I’m Frank, the Head of AI Engineering at Superblocks. We’re a platform for enterprise developers to build and host mission-critical internal apps.<p>Today we launched Superblocks AI and are excited to hear feedback from the HN community!<p>Superblocks AI combines the power of the Superblocks drag-and-drop App Builder with robust AI code generation, code optimization, code explanation, mock data generation, and API call generation across SQL, Python, JavaScript, JSON and HTML.<p>We built Superblocks AI with the intention to streamline some of the repetitive and often confusing aspects of building internal tools, here’s a brief overview:<p>Generate Code: LLMs like ChatGPT have quickly become a critical part of developers' lives and we wanted to bring that experience natively into our product. The response is streamed for the best UX, similar to Chat-GPT, and can be modified before use.<p>Explain Code: Understanding someone else’s code for large internal engineering teams is often a challenge. Highlight code and Superblocks AI offers an explanation that you then have the ability to add as a comment.<p>Edit Code: You can make alterations to code by highlighting it and choosing prompts like optimize performance, enhance readability, or something custom. We've added an inline code diff view for easy verification of changes.<p>Generate API Calls: You often want to make an API call but need to spend 30 minutes looking up the documentation and crafting the correct call in Postman. We provided a way for you to give a prompt, and Superblocks AI will generate REST and GraphQL queries for APIs like Stripe, Salesforce, your favorite SaaS provider, and more.<p>Generate Mock Data: This enables faster UI development by generating mock data for components like tables, charts, grids etc., which can be especially handy during prototyping.<p>I'd love to hear feedback from the HN community on how you think this feature should evolve over the next few months!
I honestly really like this, both conceptually and from an implementations point of view.<p>Assisting in the creation of single purpose functions is, in my experience, by far the most reliable use case for LLMs in regard to code creation at this point in time, yet is something I haven't seen many companies really lean into. In fact, the only other example I know of is Cloudflares Workers Workshop [0], which I have been waiting for since May.<p>What I especially like about the implementation showcased in your demo videos though was the visual component. Some subtle animations, effective use of diff, methods being editable before pasting, it all felt very cohesive and thought through, which is something I can't say about every company utilizing OpenAIs APIs. Seems more like something that can serve a real value add for users, rather than just jumping on LLMs because it's the hot thing this summer.<p>Really hope more people are experimenting with the "visual implementation" of LLMs, using animations and graphs to communicate LLM outputs has a lot of unused potential currently.<p>[0] <a href="https://blog.cloudflare.com/introducing-cursor-the-ai-assistant-for-docs/" rel="nofollow noreferrer">https://blog.cloudflare.com/introducing-cursor-the-ai-assist...</a>
This is neat, but, as with all projects in this vein, I fail to see the real value add. We've had ML-powered code generation for several years now (Copilot released publicly in late 2021), and apart from a cute novelty, there's no real use case I've found. (<i>Maybe</i> except for generating tests, but when dealing with a novel code base, the models really can't figure out dependency injection, mocks, weird edge cases I'd care about, etc.)<p>This might be a bit contrarian, but I don't really think there's product market fit here, at least for engineers. And if given in the hands of non-engineers, you'll need to have an engineer at least look over the code anyway. I haven't seen or heard of anyone that seriously uses ChatGPT to generate code and uses it in prod (apart from engagement-farming Twitter posts).<p>> Understanding someone else’s code for large internal engineering teams is often a challenge<p>Without proper context, this will fail miserably. Have you ever seen a corporate codebase before? It can be the definition of "wtf?" even for a seasoned developer that's been working there for years.<p>This AI-generated code trend reminds me of the "no-code" fad of the past decade.
This looks cool. I like that you're showing diffs and integrating the AI code and suggestions back into the original source code.<p>I've been focused on similar concepts with my open source AI coding tool. My tool is a command line GPT chat tool. You can ask it to write or edit code in any git repo. It displays live diffs as the AI edits stream in and automatically integrates them into your source files. I think this is similar to what your article is describing?<p>Folks might want to check out `aider` if they want to do the style of AI coding you're showing, but on their own git repos.<p><a href="https://github.com/paul-gauthier/aider">https://github.com/paul-gauthier/aider</a>
At first was excited when I clicked the link. I think for some people, this is a fantastic solution.<p>However, for me, what I was hoping it was/ one thing I'm still looking for:
Feed in my whole repo to the GPT-4 API, train on it, and ask questions about the code base/ particular functions. Has anyone seen that?
Superblocks is cool, but... when I tried building a simple web hook receiver, it totally failed because the sender of the web hook had their own particular notion of how authentication should work. And Superblocks was rigid in its concept of "web hook". There is no way to write arbitrary request handling code to get around this limitation.<p>As with all low-code / no-code solutions, you can't please all the people all the time. But I felt the design choice in this particular case was overly rigid and with little justification.
Airplane too, does have a coding assistant similar to this [1].<p>[1] <a href="https://www.airplane.dev/autopilot" rel="nofollow noreferrer">https://www.airplane.dev/autopilot</a>
This looks neat but the examples are terrible.<p>The fact that it's even willing to write a function which removes PII from a string is nonsense. No such function is possible.<p>And the comment for filterUsersTable is incredibly long, tedious, and unhelpful.<p>How does the SQL thing work? How does it know what my schema is? That's nowhere on screen.<p>The only believable example is the Salesforce one which anyone could do.
This is cool, but otherwise you can just use windmill as an open-source internal app builder and use Github Copilot in vscode to get the best AI assistant out there: <a href="https://www.windmill.dev/docs/cli_local_dev/vscode-extension">https://www.windmill.dev/docs/cli_local_dev/vscode-extension</a>
If you're interested, there's a number of open source alternatives to Superblocks<p>Budibase.
<a href="https://github.com/Budibase/budibase">https://github.com/Budibase/budibase</a><p>Appsmith.
<a href="https://github.com/appsmithorg">https://github.com/appsmithorg</a><p>Tooljet
<a href="https://github.com/ToolJet/ToolJet">https://github.com/ToolJet/ToolJet</a><p>And more (I'm sure others will comment)