How Generative AI is Changing the Way Humans Talk to Machines

Voice control of devices has been around for decades now; but the futuristic vision of conversational interactions with a computer portrayed in that first episode of Star Trek, in 1966, has yet to be fully realized. Speech controlled devices have required their users to learn their vocabulary and phrasing. Improvements have come - but they have been slow and incremental. More and more diversity of command phrasing has become possible with higher-end products. Yet, the underlying technology (the direct mapping of certain words or phrases to particular commands) has not changed. Until now. With the advent of generative AI, through facilitative technology like large language models, it is now possible for machines to probabilistically construct meaning out of dynamic human speech input. This means that the days of learning the machines vocabulary, syntax, phrasing, etc. are passing away. Soon all voice control will involve generative AI filters that predict meaning and map input intention to available commands. Additionally, these AI models allow for the generation of dynamic responsive language. So the days are soon coming when you won't just command machines with conversational speech, but they'll also respond to you with dynamic verbal feedback.
Comments