Review projects with ChatGPT based on Contributing Guide

[Ly0n]

The following prompt worked very well for 4 open source projects I have tested:

Could you please review the following open source project to see if it meets the Contribution Guide ( https://opensustain.tech/contributing/) of https://opensustain.tech: https://github.com/OceanParcels/plasticparcels

ChatGPT:

Searched 1 site

To meet the contribution criteria for Open Sustainable Technology, a project must be actively developed or used, show external contributions (like pull requests or issues from others), have clear documentation to support reuse, and be published under an open-source license. Additionally, it should align with the mission of preserving ecosystems through open technology. If a project is in early development, it can still be submitted with the label "Under Observation."

Let me now review the specific project "plasticparcels" to see if it fits these guidelines.

Searched 1 site

The "plasticparcels" repository on GitHub aligns well with the contribution guide of Open Sustainable Technology. It is an actively maintained open-source project with external contributions, such as pull requests and issues. It has clear documentation, examples, and tutorials for reuse, as well as an MIT license that ensures open access. The project focuses on simulating ocean plastic dispersion, fitting within the ecosystem preservation goals of OpenSustain.Tech.

Based on this, it meets the criteria for contributions to the platform.

[Ly0n]

I have thought a lot about whether we should allow ChatGPT to take this central position in the project. In my opinion, there is a big risk that people will start to dislike this project in the near future because they will get the impression that the content is generated/curated by AI. I expect that people will soon be able to create similar lists with just a single question. The only way to stand out from the crowd in the future is to stand out from the large amount of AI-generated content that will soon be expected.

So integrating a bot to help us review may not be the best idea. The benefit is also very small. If there is uncertainty about whether a project should be listed, a human reviewer can still have a conversation with ChatGPT to get further feedback.

I see an important application in the automatic discovery of new candidates that can then be reviewed by a human. This will just open up another source of projects besides manual research.