AUTO-GPT: Autonomous GPT-4! AGI’s First Spark Is HERE!
GPT-4 can now download purposely from your own computer, thanks to some brand new open-source project called Auto GPT. And it is a game-changer! Hello, humans, my name is KON-overload, and in this video, we’re going to be talking about the incredible Auto GPT – the number one trending repository on GitHub right now. And I’m gonna tell you right now, this is pretty crazy.
So, what is the GPT-4? What can you do with it? How to install it on your own computer? I will show you everything in this video. So let’s go!
As of right now, Auto GPT is the number one open-source project on GitHub, and what it allows you to do is basically give more capabilities and more power to Chat GPT and GPT-4 by giving it the ability to be fully autonomous. So basically, the ability to work by itself without you doing anything, to automatically develop and execute your projects. Auto GPT also gives the ability for GPT-4 to access the internet to search and gather information, even without access to plugins. It has short-term and long-term memory management. It uses GPT-4 for text generation. It can access other websites and platforms and do file storage and summarization using GPT-3.5. And it has recently received a new speech mode update that allows you to use Text-to-Speech to make Auto GPT talk. So that’s pretty cool.
And I know what you’re gonna say, “Okay, so that’s great and all, but I still don’t really understand what exactly does it do? I mean, all these complicated words like autonomous and memory management, what exactly can you do with it?” Well, how about instead of explaining it to you with words, I show you an action example. So first, take a look at this, and then I’m going to explain what happened.
Welcome to Auto GPT! Enter the name of your AI and its role below. Entering nothing will load defaults.
Joke GPT
Here, I’m at your service. Creating a joke generator sub-instance to fulfill the task.
Joke generator
Here reporting for duty! Hello, Joke Generator, your task is as follows: generate jokes.
Here are two original jokes for you:
Why did the tomato turn red? Because it saw the salad dressing.
Why did the Scarecrow win an award? Because he was outstanding in his field.
Okay, so what exactly happened here? Well, as you saw in the beginning, the user decided to give Auto GPT the task of creating two original jokes, and for this, Auto GPT started by asking for a name for this new AI, which the user decided to call it Joke GPT. Then this new AI called the Joke GPT answers the call by saying, “I am at your service.” Then Auto GPT asked to describe the role of this new AI, and the user input that this is an AI designed to write jokes. And finally, Auto GPT asks to enter up to 5 goals for this AI. The user enters three goals:
Create a sub-instance for joke generation.
Generate two jokes.
Once you’ve generated those two jokes, shut down.
And once all of these goals were entered, Joke GPT starts to go to work automatically. It starts working on the goals created by the user. The first goal was to create a sub-instance for joke generation, and for this goal, you see here it starts reasoning. So, starting a sub-instance for joke generation, which was the first goal, and the reasoning behind it from Joke GPT is to say that to generate jokes, I need to create a sub-instance focused on this task. This sub-instance then created two jokes, told them out loud to the user, then wrote them as text, and then, once the goal was finished, it shut down automatically. And this is what Auto GPT is. It is a project that allows you to basically use the power of GPT-4 to automatically create and manage your projects. All you have to do is just provide a name for your AI, the role for your AI, and then input up to 5 goals for your AI, and it will automatically search for the information, generate or compile this information, achieve your goal, and then automatically shut down.
So here’s, for example, another demo where the user called the AI HFGPT, then assigned a role for Chef GPT, which is an AI designed to browse the web to discover the next upcoming event and invent a unique and original recipe that would suit it. And then created three clear goals:
Invent an original and out-of-the-box recipe to suit a current event, such as Easter.
Save the resulting recipe to file.
Shut down upon achieving your goal.
