Last Updated on
If ChatGPT isn’t enough already, you’ll certainly want to know about this new autonomous AI agent. So, what is AutoGPT?
AutoGPT is the next big AI application derived from OpenAI’s API. This new AI system is built upon the same model that powers the ever-so-famous ChatGPT.
However, AutoGPT can do so much more. It promises to “autonomously achieve whatever goal you set”, running GPT-4 completely on its own.
There’s no doubt about it, the AI bot’s capabilities are impressive. By leveraging the power of OpenAI’s model, this AI application can automate multi-step tasks, chaining together “thoughts” in order to create self-prompts. This means you can tell AutoGPT your end goal and the AI will generate its own prompts that are required to complete the project.
Developed by Toran Bruce Richards, this AI application is open-source and can be accessed directly from the dedicated GitHub page.
What can AutoGPT do?
According to the developers, AutoGPT has access to the internet for “searches and information gathering”. The AI can store information in its long and short-term memory allowing it to remember past prompts to self-generate the next. It utilizes OpenAI’s GPT-4 for text generation, as well as GPT 3.5 for file storage and summarization.
As you would expect, anything you can ask ChatGPT you can also ask AutoGPT. Although, this time you can request more advanced tasks. For example, you could use Auto-GPT to create an app, build a new startup, and even order pizza.
Unfortunately, accessing the bot is not that simple. You can’t just jump on your PC and run the AI application off a browser – as you would for ChatGPT. To access it you will need to install specific software and be familiar with and comfortable running Python code.
If you are interested, instructions on how to run this model can be found in the README file, on the GitHub page.
Do you have to pay for AutoGPT?
AutoGPT is free for the public to use and is an open-source application. However in order to use this large language model you will need to have an OpenAI account and pay to access OpenAI’s API. What actually costs money is the OpenAI tokens required to run the model.
Final Thoughts
Artificial intelligence models have been getting a lot of buzz lately. With more and more generative AI applications hitting the market, it is no surprise that many developers are pushing the limits of what these models can do.
AutoGPT is an exciting new development in this space, which promises to make GPT-4 completely autonomous. This AI application is still pretty fresh on the scene. But, as time passes it will definitely be interesting to see how the AI community takes advantage of this new bot.
- NOW READ Chat GPT alternatives
