from Hacker News

Ask HN: Best page to always see the best available AI-based IDEs?

by sigalor on 1/21/25, 5:07 AM with 1 comments

I've been using Cursor for a few months and it's working well for small use cases, but I feel like it's starting to get deprecated already.

Obviously the ecosystem is evolving rapidly, and I'm really looking forward to automating the coding itself more and more, so that for me and my engineers, the flow of ideas to software is just accelerated. I'm sure that many of you are in the same position.

I don't have time to dig through so many AI newsletters, forums, blogs etc. all the time. I want an IDE that understands me and my entire code base and because there are so many competitors in that space right now, I want to iterate rapidly and continuously adopt the best, newest tools quickly. Right now, it's hard to compare these IDEs though and I don't want to waste my time with setting up a new IDE which first analyzes my code base for hours but then fails at the most basic tasks when it comes to generating code.

Right now, a lot of AI-generated code can be annoying to debug, but I'm sure that at the end of 2025, we'll be in a much better position here. Coding is exactly what the agentic revolution will necessarily need to tackle first. Personally, most importantly, I'm looking forward to AI-based IDEs for web-based applications (Next.js, Node.js, TS) which are tightly integrated with existing debugging automation and execute the generated code to check it for correctness before even showing it to me. I'm looking forward to seeing many companies offering AI-based IDEs rise and fail rapidly, as well as seeing them fighting over the best pricing and objective correctness through the emergence of standards for the correctness of AI-based code generators, similar to the "W3C Test Suites" which check the correctness of web browser technology implementations.

So, does anyone have a recommendation for a page/knowledge base which simply lists clear, actionable insights about AI-based IDEs and is continuously updated, targeted at CTOs and other decision makers in professional and semi-professional software development? Some of my desires:

- Of course it should be open source and independent from commercial interests. I'm just an end user for these AI-based IDEs and I want something that just works and gives me all the details to make good decisions quickly, no marketing fluff.

- Preferably it should be managed in a decentralized, participatory way, instead of being gate-kept by a small group of people who could quickly abandon the project or get into conflicts of interest.

- Comparisons of the tools between each other, based on reports from power users, including migration suggestions à la "If you currently use Cursor with your technology stack, consider switching to the IDE "XYZ", because it does this and that better"

- Also reports from power users regarding how well it works with common technologies, e.g. with large repositories including Next.js, Turborepo, Tailwind or whatever, including more complex dependency and monorepo setups.

- Notifications to stay aware of the commercial aspects, à la "The company behind the AI-based IDE XYZ will probably go bankrupt soon, so consider switching to something else soon" or "The company being the IDE XYZ was recently acquired by large corporate ABC, which will, according to official press releases (and a tweet by its CEO or whatever) probably mean this and that for its end users"

- For each tool, it should very transparently list the monthly pricing. I don't care about yearly subscriptions and I don't want any self-hosting hassle, as I will probably use each IDE for (at maximum) only a few months anyway, until something better comes along.

If it doesn't exist yet, let's build it!

  • by gregjor on 1/21/25, 5:39 AM

    All of the LLMs that IDEs use belong to big companies that currently lose money on their service. You can expect price increases once you get locked in. Given the costs of running a useful LLM at scale, good luck finding one you can call decentralized, participatory, open source, and independent from commercial interests. If and when we have to pay true costs -- enough to make the AI companies at least break even -- the value proposition (if any) evaporates.

    The companies flogging these things have to build momentum and market share before investors get discouraged by LLMs/AI, or the underlying technology turns into a commodity you can download with npm.

    No LLM "understands" anything, so you won't find an IDE/LLM combination that can "understand" your "entire" code base. Of course the companies and people with financial interests will claim their tools can do that.

    Given the way LLMs train by scraping sites like StackOverflow, Github, and Reddit, you can expect quality to decline over time as LLM-generated code makes its way into the training. Besides the photocopy-of-a-photocopy effect, what LLMs consider important or dominant will turn into a feedback loop until we all use Python and write code like newbs doing a tutorial. I wouldn't look to LLMs to "iterate rapidly and continuously adopt the best, newest tools quickly."

    You use "best" twice in the title and two more times in the body of the post. Since "best" doesn't describe an objective property of AIs, IDEs, tools, or anything else, you need to explain what "best" means to you. Other people will have different criteria and metrics.

    I wouldn't crowdsource opinions about the "best" programming language or tool to decide what applies to my specific application and business. Why would I trust a collection of anecdotes about something as complex and sophisticated as an IDE+LLM for writing code? I can't even get the good tech review sites to reach a consensus on the "best" power bank for my phone.