A IA MUDOU DE NOVO: Essa Nova Categoria Vai Transformar o Mercado - Dots, Muse, Grok Bot...

Video authored by the channel shown, not by IANIA; all rights belong to the creator. Transcript and summary generated automatically; the original-language content is preserved.

Original language: English

Original video: A IA MUDOU DE NOVO: Essa Nova Categoria Vai Transformar o Mercado - Dots, Muse, Grok Bot… · Channel: Billy IA · 12:55 · Watch on YouTube

Topics: AI agents, Enterprise AI applications, Task and workflow automation, Multimodal models

Summary

The video argues that AI is moving toward voice-driven, interactive agents that can operate computers and complete complex tasks with less human intervention. The speaker contrasts this emerging direction with current workflows, where agents often need setup, corrections, and supervision. Using tools such as OpenAI’s Dots and Meta’s Muse as examples, he predicts that work may become much more conversational over the coming months and year. He illustrates the potential with a business workflow in which multiple agents build a sales page, manage leads, create reports and dashboards, and prepare marketing materials in parallel.

Key points

The speaker sees AI companies focusing on agents that can interact with computers and carry out tasks through conversation.
Today’s AI agents are not fully autonomous or seamless; people still need to configure them, intervene, and make adjustments.
The speaker believes voice-based interaction will increasingly resemble talking with employees who specialize in different tasks.
OpenAI’s Dots is presented as a promise of parallel agents that can perform tasks without constant mouse or keyboard input.
Meta’s Muse is described as integrated with phones and social networks, able to analyze information and execute tasks through voice requests.
The speaker expects voice-first AI work to become much more common, potentially within about a year.
He advises focusing less on individual tools and more on applying AI to improve business productivity, speed, and costs.
A proposed workflow includes creating a sales page, collecting leads, sending notifications, analyzing submissions, and producing a dashboard and marketing plan.
The speaker says many of these tasks can already be done, but argues that the market impact will grow as they become easier and faster to initiate by voice.

Chapters

Start	Chapter	Summary
00:01	A shift toward interactive AI agents	The speaker introduces Dots and describes a broader market trend toward AI that can interact with computers and perform tasks conversationally.
03:43	From the JARVIS idea to voice-based work	He says AI work may become more like interacting with JARVIS, while noting that current agents still require setup and human intervention.
06:26	Specialized agents and the changing workplace	The speaker discusses Meta’s Muse, imagines voice conversations with task-specialized agents, and urges viewers to focus on applying AI to their businesses.
08:32	An example of an AI-powered business workflow	He outlines a voice-requested workflow involving a sales page, lead handling, reports, a dashboard, and parallel marketing tasks.
11:17	The expected market impact	The speaker says this kind of parallel execution is the direction companies are pursuing and invites viewers to consider its effect on work and the market.

Quotes

It’s not 100% automated, it’s not 100% fluid the way people sell it. — 02:41

There will be several agents working, each performing a function — 05:54

tools aren’t the most important part — 08:32

You’ll simply make a request like that and the AI will execute everything in parallel. — 11:17

Concepts and entities

Dots, Muse, Claude Code, ChatGPT, Meta, JARVIS, agentes de IA, interação por voz, Google Drive

Full transcript (original language)

00:01 Look, I’m seeing a new market trend, and artificial intelligence is going to have an impact at a level that very few people are observing or even seeing, okay? Look, Chat recently launched its “dots,” right? This right here, these are the dots, right? The dots promise to work in a totally interactive way inside your computer and execute tasks as if it were a natural conversation with a human, okay? So, starting from

00:34 this, things are beginning to change, and that’s what I’m going to bring to you in this very quick video, okay? But for you to understand where the market is going with everything changing so fast in AI—yesterday there was the OpenAI event, and they announced several new things, including the dots, and all other LLMs, Claude, Gemini, Meta, everything is focusing on this. What I want to tell you is the following: pay attention, man. Look,

01:04 there is something that is really scaring me. It really surprised me after these changes here. I’m not saying this just to make an impact, either. Every time a new AI or a new model emerges, everyone starts making that joke: “Oh, I’m building my JARVIS, “ right? You’ve probably seen several videos of people saying they made their JARVIS, which is that character from Iron Man, right, that Tony Stark talks to and simply makes voice requests, and

01:36 it executes tasks for him: “Take off in my jet,”“Design this 3D armor for me,” right? That’s what JARVIS does. So, up until now, it was kind of a joke, right ? There were even people selling this idea, right? But most people just took it as a joke. “Look, I built my JARVIS, I made this automation, I made several agents here that work for me operationally.” Really, up until now,

02:09 you could make several agents in an automated way. But not at the level that people sell it. What happens? People keep selling the idea that it’s extremely automated. It’s 100%automated . The interaction is very easy, everything flows very well. It’s not really like that, okay? As you know, if you’ve been following me for a while, you know I’m 100%transparent. And I say that the reality of working with this —I’ve been working with it for over 3 years—I’ll tell you the reality.

