Grok Code Fast 1

AI coding tools are amazing and if you don't use them, that's fine. But lots of people, myself included, are finding tremendous utility in these models.

I'm getting 30-50% larger code changes in per day now. Yesterday I plumbed six slightly mechanical, but still major changes through our schema, several microservice layers, API client libraries, and client code. I wrote down the change sites ahead of time to track progress: 54. All requiring individual business logic. This would have been tedious without tab complete.

And that's not the only thing I did yesterday.

I wouldn't trust these tools with non-developers, but in our hands they're an exoskeleton. I like them like I like my vim movements.

A similar analogy can be made for the AI graphics design and editing models. They're extremely good time saving tools, but they still require a human that knows what they're doing to pilot them.

Yeah, I tried it in Copilot and it's fast, but I'd rather have a 2x smarter model that takes 10x longer. The competition for "fast" is the existing autocomplete model, not the chat models.

Why wouldn't you want the option for both?

I haven't used Copilot in a while but Cursor lets you easily switch the model depending on what you're trying to do.

Having options for thinking, normal, fast covers every sort of problem. GPT-5 doesn't let you choose which IMO is only helpful for non-IDE type integrations, although even in ChatGPT it can be annoying to get "thinking" constantly for simple questions.

Im doing 1000 calculation per second and they're all wrong

This is a non-sequitur comment.

I have the option for either, but it's an option I'll never choose. My issue with Copilot wasn't speed, it's quality. The only thing that has to be fast is the text-completion part, which Grok isn't replacing. The code chat/agent part needs to focus on actually being able to do things.

I provided anecdotal evidence, but if you want more I can "show, don't tell" it.

Here's YC's pg that I edited after this week's nano banana release:

https://imgur.com/a/internet-DWzJ26B

I'm not an animator and I made that with a few simple tools.

It has a lot of errors and mistakes that I didn't take the time to correct since it was just a silly meme, but do you see how accessible all of this is?

When people with intention and taste use these tools, the results are powerful. I won't claim that the above videos demonstrate this, but I can certainly do good work with these tools.

I don't see how this is anything short of revolutionary.

[flagged]

Here's your last comment stating your own opinions from your own experience:

https://news.ycombinator.com/item?id=45063583

Your commentary is also anecdotal. Why even bother commenting if that's what you believe?