VOICE will replace WRITING

Typing is dead and the future of content is spoken. I had this realization right now. I'm doing some work here on my laptop and I just realized—no one wants to write anything anymore, right? No one wants to sit down and type everything because AI can do that easily.

The Core Insight

So you just prompt AI saying, "Hey, write this for me." It's that easy. So the future of content is spoken and I've been thinking of this for a long time. And not only the future of content, but the future in general. That's why I'm extremely bullish on voice AI.

First one I just mentioned—typing is extremely slow and we all know AI, ChatGPT, Claude, all of them can do it super fast. So that's why we read some content nowadays and we're on LinkedIn, on Twitter, whatever, and we're like, "Holy fuck, this is AI." There is no human element in here because the machine has no soul.

The Machine Has No Soul

What's the difference between when you are typing and when you are speaking something? You're basically transmitting the same message, but you spend way less time by speaking. Correct.

If you've probably gone on LinkedIn and tried to create a post about anything—maybe you went to an event and you wanted to share with people—you're going to spend like half an hour, 30 minutes writing the post and no one's going to see it.

That's why I'm extremely bullish on voice. You can communicate the same message but way faster. I'm telling you all of this right now.

The Interview Revolution

Another big point here: I think the future of content is going to be interviews, especially for brands. With an interview, think of this—you're interviewing the CIO of, I don't know, Starbucks. If that content is living only on YouTube, the content is dead. No one else can access the interview with the CIO of Starbucks.

Just imagine how much value, how much information is packed in one hour conversation with a C-suite level from Starbucks. Lots of value, lots of content, lots of everything, lots of information that people should have access to.

But guess what? You can't just find new information from YouTube videos. Let's say you're having a conversation with ChatGPT with any LLM and you just want to know something specific about Starbucks. If you ask the LLM, the LLM will just have access to the content that someone went online and typed.

The Transcription Game Changer

So this video, this interview on YouTube with the CIO of Starbucks is gone forever. No AI has access to that video. Now bear with me. Imagine the interviewer decides, "You know what, now it's time for me to get the transcription and publish it on a website," his podcast website.

Most podcasts have been doing this nowadays. This is the first time I had this realization—I wanted more information about a mobile app and the founder. I asked the AI to do a deep research for me and it found content. It found an interview with the founder and someone else.

That was the first time I thought, "Holy fuck, this is a great use case for transcriptions and YouTubes and interviews" because sometimes you don't want to listen to an hour of conversation just to get one piece of information. Of course there is value in watching the entire interview, but sometimes you just don't want to—you just want that one specific thing.

The AI found the information because it was on a transcription. It was something like "our first 100 users were all family and friends" or something like this. The AI found the information because it was on a transcription.

SEO vs GEO

Now you were able to understand what I mean by "content voice is the future." Just like SEO kind of died to get replaced by GEO—the generative engine optimization, whatever. GEO is the SEO for LLMs, right? And this is where transcriptions play a huge part.

Why Google Will Win

This is why Google is gonna win this shit because OpenAI basically has all the content available on Wikipedia, Reddit—they can scrape all that stuff—but they don't have access to YouTube.

If you go right now and try to copy this video link, go on ChatGPT and ask "what's this guy talking about" and paste the link, ChatGPT has no idea what I'm talking about. But now go on Gemini—even though it's the bad model, 2.5 Flash—it's capable of telling you what's this video's transcription and what am I talking about in this video.

This is fucking crazy. This is just crazy. This is probably where Google is going to win in the voice AI race because at some point we're going to all be tired of typing in chatbots and stuff and it's just going to move on to conversational AI.

The Conversational Future

You're going to be talking to the AI. Right now, for example, I see Google has here a microphone where you speak to the chatbot and it transcribes your speech and then you can just send. Probably the future there will be no chatbot interface—it might just be pure microphone.

Google has great, incredible advantage because they have how many movies are on YouTube you can buy or rent. How many interviews, how many videos—the whole world is on YouTube and everything you need for a model, conversational model to be well conversational is data to be trained on.

I remember if you try ChatGPT voice mode, you go on your phone and you use the voice, it's pretty good. Google has access to a huge library of content called YouTube when OpenAI doesn't.

The Economic Reality

This is my take on why conversation AI and voice is going to be the future and why you should be preparing for that because no one wants to type anymore. Typing is extremely time-consuming, especially if you're writing something longer than just a message to tell your friends "oh I'm here"—that doesn't take any time.

But if you're trying to create something more dense, then typing just consumes a lot of time. So be prepared for the future because the future is going to be conversational.

In 5 years from now, probably less, we are going to be talking to AIs. When you go to a drive-thru, when you go talk to a cashier, there will be no cashier. There's going to be an AI and maybe a security guard just to make sure you're not stealing.

The future more and more is going to be you talking to a robot because that's what makes companies replace humans for cheaper, and all companies are looking for profit, looking for more money. If they have the opportunity of instead of paying 20 bucks an hour for a minimum wage employee, they pay three bucks an hour for an AI agent, that's worth it for them.

The Historical Pattern

If we get to the point where AI is extremely cheap, this will happen. It will. The same way it happened with internet and Wi-Fi and networks in general 25 years ago, 30 years ago, when the web was released, it was extremely expensive. Nowadays it's extremely cheap.

This is going to be the same thing that's gonna happen to AI because it's the same thing that happened to voice calls 30, 40, 50 years ago. To make a phone call, it was expensive as balls. Today I'm here in Canada. I can just call my parents in Brazil and see them and I don't pay anything.

This is my take. I think in a few years from now, AI is going to be extremely cheap. The cost to produce is going to go all the way to the bottom and it's going to be accessible and you're going to see lots of people being replaced because it's cheap.

This is my take. Let me know what you think. If you disagree, awesome. If you agree, even better.