golergka

7 ms·

31.
by golergka 1y ago
I did a couple of quick searchers with help of ChatGPT, and it seems like in Norway, at least, a tenured professor would get ~$50k post-tax, a primary school teacher ~$35k, and a cleaner ~$20k. If anything, such low income inequality seems
32.
by golergka 1y ago
So all engineers who have already been using better AI tools and had no reason to downgrade to copilot and cursor were fired?
33.
by golergka 1y ago
Isn't any kind of human in the loop make system non-deterministic?
34.
by golergka 1y ago
Sometimes I do OP’s approach, sometimes yours, but in all cases, writing down what you need done in detailed English gets me to a better understand of what the hell I’m even doing. Even if I wrote the same prompts and specs and then typed
35.
by golergka 1y ago
4o-mini costs ~$0.26 per Mtok, running qwen-2.5-7b on a rented 4090 (you can probably get better numbers on a beefier GPU) will cost you about $0.8. But 3.5-turbo was $2 per Mtok in 2023, so IMO actual technical progress in LLMs drives pric
36.
by golergka 1y ago
When I pay attention to o3 CoT, I notice it spends a few passes thinking about my system prompt. Hard to imagine this question is hard enough to spend 13 seconds on.
37.
by golergka 1y ago
That’s a good thing. You want an LLM to know about product or service you are selling and promote it to its users. Getting into the training data is the new SEO.
38.
by golergka 1y ago
Ironically, cloudflare is also the reason OpenAI agent mode with web use isn’t very usable right now. Every second time I asked it to do a mundane task like checking me in for a flight it couldn’t because of cloudflare.
39.
by golergka 1y ago
This is also something that an LLM would be able to easily explain. I'm pretty confident that all 4o-level models know these facts without web search.
40.
by golergka 1y ago
Just out of pedantry, all of Mexico is in North America.
41.
by golergka 1y ago
That’s why I like using it and get more fulfilment from coding than before: I do the fun parts. AI does the mundane.
42.
by golergka 1y ago
Why would you use Adblock if you pay for premium?
43.
by golergka 1y ago
> she is a ‘minority shareholder’ in the app development company Let's suppose this means she is directly linked to this anti-ICE app, profits from it (if it even has any revenue?) and is interested in the apps success. Does it mean
44.
by golergka 1y ago
Syntax and type errors gets instantly picked up by type checker and corrected, and as long as these failures stay in context, LLM doesn’t make those mistakes again. Not something I ever have to pay attention to.
45.
by golergka 1y ago
Recently I wrote a significant amount of zig first time in my life thanks to Claude Code. Is zig a mainstream language yet?
46.
by golergka 1y ago
No, at this point it can just ingest official library documentation and produce good SO-like answers.
47.
by golergka 1y ago
This looks very similar to the pit of success concept: https://blog.codinghorror.com/falling-into-the-pit-of-succes...
48.
by golergka 1y ago
that's what lazygit in another terminal tab is for
49.
by golergka 1y ago
Ah, the Ballmer peak.
50.
by golergka 1y ago
> dDevelop guidance Good to know it was written by a human and not an LLM.
51.
by golergka 1y ago
> The author implies that Grok 3 becoming racist because of a system prompt is a bad thing. He didn't "become racist". Megahitler Grok defended completely opposite political opinions in different threads, just depending on
52.
by golergka 1y ago
Prescriptive vs descriptive. I see absolutely no reason why public utilities shouldn’t operate on exactly the same principles, and I see plenty from real world examples on why they actually should.
53.
by golergka 1y ago
That’s not how supply and demand work
54.
by golergka 1y ago
One agent and some hard code to extract doc diffs with relevant code, parallel agents for different rule groups, tool agent to look up existing patterns and related material in the codebase, consolidator agent to merge the comments and sugg
55.
by golergka 1y ago
Would you pay (a very small amount) for it? As another commenter absolutely correctly pointed out, just putting this guide and your diff into ChatGPT would yells bad results, but looks like something absolutely doable with a proper multi ag
56.
by golergka 1y ago
I don't think this makes any sense as a standalone product, but I wish I had it in my claude code and aider as an intermediate step right away. The latter tool, while less popular, already supports _exactly_ this workflow in the archit
57.
by golergka 1y ago
> something that is arguably essential for the majority of people Our perception of this changes with time, radically. A few ages ago a chair would have seemed unneeded luxury for most of the world's population. > Funny enough, t
58.
by golergka 1y ago
Dystopias are usually about things getting worse than the they are now, or at least worse than they were at some point in time. Was there some point in time when therapy was available to everybody who needed it? Or may be we just got more u
59.
by golergka 1y ago
Great point. Also, it may not be the best possible API designer in the world, but it sure sounds like a good way to forecast what an _average_ developer would expect this API to look like.
60.
by golergka 1y ago
IMO it's still good that it's family's decision. Even if it is an incorrect one.

More ›