71–80 of 118 posts
It's a bit surprising seeing all the positive replies to this thread. In my experience (using GPT-3), the time I spend debugging the code generated by the AI is almost the same as if I was to write it correctly in one go. The comprehension aspect kinda helps me sometimes, like when I need to understand some obscure shell script.
Earlier quoted context omitted.
One thing I am confused about is using ai for boiler plate code . Shouldn’t that be a macro or code generated, or even an aspect ? Having large snippets of ai generated code outside of a macro or code generation tool does not make any sense .
Its helpful to do boilerplate in languages or legacy code you touch less often, e.g. React class components. But also, it can reasonably set types for the inputs of say a React component, just based on an example (made up) usage of some props you give it. And you can even, if you are very uncouth, simply generate the sensible props for a certain kind of object given a description by say the user. It can do things lik…
There are sometimes tools available for this, but the nice thing about ChatGPT is that it's effectively a single tool that can be used for any reasonably well-known API framework in any language.
I just select the code I want to perform some changes on, hit a keybind, ask for what I want, and it does it. I've been so impressed with gpt-3.5-turbo-instruct that I defaulted to that instead of gpt-4.
I use it in Neovim [0], in my terminal [1], in many specialized tools (long live function calling), and through the chat UI when I brainstorm. I'm using Claude as well for some things.
[0] https://github.com/3rd/config/blob/master/home/dotfiles/nvim... [1] https://github.com/3rd/config/blob/master/home/bin/workflow/...
I also use !q for information that I trust can be extracted from top search results by an LLM.
I use the Kagi search engine as my default. If I find myself on a lomg webpage, I can prepend the URL with `!sum` and it will bring me to an LLM generated summary of the page. If I don't find the information I want in the ssummary, I can click "discuss this document" and ask the exact question I am looking to have answered. I also use !q for information that I trust can be extracted from top search results by an LLM.
* Copilot is great * Great for sanity checking your plans or designs (ex: "What are the best ways of securing a multi-tenant application?") * Great for bootstrapping a presentation or pitch (ex: "Please create a presentation outline about Istio, intended for an audience unfamiliar with Service Meshes" * Simple questions/reminders (ex: "How can I make it so a bash script exits on error?") edit: formatting
I use the Kagi search engine as my default. If I find myself on a lomg webpage, I can prepend the URL with `!sum` and it will bring me to an LLM generated summary of the page. If I don't find the information I want in the ssummary, I can click "discuss this document" and ask the exact question I am looking to have answered. I also use !q for information that I trust can be extracted from top search results by an LLM.
You can do this in edge/bing chat for free btw
Either way I use other Kagi features. OP specifically asked about using AI, so that's all I mentioned.
It has an HN mode, so here is its current summary.
- ChatGPT and other AI assistants can help improve productivity by answering questions faster than searching online or asking another person. This includes explaining technical concepts, providing code samples, and helping with minor tasks.
- However, the quality of code generated by AI is sometimes inconsistent, and debugging may take as long as writing it manually. For complex tasks, AI may not be much faster than a human.
- AI is most useful for getting around poor documentation, asking "stupid questions" without bothering others, and learning new concepts through interactive "interviews."
- AI can help with one-off text transformations and formatting tasks that aren't worth writing custom scripts for.
- While AI may struggle with writing production code, it can help with boilerplate, stubs, and minor repetitive coding tasks.
- Different AI systems have varying capabilities. ChatGPT is best for interactive explanations, while Copilot is more like an "autocomplete on drugs."
- It's important to understand an AI's limitations and use good judgment about what types of tasks it will and won't handle well. Day-to-day coding is often too complex.
- AI search engines can provide code samples and quick answers to common "how do I" questions, saving time over traditional search engines.
- Summarization, translation, and documentation generation are other useful applications of AI for productivity.
- By offloading minor, non-core tasks, AI helps users focus on more creative and challenging work.
I’ve been experimenting with pairing a tool I wrote called Promptr [1] with another tool called Open Interpreter [2]. I start with a prompt that teaches Open Interpreter how to use Promptr, and then I discuss what I’m trying to accomplish. It’s certainly not perfect, but there’s definitely something good that happens when you can iterate using dialog with a robot that can modify your file system and execute commands…
That's an interesting idea! Would you mind sharing more about that? Do you have an example?