KronisLV

8 ms·

61.
by KronisLV 2mo ago
I live in the EU, particularly in the part that Russia is looming over, Latvia. I'm quite glad that we're not a part of Russia, where even queer people are considered extremists and critiques of the illegal war will either carry a
62.
by KronisLV 2mo ago
> Prove that you are a robot > 60.0 s remaining > Write a 300-word essay about Napoleon in less than a minute. Heh, charming.
63.
by KronisLV 2mo ago
On a tech level I’d say that Kimi and GLM 5.3 on Max reasoning are good enough for non-trivial planning and exploration and on High are good enough for various implementation tasks. They can easily replace Opus 5 for me and mostly even Fabl
64.
by KronisLV 2mo ago
> Pre-paying for tokens is some kind of ideological red line for a lot of developers. If I can get a few subscriptions for 100 - 200 EUR month and NOT have to pay 3000 - 9000 EUR (based on ccusage and some other stats) in tokens then it’
65.
by KronisLV 2mo ago
Ohh this comes in a really good time! I mean personally I'd just like to use OpenCode with all the providers, if their desktop app was a bit more polished and Anthropic wasn't so restrictive. There's also Paseo, but it has so
66.
by KronisLV 2mo ago
Not affiliated with them, but this lets you view Claude Code, OpenCode and I guess other harnesses like Codex in the same session https://paseo.sh/ I didn't really care about their mobile app and the worst experiences
67.
by KronisLV 2mo ago
> something started changing around March. The economics catching up with the providers in regards to how much compute they can burn per request and have it make sense for them financially? A sort of model collapse where Opus 5 seems to
68.
by KronisLV 2mo ago
> just short-circuits to "there is no information here" That is my experience with the way the models write by default, often even when instructed not to do that. With enough effort you can get even them to slightly unslop the
69.
by KronisLV 2mo ago
> loser mindset. You can disagree without that. https://news.ycombinator.com/newsguidelines.html At the same time, I will also be the first one to admit that I don't have like 20-50k EUR laying around for the hardwa
70.
by KronisLV 2mo ago
> That's very defeatist. Do you have any concrete reason to think the major providers are lying to every one of their business/API customers about not training or storing the data? The business loss of trust would outweigh any
71.
by KronisLV 2mo ago
> the credit-taker Isn’t this what society already largely rewards? See the personality cults around CEOs and not even founders.
72.
by KronisLV 2mo ago
> it's speaking its own dialect, and you get used to it. Some might, I didn't - it just filled me with a sense of frustration and rage, alongside disgust because there is no good reason for that slop writing to drag everything
73.
by KronisLV 2mo ago
https://huggingface.co/Lord-H4D3ZS/Qwen3.8-Distill-35B-A3B-C... > Base / architecture: Qwen/Qwen3.6-35B-A3B (Qwen3_5MoeForCausalLM, 256 experts, ~3B active). The "3.8" in the name refers to the t
74.
by KronisLV 2mo ago
What were your experiences with ZeroMQ?
75.
by KronisLV 2mo ago
> Perhaps you can think of a way to design an alerting system which cannot be abused - but I can't. I'm pretty sure that the men with guns wouldn't be pleased at the system being so tamper proof that you can't do what
76.
by KronisLV 2mo ago
> Anyway, it is a basic practice of keeping test and dev environment as close as feasible to production, to avoid missing issues and wrong assumptions. Containers are great for this during development. Testcontainers are great for tests
77.
by KronisLV 2mo ago
> PostgreSQL allowed us to use a fulltext search plugin to do everything in one system. No need to sync any data. No need to maintain and run two systems. It just worked and made us smile (after some tweaks of course). Simplicity. I foun
78.
by KronisLV 2mo ago
> Should NVIDIA do a coding subscription too? Yes, obviously! Well maybe not a subscription but definitely an inference service. https://build.nvidia.com/ https://resources.nvidia.com/en-us-inference-infr
79.
by KronisLV 2mo ago
> Why they should go with Chinese models Because they generated some buzz and are near-SOTA and would be a great benchmark for a PoC subscription that doesn't necessarily aim to compete with other vendors at a similar scale (since t
80.
by KronisLV 2mo ago
> Why would they offer a coding subscription and start competing with some of their biggest customers; when they are capacity-bound and companies like OpenAI will take however many wafers that Cerebras sells to them? Because they already
81.
by KronisLV 2mo ago
> I've worked with enough people with fine coding skills and terrible English What about people with fine English and terrible coding skills? What about people with both terrible English and terrible coding skills? > Now, I often
82.
by KronisLV 2mo ago
Slop-ish site but the live diagram is neat!
83.
by KronisLV 2mo ago
> I was planning on using it for a design tool, but moved over to luna since it's comparable speeds and cost for a lot more intelligence. Everyone should occasionally go back to the old models to see how much worse they were, like e
84.
by KronisLV 2mo ago
Pretty bad efficiency then unless that only applies to Fable class but even then - Kimi K3 is around 3T and does similarly well in most benchmarks.
85.
by KronisLV 2mo ago
God I wish they'd back up all of those claims by offering a subscription of Kimi K3 and GLM 5.3, not some outdated GLM 4.7 instance that they then proceed to call a preview model and say that they'll remove it, leaving users only
86.
by KronisLV 2mo ago
> I think running a model unsupervised in anything above medium is a sucker move to burn more tokens. What about errors compounding and any on-the-go decisions being made being the wrong ones? That kills anything long form, no matter h
87.
by KronisLV 2mo ago
So I upgraded to their Max 20x for nothing, then. Guess Kimi K3 and GLM 5.3 it is, then.
88.
by KronisLV 2mo ago
Where’s the bullet point about the way we do programming being horrible? Not even the work, but the tools and languages afforded to us. github.event.issue.title is very obviously data. It should never be POSSIBLE to treat that as an instruc
89.
by KronisLV 2mo ago
OpenAI and Kimi are both pretty okay alternatives! I guess GLM 5.3 on Max reasoning as well but for more limited domains.
90.
by KronisLV 2mo ago
This one is really pleasant: https://woodpecker-ci.org/

More ›