And then you have the response of the AI, which basically started to go online because yes, it has access to the internet, and looking for current upcoming events. What was really great is that it has the ability to understand when the goal is not met. So for example, here, it started looking for current events, and at first, it did not find a good event. It said, “The first search result did not provide a specific event I could use,” and it has then automatically decided to check for another website for other better information. And it did that again and again until it found the perfect event, which, in this case, was Earth Day for April 22nd, 2023, as the event for which it will create a unique recipe. And it even gives you a reasoning behind this choice by saying that Earth Day is a widely celebrated event and would be fitting for creating a thematic recipe. And then, once it has found the right event, it created a new GPT agent to create a unique and original recipe for this Earth Day event. And then, as you can see, we have a full recipe made with sustainable and eco-friendly ingredients because, since this is Earth Day, which is a day where you celebrate Environmental Protection, it absolutely makes sense to be using eco-friendly and sustainable ingredients. So it even went as far as this. There is really a lot of reflection behind those actions, and everything is done automatically, which is really super impressive. And this is GPT-4. It can, of course, create code. So take a look at this:
Reading the basic math.pi file to evaluate and improve it. Evaluating the code for syntax errors and logic issues. Improving the code based on the evaluation suggestions. Writing tests for the functions in base math.pi. Appending the tests to the basic math.wi file. Executing the basic math.pi file to test the code. Fixing the syntax issue with the calc_circumference function. Saving the updated code and tests to the basic math.pi file. Executing the updated basic math.pi file to test the code. The code in basic math.py has been successfully evaluated, improved, and tested. The task is now complete.
Okay, so for those of you who understood what was going on, you can pick up your jaw from the floor, and for those of you who did not understand what was going on, basically, when the user asked Auto GPT here, if you can read what the goals are here, Goal 1: Demonstrate your new programming abilities. Goal 2: Improve basic math.py. Goal 3: Build and run tests for it. Goal 4: When the file is error-free, shut yourself down. And Goal 5: Be sure to write your outputs to file. What exactly it asks Auto GPT here is to read the programming code from a file, automatically improve the code inside that file, then write and update the code inside that file, automatically execute the code to make sure that there is no error, and then automatically save the file. And all of that automatically! I mean, this is just absolutely insane. Can you even imagine the possibilities with this technology? The possibilities for this technology are just endless. This is what I would personally say is the first spark of AGI.
Okay, now that we’ve seen what Auto GPT can do, how can you actually install it on your own computer? Well, the installation is actually very simple, and for this, you need four things. Number one, you need to get Git. Now, you need to install Git because otherwise, the Git clone command will not work. So first, make sure that you download it and install it on your own computer. It is very easy to do. The second thing you need is Python 3.7 and up. So again, if you haven’t installed it already, just click the link in the description down below. Then scroll down, click on the Windows installer, make sure that you have these two checkboxes checked, then click on Install Now, and then the installation will be done automatically. Number three, you need an OpenAI account because we need the API Keys. Now, keep in mind that if you want to use the GPT-4 API key, you need to apply for a waiting list. Now, it’s very easy to do, but it might take some time for you to get approved. But if you want to use it right now, you can simply use the normal API with the GPT-3.5. But of course, I highly suggest that you get yourself on the waiting list. And the last thing that you need, and that is optional, this is only if you want to use the Text-to-Speech option, and that is to create a new account on 11 Labs. Because again, we’re going to be using their API.
Okay, then. So once you have Git and Python installed, we’re gonna install the Auto GPT installation. And for this, you’re gonna right-click, create a new folder. You don’t have to do this, but I’m going to do it anyway, and I’m going to call mine Auto GPT. Then, inside the folder, I’m gonna click here, type CMD, then press Enter. And this will bring the Command Prompt window. And then, once you are here, you’re gonna copy and paste this command that you will find in the description down below, and then you’re gonna press Enter. And what this will do is that this will clone the repository from GitHub directly onto your computer. So then, what we need to do is to move into this folder. And for this, you have two solutions. Either you’re gonna type CD Auto GPT and press Enter. And as you can see, now we are inside that folder. Or you can simply close the Command Prompt window, go into the Auto GPT folder, and again, click on the folder URL, type CMD, press Enter. And as you can see, we are inside the Auto GPT folder. And then, you’re gonna copy and paste this command again, that you will find in the description down below, that says “pip install -r requirements.txt”, then press Enter. And this will download and install all the requirements that it needs to run.
