Voice Over Text: ElevenLabs Bets on the Death of the Keyboard

Let's be honest: typing on a keyboard, especially a glass one on a smartphone, is a crutch. We invented this way of talking to machines because machines were dumb and understood only exact commands. But at Web Summit Qatar, ElevenLabs CEO Mati Staniszewski said out loud what is hanging in the air in Silicon Valley: this crutch will soon be sent to the dustbin of history. Voice is the real, native interface of the future.

The statement didn't come out of nowhere. Look at what is happening around us. For years we trained to be good "prompt engineers," choosing words for chatbots, but the industry is sharply changing direction. OpenAI is rolling out voice modes that can giggle and interrupt you, and Google is integrating Gemini into Android so that you talk to your phone instead of tapping it. Apple, according to rumors, is preparing a Siri that will finally stop being a punchline. Staniszewski simply highlighted an obvious trend: the barrier between a person's thought and a machine's action should disappear.

Why does this matter right now? Voice assistants used to be dumb. They heard words but didn't understand context. Now that a powerful LLM sits behind the "ears," the situation has changed. ElevenLabs, which started as a voice-cloning startup (and made a lot of noise in the world of deepfakes), now positions itself as an architect of this new world. Its technology lets AI do more than drone through text: it can convey emotion, pauses and intonation. That removes the "uncanny valley" effect and makes talking to a machine unnervingly natural.

What's interesting is how this changes our interaction with hardware. If voice becomes the primary interface, why do we need 6.7-inch screens? That opens the door for the very wearable gadgets that are still flopping in sales (hello, Humane AI Pin and Rabbit R1). Perhaps the problem was not the form factor, but that the software wasn't yet ready for full conversation. ElevenLabs and its peers are trying to solve the problem of response delay, or latency. To replace the keyboard, AI must answer instantly, like a live interlocutor, rather than spend three seconds thinking about the weather.

Privacy questions remain, of course. Talking to an AI on the subway or in an open-plan office is a poor idea. But at home, in the car or through earbuds, it is becoming the new norm. We are returning to where human communication began, the spoken word, except that now our conversation partner will be a silicon chip that never tires and knows everything in the world.

The takeaway: interfaces are becoming invisible. If the forecast comes true, app and website design will move into the background, and the main competitive advantage will be how "human" and smart your service's voice is.