For the uninitiated, Wolfram Alpha is a unique online service that answers factual queries directly by computing the answer from structured data, rather than providing a list of documents or web pages like you would get from a search engine like Google.
Like ChatGPT, it's designed to handle natural language inputs, meaning you can ask it questions just like you would ask a human. However, it is powered by the Wolfram Language, a highly advanced computational language. Wolfram Alpha can solve equations, deliver factual information, make graphs, and even generate visual representations for a wide variety of topics, including mathematics, physics, engineering, geography, history, and much more.
In a nutshell, it's like having a pocket-sized polymath ready to provide expert knowledge on a vast range of topics. So what would happen if you could combine it with the power of large language models (LLMs) like GPT-4?
That is the exact question that Stephen Wolfram is looking to answer. In his latest exploration, he introduces LLM Functions to the Wolfram Language. This allows these AI models to be routinely used as components within a larger framework, with their outputs utilized as inputs for other functions, a move that significantly broadens the scope of possibilities for computation and AI implementation.
