From thumbs to voice: the next AI interface revolution

Article link: https://open.substack.com/pub/johanos...

The article argues that we are entering a major interface shift: after decades where our thumbs and fingers dominated how we “spoke” to computers through keyboards and touchscreens, AI is making voice the next default way to direct technology. Voice is faster, more natural, and better aligned with how humans think, especially now that large language models can interpret messy language and infer intent. In this view, voice prompting is not just a convenience feature; it represents a new phase in human–computer interaction where we move from operating software through menus and commands to delegating outcomes through conversation.

But the article also stresses the big catch: voice-first AI still does not serve the world’s languages equally well. Accents, dialects, code-switching, and low-resource languages are often less well supported because of data and model limitations, meaning some people experience voice as effortless while others must slow down, switch languages, or abandon voice entirely. It notes that while ChatGPT currently offers one of the smoother voice experiences in practice, the broader point is that language support must become a core measure of quality, not a footnote. The conclusion is that the voice era will only be a true leap forward if it becomes genuinely inclusive, with users and organisations also learning the discipline of verification as voice makes AI feel more human and trustworthy.