Live data from Hacker News

Grok3 Launch [video]

x.com

921–930 of 1001 posts

Re: Grok3 Launch [video]

#921
post #250

Earlier quoted context omitted.

I use it everyday. For coding, I use cursor composer to gather context about the existing codebase (context.md). Then I paste that into DeepSeek R1 to iterate on requirements and draft a high level design document, maybe some implementation details (design.md). Paste that back into composer, and iterate; then write tests. When I'm almost done, I ask composer to generate me a document on the changes it made and I doub…

Correct me if I am wrong, but the whole premise is of Cursor and Windsurf is that this architect-editor mode already being built in into the editor. This is why there is a distinction between composer (Editor) and chat function (arhitecture).

Haven't tried Windsurf yet.

Chat function is just chat; it can't edit your files.

Composer probably relies on prompt engineering to do editor-architecture, as it reads and writes to your codebase. But it's heavily heavily to Sonnet 3.5 and tool-calling.

For architecture-type stuff, I prefer DeepSeek R1 as reasoning models do better on high level design. Which is why I will copy and paste in and out of compsoer.

Re: Grok3 Launch [video]

#922
post #427

Earlier quoted context omitted.

>Not available in Europe currently I hate how its the same story for every new AI technology. If someone can tell me who to vote for or where to protest to change this awful EU law, that would be great.

Have you tried asking the AI people to ship an AI that complies with EU law?

EU regulation are hit and miss. USB C and Opening up apple store are hits. AI regulation and cookie banners and idiotic bottle caps are miss.

Re: Grok3 Launch [video]

#923

Earlier quoted context omitted.

You'd still need a fairly large amount of compute power to be able to run DeepSeek R1 locally, no?

Well yes, but not so large that it's completely prohibitive. People have been running the full models on computers going as low as $6000: https://x.com/carrigmat/status/1884244369907278106 Of course this is for a personal instance, you'd need a much more expensive setup to handle concurrent users. And that's to run it, not train it.

I have a used workstation I got for $2k (with 768GB of RAM) - using the Q4 model, I can get about 1.5 tokens/sec and use very large contexts. It's pretty awesome to be able to run it at home.

Re: Grok3 Launch [video]

#924

Earlier quoted context omitted.

>Not available in Europe currently I hate how its the same story for every new AI technology. If someone can tell me who to vote for or where to protest to change this awful EU law, that would be great.

Companies need to adhere to GDPR in order to enter the European market, poeple have the right to request to delete their PII. It's a good law, actually this should be applied everywhere. As an European, though, I'm scared: what if companies are actually testing if excluding us is so much of a problem for their business?

They are doing it. At some point we have to agree that Brussels are idiots and rarely savants.

Re: Grok3 Launch [video]

#925
post #367
post #250

Earlier quoted context omitted.

I use it everyday. For coding, I use cursor composer to gather context about the existing codebase (context.md). Then I paste that into DeepSeek R1 to iterate on requirements and draft a high level design document, maybe some implementation details (design.md). Paste that back into composer, and iterate; then write tests. When I'm almost done, I ask composer to generate me a document on the changes it made and I doub…

This sounds great - would love to hear a little more about the prompts. Are you literally just asking 'write me a context.md that explains how feature x works' or something like that?

For context.md, it's that simple because it's unstructured data extraction from your codebase and working with a regular LLM model.

For design.md, I have a prompt because we're now working with a reasoning model and doing structured data extraction: create me an issue on Linear with a title, description, and a to-do list.

I would recommend trying the approach yourself and saving the prompts if you can nail down the repetitive asks.

Re: Grok3 Launch [video]

#927

Earlier quoted context omitted.

>Not available in Europe currently I hate how its the same story for every new AI technology. If someone can tell me who to vote for or where to protest to change this awful EU law, that would be great.

Companies need to adhere to GDPR in order to enter the European market, poeple have the right to request to delete their PII. It's a good law, actually this should be applied everywhere. As an European, though, I'm scared: what if companies are actually testing if excluding us is so much of a problem for their business?

I would be very surprised if they aren't monitoring the cost-benefit curve of delaying EU launches. Why wouldn't they? It costs extra money, time, and legal risk to launch in the EU. It's especially bad for XAI due to Musk's involvement.

(Note that it's not just GDPR, there's also the EU AI Act which has a whole extra set of requirements for compliance.)

Re: Grok3 Launch [video]

#929
post #663

Earlier quoted context omitted.

Naive question from a bystander , but since DeepSeek is open source and is on par with o1-pro (is it?), shouldn't we expect that anybody with the computer power is capable to compete with o1-pro?

No DeepSeek model is open source; they're freely redistributable, but without source.

I guess when it comes to LLM's what is considered the "source" - the weights or the code used to build the weights?

Re: Grok3 Launch [video]

#930

[flagged]

I work, and the fact that my tax money is going into a black hole makes my blood boil. God bless Musk and DOGE for what they do. Here is just one headline from today, The Elon Musk-led Department of Government Efficiency (DOGE) on Monday revealed its finding that $4.7 trillion in disbursements by the US Treasury are "almost impossible" to trace, thanks to a rampant disregard for the basic accounting practice of using…

> Here is just one headline from today, The Elon Musk-led Department of Government Efficiency (DOGE) on Monday revealed its finding that $4.7 trillion in disbursements by the US Treasury are "almost impossible" to trace, thanks to a rampant disregard for the basic accounting practice of using of tracking codes when dishing out money.

And you believe them?

This is a department that fired multiple different nuclear weapons inspector and maintenance teams without knowing what their job was.

Had to re-hire them. They weren't redundant teams. DOGE just didn't understand what they (or the teams) were doing.

Now, I'm very happy for the US nuclear stockpile to shrink. I sure think you have too many of them. But then, I'm foreign and a hippy, so I would. But (1) do you?, and (2) do you want it to shrink by some of them accidentally exploding? Or being relocated by a hostile power taking advantage of the inspectors all being on early retirement?

Post reply on HN