy. This is where Explainable AI (XAI) comes into play. In this blog post, we will discuss the importance of XAI and why it is critical for building trustworthy AI systems in the business environment.

πŸ—žοΈTop News of the Week

Reddit, a social media platform with over 57 million daily active users, is planning to begin charging companies for access to its application programming interface (API). The API allows developers to access and process Reddit's vast amount of user-generated content, which has become a valuable resource for artificial intelligence (AI) developers. Stack Overflow, a popular online forum for programmers, is planning to charge AI developers for access to its data. The company says that it is doing this to ensure that its users are compensated for their contributions and to improve the quality of AI models.

In recent years, companies like Google, OpenAI, and Microsoft have been using Reddit and Stackoverflow data to train their AI systems. These systems are capable of performing a variety of tasks, such as generating text, translating languages, generating code, and writing different kinds of creative content.

However, Reddit, Twitter, and Stack Overflow have not been compensated for the use of their data by AI developers. They say that this is unfair to its users and that it is contributing to the spread of misinformation. By charging AI developers for access to its data, Stack Overflow, Twitter, and Reddit hope to generate revenue and ensure that their users are compensated for their contributions.

Microsoft is developing its own AI chip, codenamed Athena, to reduce costs and compete with Google. Athena is currently being tested internally and is expected to be available in Azure services next year. The high demand for AI services is driving up the cost of running AI systems, such as ChatGPT, which costs OpenAI $700,000 per day to run. AI companies, including Microsoft, need to buy GPUs from the likes of NVIDIA in bulk orders to help maintain their AI services. Athena could help Microsoft reduce its reliance on NVIDIA and lower the cost of running its AI services.

Stability AI, a company that develops open-source artificial intelligence (AI) tools, has released a new language model called StableLM. The model is available in two versions: 3 billion and 7 billion parameters, with 15 billion to 65 billion parameter models to follow.

StableLM is trained on a new experimental dataset that is three times larger than the dataset used to train GPT-3. The model is able to perform surprisingly well in conversational and coding tasks, despite its small size.

Stability AI is also releasing a set of research models that are instruction fine-tuned. These models are intended for research use only and are released under a noncommercial license.

The company is committed to developing AI technology that is transparent, accessible, and supportive. StableLM is an example of this commitment. The model is open-source, so researchers can inspect, use, and adapt it for their own purposes. It is also designed to be accessible to everyday users so that they can run it on local devices.

Stability AI believes that AI should be used to benefit everyone, not just a select few. The company is working to make AI technology more transparent, accessible, and supportive so that everyone can benefit from it.

πŸŽ™οΈ Interview of the week: Talking to an AI expert, TED & Keynote speaker, Tanmay Bakshi

I had a great talk with Tanmay Bakshi, TED & Keynote speaker, Google Developer Expert for Machine Learning, IBM Advisory Software Engineer, and a huge fan of SpongeBob SquarePants πŸ˜„. He builds technology that augments human lives and aspires to empower 100,000 people with the technology they need to change the world. In the last 8 years, he addressed over 200,000 executives, developers, and students in over 30 countries, namely Apple, Walmart, Microsoft, United Nations, KPMG, Linux Foundation, and IBM. His work has been featured in Forbes, Bloomberg Businessweek, The Wall Street Journal, The New York Times, CNBC, and more. Enjoy!

Thank you for taking the time to read! When you subscribed to this newsletter, my commitment to you was to be your helpful guide in navigating the world of AI. How did you find today's issue? Please reply to this email with your feedback and any suggestions for future topics.

See you next Sunday!

Armand 😎

Reply

or to participate.