This is my current AI coding workflow, how are you using AI when coding? 🤖

André J
14 replies
- VSCode with Github Copilot (this is my all-round AI coder) - phind.com and codemate.bot (I usually ask both the same code realted question) (replaces stackoverflow) - Perplexity.ai if the question isn't about code directly (replaces google searches) - Gemini and chatgpt for general questions or exiting textual material - Grammarly AI chrome extension for rephrasing code comments etc. Or copy on a website. - v0.dev (from vercel) for mocking up website structures by prompting I also write about Prompt coding here: https://gist.github.com/eonist/22e7458f9b38424af9d1bfd791b796a4 I'm curious to hear what AI coding tools or tips you might have?

Replies

Cara (Borenstein) Marin
One of my favorite things I've been trying recently is https://www.brainstory.ai/ for rubberducking. It took a moment to get used to because you speak instead of type, but it's pretty cool for getting clarity
Share
Benjamin Garrard
Github copilot for inline, CodiumAI for Test, Code Review and documentation, Cursor for codebase indexing and symbol referencing for Prompt creation. Wrote an article on it https://www.linkedin.com/pulse/n...
Share
Benjamin Garrard
what is interesting about Cursor, I didn't know, but I am literally coding with custom prompts in my codebase defining how I want it to code and iteratively reworking the prompt. Its hands on lol, way better than custom GPTs
Share
André J
@benjamin_garrard I like use VScode with GitHub copilot chat mostly. Sometimes I use Cursor if GitHub copilot chat can handle it. What do you find better with cursor than GH copilot chat?
André J
@benjamin_garrard Really. Nice way of using many AI coders together. Personally I'm really looking forward to when Xcode release their LLM coder which has been announced that they are working on. And also there was operantly a breakthrough with AI coding yesterday. Here is the presser: https://newsweb.oslobors.no/mess... (I don't know if this is as big as they say it is, but sounds intriguing, what do you think?)
Benjamin Garrard
@sentry_co pretty much everything. GH chat has very minimal cases where it can be used. Mainly because it loses all context in about 3 responses and fights the "current page" context the entire time. Its inline is decent, but it is as good as tabnine, cursor ++, codiumate. Github Copilot is cheap for the inline. It's chat is one of the worst I have seen.
Share
Benjamin Garrard
@sentry_co When it comes to software, Apple hasn't really made that great of stuff IMO. They have always been a hardware company, being innovative in that direction. For me, I don't look at Apple as software engineering talent. I agree with Uniscale in the article a bit about domain-driven development, but once they said they integrate with everything it became a lot less convincing, especially when they are developing in C# vs something more data science and ML related like python. It's not convincing that they are a rounded shop. These are just my 2 cents.
Share
Gurkaran Singh
Hey folks, Gurkaran here 🚀. Love your AI coding workflow setup! VSCode with Github Copilot is definitely a game-changer. Adding phind.com and codemate.bot to the mix is genius - double the brainpower, double the code solutions! Perplexity.ai is a neat twist for non-code queries. And let's not forget Gemini and chatgpt for those deep existential coding questions 😄. Throwing in Grammarly AI for code comment finesse is a slick move. And v0.dev for website structure mockups? Brilliant! As for me, I'm all about diving into data science and embedded programming these days. Always tinkering with new AI tools to enhance my coding quest. Thanks for sharing your setup - super insightful! 💻✨ Cheers to AI-powered coding adventures! 🤖 #TechLife #AIInCoding
Joep van den Bogaert
I mostly just VScode with GitHub Copilot, including the chat window for questions. Any reason to prefer phind or codemate? Not familiar with those.
BurukG
By exchanging insights into tools, methodologies, and best practices, we enhance our coding efficiency and effectiveness. Exploring different approaches and leveraging AI tools, like those discussed at https://chat.openai.com/g/g-a0UL..., encourages continuous learning and improvement. This platform facilitates knowledge sharing, enabling developers to harness AI's potential for streamlining tasks, improving code quality, and advancing software development. By participating in such discussions, we contribute to a dynamic ecosystem where AI and coding intersect, driving progress and pushing boundaries in the field.
Share