02:41 It’s not 100%automated, it’s not 100%fluid the way people sell it. To create your own agents, to put them to work operationally on your computer, for you to do that, a lot of things definitely require your intervention, okay? So you need, for example, to create an automation to perform a task, and to execute that task, it often doesn’t turn out 100%the way you want it. You have to ask for an adjustment,

03:13 right? So you have to go back to that automation, or in the middle of it, you have to stop to make some adaptations. Sometimes you have to give your opinion when you’re generating something or asking for help; for example: setting up a sales funnel, a strategic funnel, or analyzing some information. Sometimes it will bring you something, and you might not agree based on your experience. So you’ll have to go in there and adjust it. For your agent that is providing service, you need to make some tweaks, because sometimes people ask it things you never even

03:43 imagined they would ask. And that is what is happening to this day. But after these changes from OpenAI, and after what I’m seeing not only from OpenAI, but from Claude, from Meta, I saw that this story, this idea of working with a Jarvis is closer than we might imagine. Of course. Hold on, I’m not saying we’ll have a Jarvis that

04:14 will do everything you see in the movies. But our work format over these next few months will become increasingly similar to that style, the interactive voice-based format. Today, as I said, we still have to handle a lot of things on the computer so that AI works for us. You still have to type

04:44 a lot, use the mouse, and make many interventions, but from now on with these tools, for example, you will be able to work in this interactive voice mode and ask for the AI to have more autonomy than it already has. For example, to build a website nowadays, I have to go into Claude Code, open it, create and train my agent to build a site, install the skills, which are abilities for it to make a more professional site that better suits the function I’m looking for. A series of

05:18 tasks that humans still have to do. From the moment you turn on your computer, you will simply ask: “Look, open Cloud Code, create an agent specialized in setting up a sales page. I want this sales page to consult my company’s information from my knowledge base. So, you have to make a website for me that uses this information for the purpose of selling, I don’t know, a new training, a new product, or a new service I have. Do that for me.” The AI

05:54 you choose to use will start doing all of this in parallel. There will be several agents working, each performing a function, and they will execute this for you on their own, without you touching the mouse or keyboard. Yeah, that is the main promise of Dots. And that is what it is, in a nutshell, right? Yeah, but they don’t make that clear. But I’m here translating for you what I see that the AIs and the models are focusing on. Meta also recently

06:26 launched Muse, right? And with Muse, it’s integrated right there on your phone, on your social networks. It can even choose information for you. It can execute tasks, perform analysis, work with any kind of data that’s there, simply by you asking via voice, okay? This is already working; I use it myself. And since I already have my team of AI agents for marketing, for example, I work in that way of always requesting by voice, but there is still a lot of intervention. From now on, man

07:00 , that is going to change. That is going to change. I imagine that, I don’t know, in about 5 months, playing it safe, but imagine this: in a year, in a year, believe me, come back here and watch this video. I strongly believe that we will work practically only by voice, regardless of your field . AI will be much more interactive. You will, you will talk as if you were

07:32 speaking with an employee of yours, or several employees, right, who will be experts in tasks. There you will have the marketing agent, the sales agent, the agent of—oh, let me even share this with you. The, the—the Dots itself, right? It has here, look, for example, this yellow one here is the agent specialized in sales. The pink one for marketing, for example, right?

08:02 The green one I don’t recall here, but each one is an expert in a type of task . Uh, I don’t know, maybe managing your calendar, prepping daily tasks, anyway. Right? Later I’ll dive deeper to bring you more precise content on each type of agent here, and how to use dots , right? As soon as it launches, I’ll test it and show you how you can use it too. But the most important thing—and

08:32 tools aren’t the most important part— is knowing how to take advantage of this, understanding how the market is shifting, and how to put it to work in your business, regardless of what you do, to give you a competitive edge, so you can produce more, faster, with lower costs and less time, right? Uh, and AI is here to provide that. Yeah. That’s what you should focus on. I’ll give you an example of what I mean. Imagine in the middle of work, I go and ask the ChatGPT agent to do the

09:09 following: “Hey, open cloud code for me , build a sales page, and on that sales page, put a form. This form will send every lead that signs up there to a spreadsheet saved in my Google Drive. When a new lead comes in, create an automation that notifies this person at this number that a new lead has arrived . Uh, at the end of the week, I also want you to analyze all the leads who

09:40 filled out that form and give me a report. Send it to my email with the main insights I need, for example, I want to understand what kind of audience is coming in, uh, what are the pain points they are looking to solve, right? Let’s assume that’s in the form. I want that to be in the form. Imagine we’re asking the AI for this. And I want you, after you finish, or rather, in parallel to that, to have another

10:12 agent create a dashboard that centralizes all this information. I’ll have a clear view so I can track the flow of sign-ups on my site daily. I also want you to create a funnel in parallel, a funnel strategy so I can sell this new service I’m going to provide. Set up a series of, of uh,

10:47 copy scripts, right, of, of email marketing emails for me to launch, other texts to send to my WhatsApp group, and some story ideas for me to share this page and script, content ideas that I can also use to bring people to this page. From that, you will generate a schedule for me, an illustrative part to make it easy to understand how I am going to execute this. That’s the level I’m talking

11:17 about. You’ll simply make a request like that and the AI will execute everything in parallel. The AI will do all of this for you, man. without you lifting a finger on the keyboard or mouse. That is what the market is looking at. That’s what big companies , from everything I’m analyzing, the events I’m attending, that I went to, right, to the IO event, they’re talking about, it talked a lot about AI working with data, and I really, after I went to that event, a lot of what was said there is, is hitting home here.

11:51 Actually, I’ll even leave this video as a recommendation where I talked in detail about what happened at that event. And tell me here what you think. Do you think what I’m saying is too far off, or do you think we are even closer than I said? And more importantly, I want to know what you think the market will be like when this happens, when we reach this point of being able to produce at this speed. Remember, I know that everything I said

12:21 here can already be done; in fact, we do it here in our business and for other companies. Also, if you want help , take a look at the link in the bio or scan the QR code here. But I’m talking about what the market will be like, what your work format will be, our work format, and what impact this will have on the market when this point arrives. Us being able, overnight, to just use a simple voice command to get all the agents to execute this much faster, much easier than what we already do today. Send it here, I want

12:55 to know your opinion. Subscribe here if you like this type of more in-depth and 100%transparent artificial intelligence content.