Since the emergence of ChatGPT, many users and companies have searched for the best way to get the most out of what the OpenAI bot has to offer. Thus, although the most popular use is related to writing essays or reviewing computer codes, the truth is that the base of this tool ( GPT-3.5 and GPT-4 in the paid version ) is so powerful that it can be squeeze into other types of applications. And, for this reason, in recent days AutoGPT has stood out, a variant based on the automation of artificial intelligence.
Thus, although some have chosen to create an AI whose objective is the destruction of humanity, most users who take advantage of AutoGPT do so with one objective in mind: that artificial intelligence carry out routine tasks automatically. However, although this possibility is very striking, a series of basic computer knowledge is required to be able to install this option. And, for this reason, below we will explain how you can do it easily and quickly.
How to install AutoGPT to take advantage of the virtues of GPT-4
As Sully Omarr explains through a Twitter thread, any interested user can take advantage of AutoGPT if they know how to do it. Thus, when delving into this tool, they will discover an option that can answer emails autonomously, create web pages instantly and all kinds of actions that occur to those people who want to take advantage of it. But, as a minimum, it is necessary to take into account the following elements to start making AutoGPT:
- Get the AutoGPT code on Github via the following link.
- Download Python and Docker Desktop to work with what you downloaded on Github.
- Obtain a GPT-4 OpenAI API key. To do this, you will have to join the OpenAI waiting list, explain what you want the key for and wait for it to arrive.
When you start the configuration, you will see a file called “ .env.template ” in which you will have to put the OpenAI API key. It then duplicates the file and renames the second one to “ .env ”. After this, execute the command “ pip install -r requirements.txt ”, an action that will download all the Python packages (so it is very important that you have Python downloaded to continue with the procedure). At this point, you will be very close to having finished installing AutoGPT.
Make sure you have Docker Desktop active before you can continue. Once you’ve opened it, run Python scripts/main.py and the terminal will start asking questions: so if you want it to be fully self-contained, you can opt for continuous mode; On the contrary, if you want to configure it yourself, follow the relevant steps until the process is finished. If you don’t want to get complicated, for continuous mode you will have to enter “ python scripts/many.py – -continuous ” for the action to be executed, but it is recommended that you manage it yourself.
What you can do with AutoGPT
In the same Twitter thread, Sully Omarr recommends starting with simple tasks to take advantage of the possibilities and virtues of AutoGPT. Being a tool with only a few weeks of life, it is far from perfect. However, it is capable of offering users several options so that they can make the most of the capabilities of generative artificial intelligence. And, in fact, to automate actions such as checking social networks, answering emails or other similar actions, few current options are up to what AutoGPT achieves.
As an additional note, it is worth noting that it is possible to configure AutoGPT to perform with GPT-3.5 , the language model that supports the free version of ChatGPT. However, being a more limited version than GPT-4, the result will be worse than with it. Finally, it is expected that with the arrival of GPT-4 32K, the next step in the evolutionary scale of GPT-4, AutoGPT will continue to evolve and introduce tools that will fascinate those users who use the virtues of the AI-based tool.