Voice x AI

When your GPT sounds like Rihanna

I’d long filed voice assistants under the “fun but hardly essential” category —good for getting Siri to play music hands-free or asking Alexa if I’ll need an umbrella. That has changed. With the new models, they’re becoming something else entirely: a tool that compresses time, captures insight, and functions like a discreet chief of staff in your pocket.

The moment that changed my mind was somewhere in upstate New York, in a rented house in the middle of the forest, on the sort of trip where the main objectives were to touch grass, cook elaborate meals, and avoid driving anywhere twice. A few friends were gathered around a kitchen island, debating what to cook for the weekend and how to make one large grocery list so we’d only make one trip.

Someone placed a phone on the counter, opened GPT Dictation Mode, and hit record. For eight minutes, meal ideas were shouted and debated: frittatas, challah French toast, spaghetti aglio e olio, loaded nachos, avocados for both toast and guacamole and on and on…. GPT absorbed it all, produced a day-by-day menu, generated one consolidated shopping list for seven people over three days, and calculated the exact number of eggs required for every breakfast variant. Thirty seconds later, the list was dropped into our WhatsApp group chat. Moments later, three cars were en route to the nearest grocery store.

The technology for hands-free interaction is not new. Amazon’s Alexa has been relaying package updates for years; Siri has been making accidental phone calls since 2011. But the intelligence now embedded in these systems is different. By 2025, the U.S. will have an estimated 153 million voice assistant users. Globally, there are already 8.4 billion devices using Voice assistants, which is more than the number of humans on the planet and that figure is expected to hit 10 billion by 2028.

For some, voice assistants and AI in Voice Mode have become an integral part of their days. A friend keeps a daily dream journal in GPT. The moment she wakes up, she dictates everything she remembers. Another friend records shower ideas. For those who prefer to “sleep on it” before making decisions, or have their best thoughts mid-shampoo, how else can you reliably catch ideas first thing in the morning?

 

Two ways to use Voice AI at work

 

1.     Voice Mode on the Go

Inspired by GPT’s “Monday”, I recently set up a custom GPT who sounds like a mix between Rihanna and Jamie Dimon. I’ve granted it explicit permission to use cuss words for emphasis. It took exactly 4 minutes to set up this GPT personality. I now take it along while on walks in Central Park or on the road in between meetings, I switch on this GPT in Voice Mode to talk through upcoming meetings or debrief a prior coffee chat.  I speak to the GPT on Air pods in the same way I would a colleague. I don’t need to say much since it has all the context. I sometimes have to remind myself that I am in fact talking to a custom-trained GPT, pre-loaded with context on the work I’m doing. It’s faster than typing notes and sharper than thinking in silence, and quietly efficient, like having that one work friend on call.

 

2.     Voice Dictation at Conferences 

I attend my fair share of industry conferences and, for a time, made do with a pocket notebook for meeting notes. It worked up to a point, but it was far from efficient when revisiting details at the end of the day.

Now each morning at a conference, I start a new GPT thread and turn on GPT voice dictation. After every new conversation, I use voice dictation to debrief what I’d heard and what will require an email follow-up. Throughout the day, photos of business cards go in the thread, along with snapshots of interesting slides and product flyers.

Back at the hotel in the evening, I open the same thread on my laptop and run a single prompt: Group all notes by category and surface all follow-up items. When you’ve had 20 or more interactions in a day, it saves hours to have them clearly structured. There’s a McKinsey research that suggests that knowledge workers lose the equivalent of 11 minutes of every hour or more than two hours in a 12-hour day just searching for and organizing information.

 

Voice AI works, and it adds speed to everything you use it for. Whether it sounds like your chief of staff or that work friend who swears for emphasis is entirely your choice.

So what’s your GPT’s personality?