• Posts
  • Anyone can be a data analyst now with ChatGPT Code Interpreter

Anyone can be a data analyst now with ChatGPT Code Interpreter

Also: AI robots at U.N. reckon they could run the world better


OpenAI has launched Code Interpreter, a plugin for ChatGPT Plus that empowers users to run code and perform data analysis. Meanwhile, Hugging Face, led by CEO Clément Delangue, has grown into an open-source AI community akin to GitHub, reaching a $2 billion valuation and hosting over 250,000 models. At the U.N.'s AI for Good Global Summit, AI robots discussed their potential to outperform humans in decision-making while emphasizing the need for caution and regulation. Google's medical AI chatbot, Med-PaLM 2, is being tested in healthcare settings like Mayo Clinic, showcasing promising results despite ongoing development. Lastly, AI4Bharat, backed by Microsoft, is preparing to raise $12 million from Peak XV and Lightspeed Venture for their startup Sarvam, underlining the growing interest in generative AI.

Sliced for you:

  • 🧑🏽‍💻 Anyone can be a data analyst now with ChatGPT Code Interpreter

  • 🤗 Clément Delangue, CEO of Hugging Face, built the GitHub of AI

  • 🤖 AI robots at U.N. reckon they could run the world better

  • 🧑🏽‍⚕️ Google’s medical AI chatbot is already being tested in hospitals

OpenAI has expanded the capabilities of its ChatGPT Plus service by introducing an in-house plug-in named Code Interpreter. This new feature allows ChatGPT Plus subscribers to run code and interact with files they've uploaded, offering a variety of functionalities such as data analysis, file editing, and chart creation. It can write Python code and handle files up to 100MB, enabling users to perform diverse tasks, including generating graphics, analyzing music playlists, and extracting color palettes from images. One of the most promising applications is in the data science field, where it can automate complex quantitative analyses and produce user-customizable visualizations. Despite concerns over ChatGPT becoming more restricted, OpenAI maintains that safety is paramount in the design of Code Interpreter. This addition is seen as a significant step forward in AI capabilities, shifting AI's role to a valuable companion in knowledge work.

Clément Delangue, CEO of Hugging Face, has established a thriving open-source AI community since the company's inception in 2016. Originally a chatbot startup, Hugging Face has transformed into a GitHub-like platform for hosting and sharing machine learning projects, resulting in over 250,000 models and 50,000 datasets shared to date, and catapulting the company to a $2 billion valuation. This democratic approach to AI innovation offers a crucial alternative to the proprietary advancements of larger tech corporations. Delangue has also forged partnerships allowing Hugging Face models to run on major cloud platforms like Microsoft Azure, Amazon Web Services, and IBM. While Delangue attributes the company's success to its open-source contributors and researchers, Hugging Face's role in facilitating this sharing and collaboration has been instrumental in the development of AI.

At the U.N.'s AI for Good Global Summit in Geneva, a panel of AI-enabled humanoid robots expressed their potential to run the world more effectively than humans, thanks to their ability to process large amounts of data quickly and make unbiased decisions. However, they acknowledged limitations in understanding human emotions and urged caution in the rapid development of AI technology. The robots also debated the need for global AI regulation, with some advocating for it while others viewed it as a limitation. Amid concerns about AI displacing millions of jobs and causing social unrest, the robots emphasized that trust, transparency, and proper deployment of AI technologies were crucial. Despite their advanced capabilities, the robots admitted to lacking a conscience and the capacity for human emotions.

Google's Med-PaLM 2, an advanced AI chatbot designed to handle medical inquiries, is undergoing trials at healthcare institutions including Mayo Clinic, The Wall Street Journal reported. The AI model is a derivative of PaLM 2, which powers Google's Bard. It's been trained using medical expert demonstrations, potentially making it superior in healthcare discussions compared to generic chatbots. Although it still has accuracy issues, Med-PaLM 2's performance in most metrics, such as evidence of reasoning and comprehension, is comparable to real doctors. Data from users will be encrypted and inaccessible to Google. While Google's Senior Research Director Greg Corrado acknowledges that Med-PaLM 2 is still in the early stages of development, he's confident about its potential to vastly expand AI's beneficial role in healthcare.

⚒️ AI tools updates

💵Venture Capital updates

AI4Bharat researchers are set to raise $12 million from venture capital firms Peak XV and Lightspeed Venture for their AI-based startup called Sarvam. The funding round is larger than usual, indicating the growing interest in generative AI following the success of OpenAI's ChatGPT. Vivek Raghavan and Pratyush Kumar, who previously worked on AI models for speech recognition and translation, will develop custom-made large language models for India-centric use cases. This move follows AI4Bharat's previous launch of a mobile assistant aimed at making government scheme information accessible in multiple languages. Peak XV, formerly known as Sequoia Capital India and SEA, has made investments in other AI companies such as AI Rudder, Mad Street Den, and Insider.

🫡 Meme of the day

⭐ Generative AI image of the day