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

Grok 4.3

docs.x.ai

Post reply on HN
Print view View original

591–600 of 608 posts

Re: Grok 4.3

#591

Post by sergiotapia »

sroussey wrote: post #581
sergiotapia wrote: post #572

Earlier quoted context omitted.

what do you mean?

It was a joke about people relying on AI and it doing absolutely terrible things. Coding is an interesting area -- it can code, then compile to see if that part worked, then test to see if more worked. With taxes, it sets things up and the review phase is the IRS fining you.

in this case my tax situation is so retarded simple, I could verify what it suggested step by step and I performed the actions on one of those free tax usa website. the irs accepted my returns and everything went fine. if you're in a simple tax situation, try it!

Re: Grok 4.3

#592

Post by pixel_popping »

rglullis wrote: post #221
pixel_popping wrote: post #82

Earlier quoted context omitted.

There is bots everywhere, it has nothing to do with the platform, it has to do with attackers having an incentive to do mass account farming, no platform is secure against it.

Super easy, just make a web-of-trust type of thing: messages are only visible to those who already vouched for you. Otherwise, you pay $0.01/per message/per user reached.

How would that solve it? If I pay, I can still push the content I want (factual or not) which is the same equivalent as paying for accounts directly.

Re: Grok 4.3

#593

Post by gadders »

KaiserPro wrote: post #587
gadders wrote: post #249

Earlier quoted context omitted.

Obviously the intelligence material being gathered was too valuable.

I'm sorry you're gonna have to explain that, what intelligence and why was it valuable? so valuable that only the press paid any money for it?

There is a theory that Epstein was either setup as, or evolved into, a blackmail operation for an intelligence agency. Views differ as for which nation state.

Re: Grok 4.3

#594

Post by gadders »

vrganj wrote: post #590
gadders wrote: post #585

Earlier quoted context omitted.

I see. This is some sort of weird purity spiral, where no party is left wing unless they meet your arbitrary chronically online standards that no-one adheres to in real life. Touch grass dude.

You still have not brought a single argument and continue your attempts at derailment. It's okay to admit you were wrong.

I'm not in the habit of posting AI content, but as a 3rd party with no skin in the game in this conversation:

AI Overview

The UK Green Party is generally considered further to the left (left-wing), while the Labour Party is positioned in the centre-left of the spectrum. The Greens are seen as more progressive and socially liberal, often holding more radical policies, while Labour is described as an alliance of social democrats and democratic socialists.

UK Green Party

Position: Solidly left-wing.

Ideology: Eco-populism, social liberalism, and environmentalism. They are often considered the most left-wing of the main UK parties.

UK Labour Party

Position: Centre-left.

Ideology: Social democracy and democratic socialism.

Context: While traditionally a left-wing party, it has been described as moving closer toward the center in recent years under Keir Starmer. It is often described as having a wider range of views than the Greens, spanning from the centre to the left.

Left-Right Spectrum Breakdown (UK Parties 2024-2026)

Further Left: Green Party

Centre-Left: Labour Party

Centre: Liberal Democrats

Right/Right-Wing: Conservatives

Further Right: Reform UK

Re: Grok 4.3

#595

Post by Krasnol »

1234letshaveatw wrote: post #482
Krasnol wrote: post #399

Earlier quoted context omitted.

Elon Musk has manipulated Groks outputs to target certain demographics. It is important to highlight this fact, as some people perceive the AI as an objective tool rather than a curated one. Furthermore, I found your final paragraph unclear: are you implying that since harassment is a perennial issue, we should disregard any standards that might mitigate it?

Is it your perception that other AIs are unmanipulated? Objective rather than curated?

Did they mention another one with such a significant racism issue or are you trying to whatabout this discussion?

Re: Grok 4.3

#596

Post by MrDrDr »

IshKebab wrote: post #85
MrDrDr wrote: post #22

Earlier quoted context omitted.

Please elaborate.

Some people have this strange idea that only "whatever humans do" counts as intelligence, despite the fact that a) we don't really have a clue what humans do, and b) "intelligence" is definitely not that strictly defined. I think they're just trying to feel like they know some important truth that other people don't.

Agreed. I see this debate as an active discussion as to what intelligence is, not how it's currently (poorly) defined This is a philosophical discussion, and there is no correct answer, but IMO some answers will prove more useful than others. I would like to define intelligence as the ability to solve problems. Lots of other life forms have this ability, and its clear that LLMs also have this. Now, while they may not be poetic (in the literal sense of the word), or conscious, in that 'they' do not experience the world. I think there is a strong case for arguing they conform to a meaningful definition of intelligence. They solve problems.

Re: Grok 4.3

#597

Post by rglullis »

pixel_popping wrote: post #592
rglullis wrote: post #221

Earlier quoted context omitted.

Super easy, just make a web-of-trust type of thing: messages are only visible to those who already vouched for you. Otherwise, you pay $0.01/per message/per user reached.

How would that solve it? If I pay, I can still push the content I want (factual or not) which is the same equivalent as paying for accounts directly.

By buying accounts, you are buying reputation. By paying for the posts, you are maybe paying for reach at first, but (a) it will be costly and (b) it does not guarantee that the reached ones will spread anything further.

Re: Grok 4.3

#598

Post by tistoon »

tornikeo wrote: post #30

So, we have: - claude for corps and gov - codex for devs - grok for what, roleplay, racism? Those are the two things I've ever heard grok associated with around me.

Maybe you live in a (fascist) cave. Hope you will get out one day. Grok is awesome. Peace.

Re: Grok 4.3

#599

Post by pixelmelt »

Scroll_Swe wrote: post #310
2ndorderthought wrote: post #303

Earlier quoted context omitted.

I don't think companies are hosting them because imagine the liability. Could be wrong though. Again I don't know much about these things I just know they exist.

Yes that is my point. It is the dropbox comment all over again. "Well you can just self-host to get uncencored same as Grok without NAZI!! Elon Musk!!" Just like you can spin up an FTP to get your own Dropbox. Well... very few people are going to actually do that.

I've been working on my own misaligned model and grok is definitely different enough with a syspronpt compared to all the other frontier models that I've considered using it to generate synthetic training data, however it leans really heavy into LLMisms which makes it not really worth it. Tangentially I also really like the idea of llms as librarians they are trying out with grokapedia.

Re: Grok 4.3

#600

Post by jadbox »

margalabargala wrote: post #259
ezoe wrote: post #206

While the tread is swapping between "OMG Claude good. OpenAI was done for" and "OMG Codex good. Anthropic was done for". I've never heard about Gemini and Grok. It works mostly similar performance, but people don't mention that much. Still, my impression is, Gemini hallucinate too much while Grok is always less capable than competitors so it's not worth using it.

Gemini is the best model for OCR bar none. It absolutely sucks at coding.

I get great results when operating at just a 'file' level. It's not so great at editing across many files.
Post reply on HN