Show HN: MutableAI (YC W22) – Copilot Alternative for VS Code
marketplace.visualstudio.comHi HN! We’re MutableAI (YC W22) (https://mutable.ai). Our mission is to reduce the time and cost to create high quality code using AI. We have worked at the full gamut of companies ranging from startups to big tech and have gotten the sense that many of the rote aspects of software development could be made easier, freeing up precious mental energy.
Devs love Copilot, but autocomplete is just one of many ways that AI should make programming easier. We’re taking a more comprehensive approach to developer tooling that bundles Copilot-like autocomplete with documentation, custom AI instruction, and some early refactoring capabilities (Python only) in one extension. We currently support Javascript/Typescript, Python, Go, and Rust, with more coming soon. Overview (w/ bg music) [1].
In addition to autocomplete, which can be triggered either automatically or manually, we can add documentation to all your methods in a Rust file, or you can ask the AI to insert missing imports in a Python file. In some cases we can get really sophisticated and ask it to program a game from scratch or update your REST interface to accommodate a new data payload.
We use a combination of AI (e.g. OpenAI codex) and AST transformation / metaprogramming techniques on the backend. We are also working on providing other backend solutions for varying needs, including on-prem deployments.
We soft launched our product with a small cohort of users and want to welcome more of you to our pilot beta. We hope you enjoy the product and look forward to learning from you.
We are currently in an extended free trial phase for early adopters and plan to keep a free tier for solo devs and open source contributors. We also offer a generous discount to startups. For enterprise please reach out to info@mutable.ai
We want to thank our very earliest users and invite the HN community to try the product installing it via the VS Code marketplace [2]. We're looking forward to hearing your comments and feedback, or feature suggestions!
[1] https://www.youtube.com/watch?v=x-gKEbgyzCg
[2] https://marketplace.visualstudio.com/items?itemName=mutable-...
Congrats on the launch and really want this to succeed. I think I write too many trivial code that can be replaced by AI like this - like writing simple auth stuff or json parsing logic etc in my projects. However, I just installed it in VS Code and tried it in my Django project.
Feedback:
1. I like the refractoring suggestions. Really cool and had some awesome suggestions right off the bat.
2. I tried to give instructions to write a method and it simply failed. I tried something like: "write a method to parse json object and parse non-zero values from it". All it did was add a method called parseJson(json) with pass as body :( I tried a few different variations to get a method written by mutable but got nothing except new lines.
Anyway, wish you the success. I would love to see more players in this space. I feel a strong need for this kind of AI that can supplement my day to day coding .
Thank you for trying us out and we appreciate your congratulations. We are moving very fast so you should notice steady improvements all the time. We are serious about training our own LLMs and believe a competitive one can be trained on a seed startups budget.
Is there a way to provide feedback for each suggestion in the editor?
Feedback that goes directly to us? We've thought about having a feature where you give us feedback and a ticket is filed with the code snippet in question (and the AI suggestion). Is that what you mean?
Yes, something like that. Even a thumbs up, or down or a way to let you know that the suggestions didn't work for me.
Makes sense. We can also learn from this!
I noticed the docs mentioned it sends upto 1000 lines on a request and reads the contents of the immediate files.
One thing I have noticed with Copilot when working on large codebases is that the suggestions are that useful. For generic code it is fine, but when trying to write code that depends on ~100% internal modules, it was pretty useless and generated a lot of noise that I rarely used it and finally uninstalled.
I have been wondering if taking a hybrid approach of creating a mini model of the entire project locally and using it with Codex as a complementary source would greatly improve the quality of suggestions. I don't know enough about all the existing tools, so this might be something already implemented. Just wanted to share this with someone who worked in this space
Congratulations on the launch.
That is a great point. We want to scale to entire codebases so that this will no longer be an issue. I think you're on to something when you say that companies, teams and ultimately individuals will have their own models. We're really excited about this future and can't wait to build it.
This does prompt hacking and calls OpenAI's Codex, which seems more like a clone than an alternative to Copilot.
Thanks for your comment. Our vision of this space goes beyond autocompletion, for example currently we have documentation and the code edit functionality. We are also building out our own LLMs to support other functionality like AI refactoring.
Congrats on the launch! What is your thinking with the on-prem deployments? Would that work along the lines of something like Hugging Face's new Private Hub?
Thanks for your question! I am not super familiar with Private Hub, but I envision companies wanting their own networks (which don't necessarily have to be on-prem) for many reasons. This allows for fine tuning the network on company/team code idioms, lower latency, and even fine tuning for very company specific applications (say translating their APIs from one language they develop in to all other languages they support).
Wouldn't it be much easier to scale this for functional languages where there is a theory of program transformations [1] and code is more composable?
[1] https://www.amazon.com/dp/013507245X
Devs don't love copilot. I don't know any good dev who uses copilot more than a novelty.
Perhaps you’re just generalizing your social bubble? I use Copilot and I find it’s pretty useful. I don’t use it to write whole functions though, but the autocomplete gets it right 7 out of 10.
Congrats, very excited to try. Any chance of supporting JetBrains IDEs?
No immediate plans to do so unfortunately (but it's an inevitability in the long term).
I'm a bit out of the loop on these, but,
1) Is there any Copilot alternative that is open source, and,
2) Is there any Copilot alternative that works with Vim?
> Is there any Copilot alternative that is open source
In fact one was posted yesterday!
https://news.ycombinator.com/item?id=32327711
TabNine isn't open source but works with Vim: https://www.tabnine.com/install/vim
Tabnine is my personal favorite as it works with most IDEs including IntelliJ, Vim and even VSCode and the suggestions are high-quality. I actually moved away from Copilot to Tabnine across all IDEs and haven't gone back since.
I actually used Copilot in neovim and it works well (because the VS Code Copilot does not work in the open source VS Code build)
Not sure about 1, but I know for 2 you can use the vim VS Code extension as a good emulator.
Add Nim support please!
I am personally quite interested to see how many Codex-wrapper-for-VSCode projects we see. It's the one project every single engineer on HN has considered. (Same goes for the GPT3 text generation flavor, the "autocomplete for Gmail/Chrome/whatever" ones).
What makes it interesting is everyone knows there will be thousands. So the people who decide to continue anyway will presumably have some significant ideas for post-processing.
It's interesting in a game theory sense.
Yes... many of these will be the obvious (generate 5 suggestions and rank them with a different neural net, similar to what Google is doing internally, etc.)
But a small portion will do interesting things. (I know my personal daydreaming session on the topic ended with many pages of possible approaches.)
There are MANY novel applications or approaches possible beyond simply "wrap Codex, add stack-specific context to the prompt, re-rank with secondary model".
I think some of the wrappers will actually succeed. But there will be so many (I predict it being a popular course project for CS classes etc).
This is so cool! Congrats on the launch!
Super useful product and really tight execution. Excited to see how this evolves.