Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
Zambyte
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
3 ms
·
31.
▲
by
Zambyte
16d ago
People would absolutely choose trains over flying if we actually had modern train infrastructure. NY to LA would be about 11 hours at 250 MPH. That's 100 MPH slower than the fastest passenger train, but seems like a reasonable estimate
32.
▲
by
Zambyte
16d ago
Now do China.
33.
▲
by
Zambyte
16d ago
My apartment overlooks a three way intersection. Drivers block the box almost every light cycle in rush hour. The most common traffic offense I see bikers make is riding on the sidewalk, which they obviously do because they fear being kille
34.
▲
by
Zambyte
17d ago
And ideally enough mass to violently crush their neighbors.
35.
▲
by
Zambyte
17d ago
Distilling is a massive achievement. I can run Qwen. I can't run GPT (TM). It's not a matter of X is better than Y. It's a matter of Y exists, X does not.
36.
▲
by
Zambyte
17d ago
I wonder if the prevalence of adblock has helped make a lot of things like this possible. Like, it's hard to imagine people putting up with this without literally protesting in the streets. But the people who can't tolerate it hav
37.
▲
by
Zambyte
18d ago
Advocating vs blaming. Blaming people for using inefficient vehicles when you eat red meat (assuming you eat enough red meat to have more of an impact than the fuel) is hypocritical. Advocating for better alternatives to inefficient vehicle
38.
▲
by
Zambyte
18d ago
Being flawed isn't the issue. Everyone is flawed. The issue is blaming people for problems like environmental damage, while at the same time needlessly causing more damage yourself. The problem is being hypocritical. Some people are ce
39.
▲
by
Zambyte
18d ago
The vast majority of people simply don't need SOTA models. I run most of my programming on a single 7900 XTX with Qwen 3.8 27b. I'm more interested in scaling out to run more instances of the model in parallel, than I am in scalin
40.
▲
by
Zambyte
19d ago
Unix philosophy: tools should be simple and composable Windows philosophy: don't get sued by AT&T
41.
▲
by
Zambyte
19d ago
Something else I want to add on to be more specific is that I use skills to document tasks that my model doesn't know how to do out of the box. For example, there is a jira cli[0] that I use for interfacing with jira. If I just say som
42.
▲
by
Zambyte
19d ago
Your agents.md is a good place for high level facts, but if you have something that requires a lot of info to explain (ie: if there is a complex build process, testing patterns, things like that), loading up your agents.md for every request
43.
▲
by
Zambyte
20d ago
AI is great for offloading problems that you don't want to solve, but you want done. The key is finding what problems you find important to solve yourself. At this point, roughly all the code I push is generated by AI. I find the code
44.
▲
by
Zambyte
21d ago
You can just add it in a user-level gitignore instead of ignoring it in every repo. See: ~/.config/git/ignore
45.
▲
by
Zambyte
21d ago
I found this ubder the Play it now! page. Funny name collision with the Scala programming language :) https://www.huygens-fokker.org/scala/
46.
▲
by
Zambyte
21d ago
I wonder what data Elon is able to slurp up through the app that he can't get through the website. Seems like a good incentive to keep the site worse.
47.
▲
by
Zambyte
21d ago
I cut my Claude Code token usage by 100% by writing the code with Qwen 27B on my own GPU.
48.
▲
by
Zambyte
21d ago
hasn't*
49.
▲
by
Zambyte
21d ago
How can you define "general" and "intelligence" in a way that has existed for years now?
50.
▲
by
Zambyte
22d ago
Seconding the TUI usage. Opencode, codex, and pi are ones that I have / my friends have had success with using locally. I mostly use pi. It feels like the most "boring tool that just does its job" out of the big options.
51.
▲
by
Zambyte
22d ago
What are people even using embeddings for these days? It certainly seems like giving an agent grep covers most of the use cases. Dare I say: grep is all you need.
52.
▲
by
Zambyte
22d ago
Qwen 3.8 27b has the juice. Try it.
53.
▲
by
Zambyte
22d ago
> What I think should happen is that it should update its memory with notes on the proficiency level of the user, so it gets the balance right over time. Which can also involve just asking for the users level of experience
54.
▲
by
Zambyte
22d ago
Genuinely, how do you make your phone so slow? I've been using my s10 for 7 years, and besides the minor crack in the corner of the screen from being dropped, and constantly juggling storage because I hoard media, it works as well as i
55.
▲
by
Zambyte
22d ago
It's efficient in terms of energy, but slow after decades of neglect. People pay for fast.
56.
▲
by
Zambyte
22d ago
Digital media cannot be transacted without being copied.
57.
▲
by
Zambyte
23d ago
Discord is simply a liability.
58.
▲
by
Zambyte
23d ago
At 1500 tps, "uh" is about 0.7 ms, instead of 200-300 ms for a human.
59.
▲
by
Zambyte
23d ago
For digital products, that would mean copyright privileges. Else there is no definition of "buy" that applies to GOG sales either.
60.
▲
by
Zambyte
23d ago
This would be case-by-case in Steam. There are some games which really do sell under "irrevocable license to play" (which, short of granting copyright privileges, is about as close as you can get to ownership for a digital product
More ›