By Ben Ansell and Lisa Baxter

In 2023, ChatGPT took Artificial Intelligence into the mainstream. Now there's a bewildering choice of human-like chatbots to choose from. Generative AIs can produce pictures and video from a text prompt, and many websites and apps are now labelled "Powered by AI".

This new technology can do lots of things and tech companies have raised vast amounts of money from investors based on its potential.

But what is AI actually for?

Certain specialised AIs have a clear purpose. AlphaFold2 can predict how proteins fold-up and won its creators the Nobel Prize for Chemistry, and Google Translate is an AI with a purpose that’s clear from its name.

But so far there is no must-have or "killer" application for the Large Language Models and Generative AIs.

The future of AI is equally hazy. Will AI somehow lead to all-purpose "Artificial General Intelligence", autonomous robots or even machine consciousness? Or is this all just the stuff of fantasy and nightmares?

Rosalind Picard, Grover M. Hermann Professor of Health Sciences and Technology and director of the Affective Computing Research Group at the Massachusetts Institute of Technology Media Lab, is among the contributors to this episode. 

Rethink is a BBC co-production with the Open University.