Viewing profile — KronisLV
KronisLV
HN member- Joined
- Sat, Sep 26, 2020, 7:32 AM UTC
- HN karma
- 8,278
- Public activity
- 3,755 items
- HN profile
- View on Hacker News ↗
About KronisLV
Recent public activity
-
comment
Comment #49223505
> GLM 5.3 bridges this gap. I’m sure the next models will only get better, when they’re released. Also super curious about what Moonshot will achieve and the full DeepSeek V4 Pro r…
-
comment
Comment #49223331
For what it's worth, I daily drive Edge on my Windows machine and it's an okay browser, haven't had a reason to install Chrome in ages (though did run Firefox for a while due to li…
-
comment
Comment #49219915
> Why, people? As someone who's had to use the Oracle database and also Oracle Linux on servers, beats me. Also stuff like Oracle ADF and JDeveloper and JBoss/WildFly, no thank you…
-
comment
Comment #49219903
> The Chinese models are not good enough for anything other than pair programming, which is just a very last-gen way of using agents. This matches my experience with DeepSeek V4 Pr…
-
comment
Comment #49219832
> You use Fable 5 right? If that’s good enough for you now, why wouldn’t a Chinese model that’s as good as Fable 5 but at 10% the cost be good enough in 6 months? This is why I qui…
-
comment
Comment #49210705
USB3-5A-10Gbps-DP-TB3-C would be pretty gnarly as a label even if it’d still be a Type C at the end of the day.
-
comment
Comment #49210533
> Deployments not using Docker will no longer be supported. This is the way (while also allowing people to use their own proxies etc., but being clear about those not being officia…
-
comment
Comment #49208657
What about quality of life per watt and just making having a certain standard of living more energy efficient? Simple examples: energy efficient lights seem to have no downsides, b…
-
comment
Comment #49207834
I feel like LLMs could produce some pretty cool fiction - like cranking up the temperature setting like you can with many models to produce nonsensical garbage that isn't even real…
-
comment
Comment #49185501
> I don't really have an issue with Oracle doing something any sane business would to be honest. I don’t have an issue with the reduction, but the communication around it could hav…
-
comment
Comment #49136707
> Claude Code injects a ton of tools into the system prompt It's so unfortunate they don't let you use the subscription with other harnesses anymore - since even if I used OpenCode…
-
comment
Comment #49134150
> will it ever terminate and say the thing is done and bug free? Will human reviewers, if you ask them to find more bugs? Will it be because there really are no bugs, or because th…
-
comment
Comment #49132865
> Then cry as the ai reveals that it didn't actually do anything close to what it said it did. If using AI to generate code, you told it generate some code, so it did. No amount of…
-
comment
Comment #49132416
> I'll await your Rust ports of Windows, Linux and MacOS then. I’d love to have that happen within the next century. I might not be around for it, but I’d say that humanity deserve…
-
comment
Comment #49120079
> LLMs are trained to write consistent code You are absolutely right, that's the load bearing assumption and it earns its keep. Personally, I didn't mind when a lot of sites looked…
-
comment
Comment #49119985
Wonder how good the proper version of V4 Pro will be. I'm still considering pulling the trigger on the annual subscription of Kimi for K3 but it's sometimes slower than I'd like (a…
-
comment
Comment #49107986
> That is true, but I feel like sometimes if the conversation contains useful rational it can help to keep it. I think sometimes it is a judgement call, I will sometimes compress t…
-
comment
Comment #49104017
Existing project and as usual, a few issues mashed together in one mostly coherent plan. The shorter Kimi sessions I mentioned were singular features, there 256k would be wholly ad…
- comment
-
comment
Comment #49102458
> Are they really that good, They did exceed the expecations of pretty much everyone! I've also blogged about using the model, it's a bit on the slow side but pretty good! > so wer…
-
comment
Comment #49102385
My current Claude Code session has been going on for like 35 hours and has used up around 400 million tokens, thankfully almost all of those being cached (95-98%) - pretty typical …
-
comment
Comment #49101789
Might be a bit more than just noise: https://marginlab.ai/trackers/claude-code-historical-perform... Edit: well, crap. Model overloaded retrying (6/10) 1m 40s Thanks, Anthropic.
-
comment
Comment #49101755
For me it got exhausted faster than I'd like, subjectively less value than the Anthropic 100 USD subscription, BUT their 5 hour limits are very generous, whereas I keep bumping int…
-
comment
Comment #49101589
> There isnt a definitive right or wrong As someone who struggled with the downright problematic lifecycle of complex JSF/PrimeFaces apps and now is in a year-long-with-no-end-in-s…
-
comment
Comment #49080211
> Whatever they do in their country is their own problem. Idk to me it feels like human compassion and ethics dictates that you at least have to care a bit even about the things in…