dimgl

What harness do you use for Kimi?

I just like Grok 4.5 and Grok Build

I agree about OP.

However after looking at all of these articles, these all seem like instances of users misusing the product. The product happens to reply on social media, so media publications immediately capitalized on this.

Seems less like malicious intent from xAI's part and more like a product with young and/or insufficient moderation controls.

Just today I saw an article where xAI is suing a creator for creating illegal content. https://www.reuters.com/legal/litigation/musks-xai-sues-grok...

Grok Build seems faster to me than `omp` and Claude Code but I can't put my finger as to why. Anecdotally, after disabling code uploads the agent doesn't respond instantly anymore (it used to respond within milliseconds).

Why snowflakes? You can use /feedback in the app.

They made it open source. Are you just trying to be bad faith here? Isn't this what the community was asking for?

The "scam" part is the hyperbolic part

A bit hyperbolic, no? A majority of code is now written through Claude and similar services.

Grok Build has had impressive performance in a couple of my projects. And fast. So this revelation has been very disappointing...

I will say, a majority of the code I'm writing now is fully through an online LLM. If a company wanted to reconstruct a project I'm working on, they could just replay all of the tool calls from their logs, if they decide to retain the data (I did this locally once to recover a project that I mistakenly clobbered in Git).

Still, this is a big overstep IMO. At the very least, they should make it clear in their terms of service and privacy policy, and not hidden through legalese. Not all usage of Grok Build will be through their enterprise plan which offers ZDR.

Want to add that I started using Grok 4.5 on Grok Build and the big codebase problem seemingly vanished. The issue was only happening on OpenCode.

> They're the only AI company that seems okay with CSAM (or at least don't do as much to stop it)

Wait, what? In the X and Grok terms of service it's pretty clear this is prohibited. Where are you getting this?

Refreshing to see this be the top comment, thank you. I agree.

To answer your question, Grok 4.5 seems to be pretty good at simple tasks and gets even some of the trickier ones correct but it tends to struggle with bigger codebases that aren't very uniform. I've noticed that it uses a fraction of the tokens to get to solutions which is really impressive compared to GLM 5.2 which tends to be an overthinker.

I'm not sure if Grok 4.5 will become part of my stack yet but I am genuinely impressed with what it's been able to achieve.

I'm also unsure if Grok 4.5 is the same base as Grok 4.3. Maybe it is and the data they've used in pretraining is additive (the Cursor data) but it feels like a completely different model than the previous Grok versions.

As opposed to what was happening before, on Twitter?

I've come to this conclusion too. It's a bit disappointing. I fully realize this is more than likely just user error. And now with AI agents the maintenance of a Postgres instance is likely less of a burden. But I can't help but yearn for the simplicity of a single SQLite file and bespoke solutions for things like queues, pub/sub, caching, etc.

Yep, I'm having the same verdict. Interestingly, other people swear by it. I'm trying to understand what's going on with that.

Which are you talking about? OpenCode or ZCode?

Yeah you're overthinking it. Their product releases and their general approach to business is harming their business.

why

Maybe I was getting downgraded to Opus 4.8 but I saw nothing even close to resembling this behavior when using Fable.

I'm gonna be real here. I think they should have cancelled this release.

> long story short his refactoring caused what was otherwise a stable system into a complete mess

Yeah that totally happened

It's not envisioned. There have been mass layoffs and a lot has been pinned on AI, even if it's true or not.

I think sleep apnea would have killed me had I not gotten the sleep study. I was having 100+ events in the span of a few hours. Anyhow, I use the CPAP everyday. I don't snore anymore and I have maybe 1 event a night. Incredible machine.

This is an overly negative response to a genuine solution. There are a million reasons you shouldn't do X or Y.

More than likely GitHub would have to maintain their own internal wallet solution for this, which is a big engineering lift. But we're all just having a discussion.

> you're unfairly conflating things and putting the blame for a lack of care or understanding on tailwind vs on the dev themselves. nothing about tailwind forces you to build inaccessible or "div soup" apps

+1

Hey, that's cool. Does this support conversation lookups? Like, "find this conversation we talked about yesterday"? I built a similar tool to this, although Regent seems much more elegant: https://github.com/divmgl/clancey/

Why do you assume it's botted? Just open up Codex on GPT 5.4 and point it at your codebase.

slop

Vagueposting in Hacker News?

In short: no.

Nothing comes close, in my opinion. Sonnet and Opus are still the best models. The Codex variants of the GPT models are also great. I've tried MiniMax, GLM, Qwen and Kimi and for anything even remotely complex these models seriously struggle.