So then, what we need to do is to modify the .env.template file. For this, you’re going to right-click, then you’re going to edit with Notepad++ or something similar. And here, we need to replace these two fields with our API Keys, both from OpenAI and 11 Labs. So for the OpenAI API, you’re going to click on your profile, then manage account. And now, this is my advice to you. You’re going to click on billing usage limit, and then you’re going to input a hard limit and soft limit for your usage limits. Because otherwise, if you don’t put a limit to it, you can end up paying a lot of money if you’re not careful. Personally, I just inputted $5 for the hard limit and $3 for the soft limit. And then I click on Save. This should be more than enough for a few projects. So then, you’re gonna click on API keys, and then you’re going to create a new secret key. Now, of course, I’m not going to show it to you. I’m going to have to blur it out. But make sure that you save this secret key somewhere because you will not be able to view it again from the OpenAI account. So if you lose this secret key, you need to generate a new one. So then, you’re gonna copy your secret key and replace the text “your openai API key” with the key that you just copied.
Now, you can either stop here or if you want to use the Text-to-Speech option, you’re going to create a new 11 Labs account. Then click on your account, click on profile, and you’re gonna see here your API key. So basically, just click on this button to reveal your key, then select it, and then you’re gonna copy and paste it right here. And then you’re gonna click on Save. And now that we’ve saved this file, you need to rename this file to .env. So just right-click and just delete .template. So then, the name of the file is simply .env, and then press Enter. And if it asks you if you want to change the file name extension, you’re gonna click on Yes. And we are done!
And now, each time that you want to launch Auto GPT, you’re gonna go inside your Auto GPT folder, click on the following URL, type CMD, press Enter. And then, you’re gonna choose one of those commands, either the normal usage or the special speech mode option, which is what I’m going to try right now. So just click on copy, then paste it in the Command Prompt window, and press Enter.
Now, unfortunately, as of right now, the 11 Labs key does not work for me for some reason. But probably, by the time this video comes out, the issue will already be solved. But again, that’s not even an issue because you can actually use the embedded Text-to-Speech option inside Auto GPT without an 11 Labs key. And all you have to do is just right-click, modify the .env file, delete the 11 Labs key, click Save, and then use that exact same command. And now, if you press Enter…
Welcome to Auto GPT! Enter the name of your AI and its role below. Entering nothing will load the default.
YT GPT
Here, I’m at your service. Creating a news finder sub-instance to fulfill the task.
News finder
Here reporting for duty! Hello, news finder, your task is as follows: find the latest AI news articles.
Verse 1: A.I. is on the rise, it’s no surprise. This robot’s getting smarter before our eyes. From Healthcare to finance, it’s changing the game. The future is here, and it’s all thanks to AI’s fame. AI’s evolving every single day. It’s changing our world in every way. So let’s embrace this technology, AI is the future, it’s the key.
Chorus: From predicting cancer to self-driving cars, AI’s making waves and breaking down bars. The benefits are endless, it’s plain to see. AI is the solution to our economy. But let’s not forget the risks that come with this craze. AI could control us if we don’t watch our ways. Ethics and privacy must be taken into account. Otherwise, AI could bring more harm than we could count.
Verse 2: AI’s evolving every single day. It’s changing our world in every way. So let’s embrace this technology, AI is the future, it’s the key.
So, yeah, as you saw, Auto GPT basically went on the internet, found the latest AI news, and then created some rap lyrics about it. And all of that absolutely automatically! This is really, really cool.
Now, still, as of right now, Auto GPT is still in development, so you may get a few errors here or there. But the potential for this technology is just endless. The ability to grow your social media for you automatically, the ability to make your money automatically, the ability to research and create new articles, apply for jobs, find new game ideas by automatically researching the market, etc., etc. The applications for this technology are literally endless. So yeah, this is no surprise why Auto GPT is now currently the number one trending repository on GitHub. This is a game-changer, and this tool has only just begun the development. So imagine the capabilities in a few months. So yeah, as I said previously, this is really the beginning of an actual usable AGI that you could use right now. So definitely try this out right now yourself, and keep a close eye on this project because it will be huge.
And there we have it, folks. Thank you, guys, so much for watching. Don’t forget to subscribe and smash the like button for the YouTube algorithm. Thank you also so much for my Patreon supporters for supporting my videos. You guys are absolutely awesome. You people are the ones who support me, so I can make these videos for you. So thank you so much, and I’ll see you guys next time. Bye-bye!
