as I went through texts and messages via discord, I got attentive to AI tools in the programming section. Here are some examples of tools:
- DeepCode, looking for ways to improve code and for example get better efficiency
- YouCode, as a Search Engine for complicated tasks
- Copilot, as an automated code finisher, usable in many different programming languages
Basically all of these will create you a finished code in whatever programming language you like or select/help you finish/improve your code.
The last one caught my interest immediately because it has an own Github site and demands money for its use.
Interestingly they show that they have certain weaknesses e.g. when using code or wanting them to write fully new and custom code, suitable your taste like Morris did with his worm, meaning it lacks creativity. Another point is that it can not provide suggestions for new codebases:
“GitHub Copilot is trained on public code. When a new library, framework, or API is released, there is less public code available for the model to learn from. That reduces GitHub Copilot’s ability to provide suggestions for the new codebase. As more examples enter the public space, we integrate them into the training set and suggestion relevance improves. In the future, we will provide ways to highlight newer APIs and samples to raise their relevance in GitHub Copilot’s suggestions.”
- What do you guys think is their limit?
- What is your opinion on these tools, especially Copilot?