Use ChatGPT Voice for More Natural Conversations (Full Transcript)

[00:00:00] Speaker 1: You can actually have a full conversation with ChatGPT, interrupt it, ask follow-up questions, brainstorm ideas, practice a presentation, learn a language, or even use it like your own personal tutor. And it feels much more like talking to another person than using a traditional chat bot. I'm Elizabeth, and in this video, I'll show you exactly how to use ChatGPT Voice, how to customize it, and most importantly, some of the best ways you can actually use it. To get started, open ChatGPT. You can use Voice from the ChatGPT app on iPhone or Android, or right here on the web. Down in the message box, look for the voice icon. One quick distinction that's worth understanding, voice is different than dictation. With dictation, you speak your prompt, ChatGPT converts what you said into text, and then you can review or edit that text before sending it. Voice is designed for a live conversation. So if you just want a faster way to type something, use dictation. But if you want an interactive back and forth, use Voice. The first time that you use Voice, ChatGPT may ask for permission to access your microphone. So go ahead and click Allow. And then once you're ready to begin your conversation, select Start Voice. What's a simple dinner idea that would include chicken, rice, and broccoli? And give me all of the ingredients that I would need for this.

[00:01:32] Speaker 2: All right, a super simple one is a one-pan chicken, broccoli, and rice bake. You toss bite-sized chicken pieces with a bit of olive oil, salt, pepper, maybe some garlic powder, spread it on a sheet pan with broccoli florets, then roast everything till the chicken is done and the broccoli's crisp tender. Meanwhile, cook some rice. Seriously, whatever you've got. White, brown, jasmine, all fine. And serve it all together. Ingredient-wise, you'd need boneless chicken.

[00:02:01] Speaker 1: Instead of ChatGPT just giving me a text response, it actually talks back to me. At the same time, I can follow along with the response right here on the screen. So this is useful even when the answer includes something that I might want to read or reference later. Where the voice becomes much more interesting is when you stop treating it like a traditional voice assistant. You don't have to ask one perfectly structured question at a time. You can talk naturally. I'm planning a three-day trip to San Diego with my family. We like beaches, good food, and activities for kids. What would you recommend?

[00:02:40] Speaker 2: Right, let me just think this through. I'm just checking current info, and I'm gonna lay it out neighborhood by neighborhood so you're not in a car the whole time. Right, here's a family-friendly three-day plan.

[00:02:57] Speaker 1: Actually, I don't want to start in La Jolla at the beach. I'd like one day where we don't do the beach at the start of the vacation.

[00:03:08] Speaker 2: Got it. Let's make day one Balboa Park with a San Diego zoo. It's an easy full day, kid approved, no beach required. Then day two, do La Jolla.

[00:03:19] Speaker 1: Actually, for day one, is there a better indoor activity that we could do? It doesn't look like the weather is going to be that good.

[00:03:28] Speaker 2: All right, just a sec, let me tweak that. If the weather's looking dodgy, I'd do the Fleet Science Center in Balboa Park. It's almost all indoors, and it's great for kids. You've got hands-on exhibits and a giant dome theater. You could do the morning there.

[00:03:47] Speaker 1: In this case, as we were having a conversation, I was interrupting the chat assistant to adjust the itinerary based off of my preferences as well as the weather. And you don't have to wait for a long answer before changing direction. You can interrupt, ask follow-up questions, or clarify what you want as you go. Can you tell me more about what to expect at the Fleet Science Center?

[00:04:13] Speaker 2: Well, think of it as a very interactive space. So you're not just looking at things.

[00:04:19] Speaker 1: Kids are really- Can you keep your answer really short unless I ask for more detail?

[00:04:25] Speaker 2: Sure. It's a hands-on science museum with a big IMAX dome theater. Kids love it.

[00:04:33] Speaker 1: Anne, what are the hours?

[00:04:36] Speaker 2: Daily 10 till five.

[00:04:38] Speaker 1: Voice gets especially useful on your phone. Go into settings, then voice, and you'll find a few additional options. One I particularly like is background conversations. When you turn this on, you can continue talking with Chat GPT while you're using another app or even while your phone is locked. That means you can brainstorm while walking, ask questions while cooking, or talk through your day's priorities without constantly staring at your phone. While we're in settings, you can also change how Chat GPT sounds. Within settings, select voice, and then you'll see the different voice options here. Try a few and choose whichever you prefer.

