AUTOGPT NEW 'AI' Takes the Industry By STORM! (8 New CAPABILITIES)

So AutoGPT is literally the next step in AI evolution. And this is insane. So AutoGPT is essentially ChatGPT on steroids. All you have to do is use the GPT4 API and it carries out tasks without input by the user. So essentially what this means is that you have a prompt, and then the prompts keep feeding back into itself and it creates an infinite loop until your desired task is done. Now, I know many of you already do know what AutoGPT is, but I'm going to be showing you some of the insane applications that AutoGPT is currently being used for and how you can use those yourself. So one of the first examples I saw this was where a user decided to give AutoGPT a task. And the objective was simple. Create a website using React and make a form, Add a header made with AutoGPT, then change the background to blue. And of course you can see that in the bottom right hand corner, AutoGPT is working away now. Now, what's really cool is that you can actually see what kinds of prompts it's giving itself and how it's working itself through the task, just like how a human would. And then of course, on the left you can also see the screen being made. Now zoomed in here you can also see the reasoning, the plan and the criticism. And of course, like we said before, this is the entire feedback loop that AutoGPT is doing. Now, the reason this mix is so effective is because it doesn't need any more human intervention, which, which means that you now have AI agents which can work on your behalf. And this just cuts your workload by around 80% because the AI is going to be able to string together a bunch of things and then execute the task. And I find this very, very interesting because we've already seen examples of where this has been very, very effective. So I wonder if prompt engineering is going to die or we're going to be seeing the rise of AI agents. Here's another example, and this one is with a much better user interface. You can see this is called Agent GPT. Someone is creating Hustle GPT. And of course you can see the goal is to create a startup with only $100 of funding and then of course, make it a success. Now, what's cool about this is you can see every single task being planned out and you can see how the AI actually researches its plan and how it calculates and makes its decisions. Even you can see that it's able to do research and it's able to effectively come up with certain solutions to problems that you didn't even know you had, which I find that is one of the most effective things. Now I'm going to show you how you can actually use this right now. But you can all see right here that this is Agent GPT where you can literally just go ahead and name your agent. So let's say we call it YouTube GPT and let's say increase viewers on my AI YouTube channel. Okay. And then of course you can see right here, I can then deploy the agent. So then you can see it's embarking on a new goal. Increase your viewers on my AI focused YouTube channel, the AI Grid. Now of course you can see right here analyzing viewer data, creating competitive video content. It's absolutely insane. Okay, now you can see that it's thinking and of course you can restart this if it's, you know, taking too long. Now what's really cool about this is you can have many different research tasks. You can, you know, I guess you could say configure this in whatever you want. Okay, Maybe you could find a job online where you know something that you need to do. You could just completely automate this task. So yeah, you can identify popular AI related conferences and create promotional videos or presentations. This is definitely really, really interesting to see how these tasks are done and just how effective this is going to be in the future. Because remember, like I said, we're still pretty much in the early days of AI in the sense that companies are now just starting to get the ball rolling. Now, if you're enjoying this video, don't forget to leave a like and subscribe. But there's also something later on in the video that we do want to discuss, which is Microsoft's tool, which is kind of like AutoGPT, but that'll be discussed a bit later. Now there's also this right here, which is Goal GPT, which is really, really, really cool. You can see that Gold GPT is essentially quite the same as Agent GPT where you simply enter your task and then of course it gives you some real life examples and real life strategies based on your task. And of course you can see in that example, someone was trying to become a millionaire from a very small amount. So let's check out Goal gbt. So all links will be in the description, but when you go over to Go GPT, you can see right here that there are many different things that you can do. So you can put your task in the first bar and then of course you can put your API key in the Second bar. So I'm going to go ahead and ask it to create a simple script and see if it's able to do it. So here I've asked it to create a script on neuromorphic computing and AI for a YouTube video and then of course I'm just going to click deploy. So I'm going to see if it's actually able to create this task for me and see. So instantly we get create an outline of the topics to cover in the video, research relevant content for each topic, including images and short videos, then draft a script for the video incorporating all relevant information. And then you can see right here it is creating some stuff. So. Or it's already created the outline right here. And of course you guys can see it's created the outline of the video, yada yada yada, a brief overview on the challenges and pose. And then of course it says finally describe some efforts to accelerate innovation in the field, including some special programs developing some new technology. So I'll get back to you once this is all done. So I've left this AI bot running for just about around two minutes and you can see right here exactly what it's been able to do. It's actually