Auto-GPT

Auto-GPT is an open-source "AI agent" that, given a goal in natural language, will attempt to achieve it by breaking it into sub-tasks and using the Internet and other tools in an automatic loop. It uses OpenAI's GPT-4 or GPT-3.5 APIs, and is among the first examples of an application using GPT-4 to perform autonomous tasks.

Background
On March 30, 2023, Auto-GPT was released by Toran Bruce Richards, the founder and lead developer at video game company Significant Gravitas Ltd. Auto-GPT is an open-source autonomous AI agent based on OpenAI's API for GPT-4, the large language model released on March 14, 2023. Auto-GPT is among the first examples of an application using GPT-4 to perform autonomous tasks.

Richards developed Auto-GPT to create a model that could respond to real-time feedback and to tasks that include long-term outlooks. Users are prompted to describe the Auto-GPT agent's name, role, and objective and specify up to five ways to achieve that objective. From there, Auto-GPT will independently work to achieve its objective without the user having to provide a prompt at every step.

In October 2023, Auto-GPT raised $12M from investors.

Usage
Auto-GPT is publicly available on GitHub. To use it, users must install Auto-GPT in a development environment such as Docker. Also, users must register it with an API key from OpenAI, which requires users to have a paid OpenAI account.

Capabilities
The overarching capability of Auto-GPT is the breaking down of a large task into various sub-tasks without the need for user input. These sub-tasks are then chained together and performed sequentially to yield a larger result as originally laid out by the user input. One of the distinguishing features of Auto-GPT is its ability to connect to the internet. This allows for up-to-date information retrieval to help complete tasks.

In addition, Auto-GPT maintains short-term memory for the current task, which allows it to provide context to subsequent sub-tasks needed to achieve the larger goal. Another feature is its ability to store and organize files so users can better structure their data for future analysis and extension. Auto-GPT is also multimodal, which means that it can take in both text and images as input. With these features, Auto-GPT is claimed to be capable of automating workflows, analyzing data, and coming up with new suggestions.

Software
Auto-GPT can be used to efficiently develop software applications from scratch. Auto-GPT can also debug code and generate test cases. Observers suggest that Auto-GPT's ability to write, debug, test, and edit code may extend to Auto-GPT's own source code, enabling self-improvement.

Business
Auto-GPT can be used to do market research, analyze investments, research products and write product reviews, create a business plan or improve operations, and create content such as a blog or podcast. One user has used Auto-GPT to conduct product research and write a summary on the best headphones. Another user has used Auto-GPT to summarize recent news events and prepare an outline for a podcast.

Other
Auto-GPT was used to create ChefGPT, an AI agent able to independently explore the internet to generate and save unique recipes. Auto-GPT was also used to create ChaosGPT, an AI agent tasked to “destroy humanity, establish global dominance, cause chaos and destruction, control humanity through manipulation, and attain immortality”. ChaosGPT reportedly researched nuclear weapons and tweeted disparagingly about humankind.

Limitations
Auto-GPT is susceptible to frequent mistakes, primarily because it relies on its own feedback, which can compound errors. In contrast, non-autonomous models can be corrected by users overseeing their outputs. Furthermore, Auto-GPT has a tendency to hallucinate or to present false or misleading information as fact when responding.

Auto-GPT can be constrained by the cost associated with running it as its recursive nature requires it to continually call the OpenAI API on which it is built. Every step required in one of Auto-GPT's tasks requires a corresponding call to GPT-4 at a cost of at least about $0.03 for every 1000 tokens used for inputs and $0.06 for every 1000 tokens for output when choosing the cheapest option. For reference, 1000 tokens roughly result in 750 words.

Another limitation is Auto-GPT's tendency to get stuck in infinite loops. Developers believe that this is a result of Auto-GPT's inability to remember, as it is unaware of what it has already done and repeatedly attempts the same subtask without end. Andrej Karpathy, co-founder of OpenAI which creates GPT-4, further explains that it is Auto-GPT's “finite context window” that can limit its performance and cause it to “go off the rails”. Like other autonomous agents, Auto-GPT is prone to distraction and unable to focus on its objective due to its lack of long-term memory, leading to unpredictable and unintended behavior.

Reception
Auto-GPT became the top trending repository on GitHub after its release and has since repeatedly trended on Twitter.

In April 2023, Avram Piltch wrote for Tom's Hardware that Auto-GPT 'might be too autonomous to be useful,' as it did not ask questions to clarify requirements or allow corrective interventions by users. Piltch nonetheless noted that such tools have "a ton of potential" and should improve with better language models and further development.

Malcolm McMillan from Tom's Guide mentioned that Auto-GPT may not be better than ChatGPT for tasks involving conversation, as ChatGPT is well-suited for situations in which advice, rather than task completion, is sought.

Will Knight from Wired wrote that Auto-GPT is not a foolproof task-completion tool. When given a test task of finding a public figure's email address, he noted that it was not able to accurately find the email address.

Clara Shih, Salesforce Service Cloud CEO commented that "Auto-GPT illustrates the power and unknown risks of generative AI," and that due to usage risks, enterprises should include a human in the loop when using such technologies.

Performance is reportedly enhanced when using Auto-GPT with GPT-4 compared to GPT-3.5. For example, one reviewer who tested it on a task of finding the best laptops on the market with pros and cons found that Auto-GPT with GPT-4 created a more comprehensive report than one by GPT 3.5.