←back to thread

504 points Terretta | 6 comments | | HN request time: 0.001s | source | bottom
Show context
NitpickLawyer ◴[] No.45066063[source]
Tested this yesterday with Cline. It's fast, works well with agentic flows, and produces decent code. No idea why this thread is so negative (also got flagged while I was typing this?) but it's a decent model. I'd say it's at or above gpt5-mini level, which is awesome in my book (I've been maining gpt5-mini for a few weeks now, does the job on a budget).

Things I noted:

- It's fast. I tested it in EU tz, so ymmv

- It does agentic in an interesting way. Instead of editing a file whole or in many places, it does many small passes.

- Had a feature take ~110k tokens (parsing html w/ bs4). Still finished the task. Didn't notice any problems at high context.

- When things didn't work first try, it created a new file to test, did all the mocking / testing there, and then once it worked edited the main module file. Nice. GPT5-mini would often times edit working files, and then get confused and fail the task.

All in all, not bad. At the price point it's at, I could see it as a daily driver. Even agentic stuff w/ opus + gpt5 high as planners and this thing as an implementer. It's fast enough that it might be worth setting it up in parallel and basically replicate pass@x from research.

IMO it's good to have options at every level. Having many providers fight for the market is good, it keeps them on their toes, and brings prices down. GPT5-mini is at 2$/MTok, this is at 1.5$/MTok. This is basically "free", in the great scheme of things. I ndon't get the negativity.

replies(10): >>45066728 #>>45067116 #>>45067311 #>>45067436 #>>45067602 #>>45067936 #>>45068543 #>>45068653 #>>45068788 #>>45074597 #
jameshart ◴[] No.45067602[source]
If the Grok brand wasn’t terminally tarnished for you by the ‘mechahitler’ incident, I’m not sure what more it would take.

This is an offering being produced by a company whose idea of responsible AI use involves prompting a chatbot that “You spend a lot of time on 4chan, watching InfoWars videos” - https://www.404media.co/grok-exposes-underlying-prompts-for-...

A lot of people rightly don’t want any such thing anywhere near their code.

replies(16): >>45067741 #>>45067793 #>>45067834 #>>45067845 #>>45067876 #>>45067950 #>>45068178 #>>45068224 #>>45068385 #>>45068645 #>>45068805 #>>45068858 #>>45069087 #>>45069800 #>>45070448 #>>45071147 #
ralfd ◴[] No.45067845[source]
> terminally tarnished for you by the ‘mechahitler’ incident

It is forgivable because there is no real understanding in an llm.

And other llm can also be prompted to say ridiculous things, so what? If a llm would accept a name of a Viking or Khan of the steppes it doesn’t mean it wants to rape and pillage.

replies(1): >>45068031 #
1. tenuousemphasis ◴[] No.45068031{3}[source]
It's not about the model, it's about the ethics of the company intentionally building the model, and what they might do in the future.
replies(1): >>45068842 #
2. simianwords ◴[] No.45068842[source]
What was the alternative? This was clearly an oversight and this much was admitted.

Your suggestion that an oversight like this is reason enough to not use the model?

I don’t get the big problem over here. The model said some unsavoury things and the problem was admitted and fixed - why is this making people lose their minds? It has to be performative because I can’t explain it in any other way.

replies(2): >>45069060 #>>45069240 #
3. bhauer ◴[] No.45069060[source]
Yes, it is performative. As is most of the outrage in this thread.
replies(1): >>45071400 #
4. jameshart ◴[] No.45069240[source]
That’s an uncharitable world view. ‘People who reach different conclusions to me based on the same events must be being dishonest’?

From the outside, the Grok mechahitler incident appeared very much to be the embodiment of Musk’s top-down ‘free speech absolutist’ drive to strip ‘political correctness’ shackles from grok; the prompting changes were driven by his setting that direction. The issues became apparent very early that the prompt changes were leading to issues but reversion seemed to be something that X had to be pressured into - they were unwilling to treat it as a problem until the mechahitler thread. This all speaks to his having a particular vision for what he wants xAI agents to be – something which continues to be expressed in things like the ani product and other bot personas.

The Microsoft ‘Tay’ incident was triggered through naivité. The Grok mechahitler incident seems to have been triggered through hubris and a delight in trolling. Those are very different motivations.

replies(1): >>45069330 #
5. simianwords ◴[] No.45069330{3}[source]
> ‘free speech absolutist’ drive to strip ‘political correctness’ shackles from grok;

Say no more. I’m already sold.

6. bigyabai ◴[] No.45071400{3}[source]
You wouldn't know where performance ends and the market begins. Elon bought his audience with performative outrage, he'll be locked in the pillory of public perception until he's a corpse with a dainty "T" logo tattooed on the asscheeks. This is what he wanted - dark comedy, transgressive politics, edgy juvenile quips, now it's all "performative outrage" when people react? When taxpaying Americans and corporate entities respond rationally to racism, antisemitism and sexism?

Elon never outsmarted the federal admin, and he can't convince anyone that he was too retarded to understand the consequences. He's the most embarrassing type of failure, now - a midwit, the man with no plan who went for the king and missed. He be bet it all on black, and struck out hard. He didn't even manage the shoo-in proof for Trump being a pedophile. Now bipartisan politics will resent him forever, and ensure he and his businesses would rather be dead. All because Big Balls told Mr. Silly he could make a killing in politics, what a touching little sob story.

I say this as a Starlink early adopter, general Elon apologist and space buff for life: if you actually think this is an insincere reaction, try copying any of Elon's mannerisms around normal people and watch how they treat you. You'll be a social pariah come Monday.