[00:05:20] Speaker 2: Yo, how's it going? I've got time to talk if you do.

[00:05:24] Speaker 3: Hey, what's up? Feel free to ask me about whatever's on your mind.

[00:05:29] Speaker 4: Hey there, how's your day going?

[00:05:32] Speaker 1: You can also set your preferred language. This can be especially helpful if Chat GPT occasionally misunderstands what you're saying or if you regularly speak a language other than English. One of my favorite uses for voice is brainstorming because sometimes ideas come much faster when you say them out loud instead of trying to type everything. I want to make a video about Microsoft Excel for beginners. Help me brainstorm five topics people would actually find useful. Chat GPT is giving me some ideas, and now we're developing the idea together.

[00:06:09] Speaker 3: Okay, how about these? Number one, Excel basics in 10 minutes. So think the ribbon, cells, simple formulas. Number two, five formulas you'll use every day. Some if, VLOOKUP or XLOOKUP.

[00:06:25] Speaker 1: I like number four. Can you give me a few different ways I could structure that video?

[00:06:31] Speaker 3: Oh yeah, for sure. Okay, how about option one, project first, then explain. So you build a simple budget live, then you circle back and explain each piece.

[00:06:43] Speaker 1: Option two, you could do this with presentations, business ideas, lesson plans, vacation, home projects, really anything where you want someone to bounce ideas off of. You can also tell Chat GPT to hold its response while you think. Then I can basically think out loud, and once I'm done, ask Chat GPT to organize everything I just said. We're used to using Chat GPT to summarize data for us, whether that's in an Excel spreadsheet or just reading from a PDF. But with Chat GPT Voice, instead of just reading your summary, you can have a conversation about it. Can you walk me through this summary, but only walk me through the really important parts?

[00:07:35] Speaker 4: Sure, I'll pull out the high-impact headlines only and give you a simple walkthrough. Just a sec, let me double-check the numbers. Here are the four headlines that really matter from September. First, total sales were about $1.3 million, but the sheet actually labels them as quarterly sales. So that may not all be from September. Second, Alice Johnson is your top performer and Ella Martinez is close behind. Together, they're about 50% of all sales.

[00:08:04] Speaker 1: Actually, can you stop for a second and tell me what the difference in revenue between those two salespeople are?

[00:08:14] Speaker 4: Sure, I'll calculate the gap quickly. Alice brought in $31,480 more than Ella, about 10% more.

[00:08:23] Speaker 1: Another really useful feature is role-playing. Let's say you have a job interview coming up. You could say, pretend you're interviewing me for a marketing manager position. Ask me one question at a time, and after each answer, give me some feedback. Now, Chat GPT becomes your practice interviewer. You can do the same thing with a presentation. I'm going to practice the opening of my presentation. When I'm finished, tell me where I sound unclear or where I could be more concise. Or you could practice a difficult conversation. Pretend you're my manager and I'm asking you for a raise. Push back on some of my arguments so I can practice responding. This is where voice can be much more useful than simply typing back and forth because you're actually practicing the conversation out loud. Voice is also fantastic for learning a language. For example, I could say, let's have a beginner-level conversation in Spanish. Speak slowly, and if I make a mistake, explain it to me in English.

[00:09:34] Speaker 3: Esta bien. Vamos despacio. Yo te corrijo en inglés si hace falta. Hola, ¿cómo estás hoy?

[00:09:43] Speaker 1: Hola, ¿cómo estás hoy? Estoy bien.

[00:09:48] Speaker 3: Muy bien. A tiny correction in English, it's, ¿cómo estás hoy? With an accent on estás. Bien. ¿Y tú? ¿Cómo te sientes hoy?

[00:09:59] Speaker 1: Now I can actually practice speaking instead of just memorizing vocabulary. I could also say something like, how would I order a coffee in Italian?

[00:10:12] Speaker 3: You can keep it simple. Un café, per favore. That's a coffee, please. In Italy, that usually means espresso. If you want it with milk in the morning, you might say, un cappuccino, per favore.

[00:10:29] Speaker 1: Then I could ask Chachi PT to repeat the phrase more slowly or have it role-play the person working at the cafe. Because it's conversational, I can keep adjusting the difficulty as I improve. And that's how to use Chachi PT Voice. Thanks for watching. And make sure to subscribe to our newsletter for more tips and tricks like this.