given me a lot of information that I actually didn't know and it's shown me some real life examples, such as in the University of Manchester, they're using neuromorphic computing to model brain activities. And essentially, if you don't know what that is, essentially neuromorphic computing is just where they're trying to design chips that basically replicate how our brain works and just make them pretty much just 100 times more intelligent and more faster than our currently designed computer chips. So this is honestly insane. You can see right here, it's researching the relevant content and then drafting a script incorporating all the relevant information. Then, then it's giving me real life examples and it's also saying organizing the outline so it's easily understandable for the viewers. Then of course write a clear narrative for the video so it smoothly transitions. Then it says incorporate in examples again. So it's really, really cool. And also it was saying, you know how you should get images from Pixabay. It's definitely really, really cool. Okay. And the crazy thing about this is that what happens when this is essentially installed on your computer, you can actually use this AI to actually execute the task and actually run the programs for you. Which is definitely going to be pretty game changing once you're able to integrate this into your system. So I've got to Be honest with you, this next one here is honestly going to completely change the game in the way that people research papers and the way that people write stuff. Now this tool is not available yet, but I was browsing Twitter and I came across this, which I found to be absolutely incredible. So take a look at this. Okay, this is essentially a bot that you can do to do research. This is a bot that is specifically trained to, to research papers and write papers. And you can see right here it says write an article on the recent starship launch on 23 April and its implications for space travel. Now what's key about this is that this one is actually able to access the Internet and I'm not sure how it's able to do that. You can see that task one is to perform a Google search for the recent starship launch on the 23rd of April and you can see right there that they're able to get that information from a recent article. And then of course eventually what happens is, is that the AI is able to quote this article just like the average script writer would if you were to hire them or if you were to make the script yourself. Honestly, I think this tool is going to be game changing. You can see this person going through all the links and all that kind of stuff. And the task list is really, really good. And you can see at the bottom it does also have a bar of the completed tasks. And if I'm being honest with you, the output that you can see right here looks very, very effective. I mean, if you're someone who is trying to generate scripts quickly, maybe for your blog, maybe for your article, I'm pretty sure news agencies are going to be literally using this because you can see that it's able to cite these sources as well in the exact way that sources would likely be cited in a traditional article. So this tool currently hasn't been released yet. Person who's created the tool has said that it's going to be released next week. If it is released next week, I will cover this or maybe I'll just leave a link in the description when we do the weekly roundup of AI. But I, but this was one of the tools that I genuinely found to be sort of groundbreaking in the sense that it can really, really help out in some mundane tasks. And the fact that you can access the Internet is something that many tools don't do. And even when they do do it like Bing, they just don't do it effectively at all. Could try it for yourself. But anyways, let's Take a look at this one right here. So this one is very, very interesting. So this one is specifically designed for research as well. And this one is designed to get links from the Internet. Now this one was designed to find all hack and use posts that are GitHub repos, to summarize the repos and then get the Repos URL. And you can see right here that it is able to do that very, very quickly and very, very effectively. Now what's interesting about this is that we are seeing a quick and fast transition from standard prompting to a one prompt solution, where the prompt solution essentially just gives you entirely what you need based on an Auto GP sort of architecture. So it's crazy to see exactly what people are building on top of this. And to be able to access the Internet, it's definitely going to allow a lot of these apps to do just unspeakable things. I mean, you know, being able to get all of these links, being able to write research papers that we saw before, these softwares are definitely going to be able to compete with Bing. And I'm wondering if the tech giants are going to be integrating this stuff into their products. Now what's also interesting about this is that this is an Auto GPT we are looking at right here. All you are looking at is Microsoft's Jarvis. And essentially you can see right here it's got task planning, model selection, task execution, execution and response generation. So some people could argue that this is essentially like Auto GPT, but I'm wondering if Microsoft is going to refine this process and then actually release something where they actually call it Jarvis on Microsoft computers that's literally going to be able to do any task. And I do think that if they do do that in the future, that is going to just put them miles and echelons ahead from Google. So it will be interesting to see how quickly this software does develop because boy oh boy, things are getting really, really interesting. And I think once again we're starting to see the rise of AI tools that are really going to wipe out certain industries and certain parts of the workforce because it just seems to do things much more effectively and much quicker than the average person simply can.