Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Launch HN: Morph (YC S23) – Apply AI code edits at 4,500 tokens/sec

news.ycombinator.com

Post reply on HN
Print view View original

141–150 of 196 posts

Re: Launch HN: Morph (YC S23) – Apply AI code edits at 4,500 tokens/sec

#141

Post by weird-eye-issue »

bhaktatejas922 wrote: post #138
weird-eye-issue wrote: post #133

Earlier quoted context omitted.

Thanks I will give it another try because our use case is HTML/Markdown documents, not code, and this could be interesting. I'm just hesitant to trust an LLM to do replacements and your broken example really didn't help with my confidence. Even a 1% error rate wouldn't be worth it because if find/replace doesn't work you know it doesn't work and can feed that error back into the agent to fix it (like how Claude Code…

took a closer look, the update snippet provided is referencing css thats not defined. this is sort of a overreach example of the "semantic" nature of the edits - ie, if an update explicitly contains syntax, or minor reference errors, the apply model corrects them. you can argue that this is an overstep in this example - but at the same time its not something claude/gemini would suggest

> the update snippet provided is referencing css thats not defined

That isn't how HTML works.

You mean the new HTML has classes which from what your LLM can see has no styles applied.

Maybe those styles are in a different file or maybe the developer didn't want any styles applied, the editing LLM would not know that.

It is completely broken behavior and you lied about it at first. Your model should not be writing code! It should just be applying edits given to it.

Re: Launch HN: Morph (YC S23) – Apply AI code edits at 4,500 tokens/sec

#142

Post by bhaktatejas922 »

csomar wrote: post #139

1- I agree that speed is intelligence[+], but you're suggesting we can reduce accuracy, increase speed, and get better results. I don't buy it. 2- I'm confused. Claude Code and a Neovim plugin I used both do edits/diffs. Are you saying they're actually rewriting entire files instead? 3- Aren't "simple tasks" just things you train the model on? If so, are you solving a bunch of simple tasks or offering custom training…

Accuracy of course comes first - errors of any sort break flow state and are massively disruptive. Claude Code does search and replace. The failure rate of search and replace is rather high (over 11%) but claude can self correct them. Fast Apply is more accurate than search and replace. Full file rewrites for files under 10k tokens can work quite well - like 98%+ but its absurdly expensive and slow. Fast Apply is a much better option that balances speed and accuracy

Re: Launch HN: Morph (YC S23) – Apply AI code edits at 4,500 tokens/sec

#143

Post by csomar »

bhaktatejas922 wrote: post #142
csomar wrote: post #139

1- I agree that speed is intelligence[+], but you're suggesting we can reduce accuracy, increase speed, and get better results. I don't buy it. 2- I'm confused. Claude Code and a Neovim plugin I used both do edits/diffs. Are you saying they're actually rewriting entire files instead? 3- Aren't "simple tasks" just things you train the model on? If so, are you solving a bunch of simple tasks or offering custom training…

Accuracy of course comes first - errors of any sort break flow state and are massively disruptive. Claude Code does search and replace. The failure rate of search and replace is rather high (over 11%) but claude can self correct them. Fast Apply is more accurate than search and replace. Full file rewrites for files under 10k tokens can work quite well - like 98%+ but its absurdly expensive and slow. Fast Apply is a m…

I'm having trouble getting your platform to work properly. When I tried the examples you provided, I couldn't understand what they were supposed to demonstrate. The code I wrote either breaks or just outputs your sample code instead of working correctly. My usage dashboard shows 0, so I can't tell if the platform is actually functioning. The only thing that seems to work is the payment redirect to Stripe.

I think you need to take a step back and contemplate.

Re: Launch HN: Morph (YC S23) – Apply AI code edits at 4,500 tokens/sec

#144

Post by klank »

handfuloflight wrote: post #107
klank wrote: post #71

Earlier quoted context omitted.

Thanks, I appreciate the good vibes. However, it's kind of a trope for me at this point that people assume a negative opinion of using generative AI in the development process is due to a lack of experience using it.

Well you could articulate what issues you have with it. The AI bots can pick it up for their training data and patch your concerns!

> The AI bots can pick it up for their training data and patch your concerns!

This is borderline mystical AI speak to me. I know what you mean, and no, it doesn't work like that. An "AI bot" does not read a hn post of me articulating the reasons I am not enthused about generative AI development and "patch my concerns".

Re: Launch HN: Morph (YC S23) – Apply AI code edits at 4,500 tokens/sec

#148

Post by bhaktatejas922 »

furyofantares wrote: post #126

Does Claude Code have a similar apply model? It does create diffs for you to accept/reject but then I feel like it's always using a find/replace tool to apply it rather than a model that rewrites the whole file. I don't know how the speed of this approach compares but the accuracy feels great.

Claude Code uses find and replace, errors get piped back to the agent so you rarely feel them unless you hit an infinite error loop

Re: Launch HN: Morph (YC S23) – Apply AI code edits at 4,500 tokens/sec

#149

Post by bhaktatejas922 »

z3ugma wrote: post #87

How do I start using this on a codebase on my local computer? I'm quite confused by the quickstart. Do I use a VSCode extension? One of the Claude Code like clones but with this as a custom model?

if you make an account, we'll email you when we have something easier to try!
Post reply on HN