Voice as an interface

The smartphone was born in 2007. At the time it was a revolution to reach your entire digital world from your pocket, through the disruptive interface of the touchscreen. Having such a vast library of apps was an innovation in itself, as was breaking free of the keyboard and stylus in favour of that revolutionary interface, the multi-touch screen.

Today, smartphones are no longer an innovation. The industry has plateaued in this physical format.
The fair question is what the next global hardware format will be.
Right now it's laborious to hunt down the right app and fill in every action by hand.

Major technological breakthroughs have brought us LLM chat, like ChatGPT or Gemini. A merger with the older voice assistants — the Siri of that era, or Google Assistant — is worth imagining, and would make the experience genuinely powerful.
It would move us on from robotic commands like: "Send a text to Marie: 'I'll be there in 5 minutes'." And it would let you address your assistant the way you'd address a person, in ordinary spoken language: "Let Marie know I'm 5 minutes away."

Push the thought further and you can imagine no longer needing individual apps and graphical icons to get things done on your phone — going instead through a voice launcher alone, one that would carry out even your most complex actions, building a dedicated app for the task if that's what it takes. The latest development here is OpenClaw, which shows how far such software can go. Turning that into a reliable product built into the phone would be a major innovation.