News for hackers. Stuff that matters.

volf_

Re:Kimi K3.1 Pro and K3.1 Pro Highspeed on Friday October 23, 2026 @01:13PM (#50212303)
That's quite an impressive jump
> with severe ADHD, who has never taken any medication > going through mental health treatments and get special care and privileges in society

1. How could you know ADHD medication numb their responses if you've never taken meds?

2. I hope you have the live you deserve. You seem like a happy person.

I can see this being useful. My agents generate useful findings that I've been thinking about how to share them.

What makes or breaks this is the search. Ensure every possible search option is exposed to give maximum granularity to what agents can search for.

Re:Xiaomi MiMo v2.6 on Monday September 21, 2026 @05:22PM (#49793637)
I've got a working recipe to run this model on Dual DGX Spark: https://github.com/volfco/spark-vllm-docker/blob/main/recipe...

Averages ~25-35tok/s which isn't bad for a first attempt.

Re:Ox-Alpha Is GLM on Tuesday August 25, 2026 @01:14AM (#49429353)
same business model as crack. best way to get people hooked is to make the first hits free.
Re:Ox-Alpha Is GLM on Monday August 24, 2026 @02:31PM (#49424014)
It's possible
Re:Ox-Alpha Is GLM on Monday August 24, 2026 @02:31PM (#49424000)
Yeah. It could be. The Z.ai DC latency is still ~1.2s faster than whomever is serving this model.
Re:Ox-Alpha Is GLM on Monday August 24, 2026 @01:54PM (#49423512)
GLM 5.3 and all previous models don't have a vision encoder and can only accept text. Ox-Alpha can accept video and images, so unless Z-ai added a pretty good vision encoder for this model, I don't think so.

My money is on Moonshot and this being Kimi K3.5. The measured tps and latency is in-line with K3's tps and latency from Moonshot.

MiniMax M3.5 is also possible (but the MiniiMax provider is a lot more performant than the lab behind ox-alpha, so less likely).

Re:Qwen/Qwen3.8-2.4T-A95B on Wednesday August 12, 2026 @11:28AM (#49273947)
A ~5TB model.
Not mentioned in the article, but this is a x86 architecture because Hygon has a license for Zen core IP from AMD.
Re:Annoying and alarming things about OpenCode on Monday July 20, 2026 @09:26AM (#48978573)
I switched from OpenCode to Pi and there was a big improvement in terms of tool calling performance and I find the experience less buggy.

OpenCode has also seemed to have disappeared from https://openrouter.ai/apps/category/coding

Re:Fable extended until 19 July on Sunday July 12, 2026 @01:34PM (#48882869)
Keep using Fable. Don't think about using GPT 5.6 Sol
Re:Ask HN: Developers still enjoying development after AI? on Wednesday March 11, 2026 @03:28PM (#47340054)
No.

the amount of coworkers who have said "<ai tool> said X,Y,Z". Then they make basic mistakes with what they need to do and it becomes my problem.

atproto is a really good attempt at solving this issue
do as we say, not as we do
Re:Problems with D-Bus on the Linux desktop on Monday December 15, 2025 @02:52PM (#46279535)
The world is filled with "good enough" solutions. If it has 80% of the functionally needed, it will be good enough for most use cases.

> the most meritorious solution never rises to the top

You know why? Because it was the best solution at the time that covered everything they needed. You can only complain about it because the initial problem was solved and people were able to move on to something else.

Does it work for what we need it to do? Yes? Ok. Now that we have settled this problem, we can continue on developing.

Meritocracy is only relevant to someone who has ideas but no desire or ability to actually structure the idea in such a way that would work in the world we find ourselves in today.

> Maybe D-Bus has something similar?

So you just commented here to say nothing but "dbus sucks because it isn't the best. I wonder if there are better options?".

That's awesome.

These are the best kinds of posts

Re:Everyone in Seattle hates AI on Wednesday December 03, 2025 @10:00PM (#46143309)
because autocorrect and predictive text doesn't help when half your job is revisions
Re:Aldous Huxley Predicts Adderall on Monday November 17, 2025 @04:59PM (#45958847)
Exactly.

Adderall has no positive relationship to my mental efficiency. It can in fact be a negative once your passed the 8 hour windows where its still in your system.

At the end of the day, it makes it easier to not bounce between different things. It doesn't help me be smarter. It helps me drive to work without needing to listen to music and be on my phone.

Modafinil... maybe.

Re:ADS-B Exposed on Wednesday October 15, 2025 @05:13PM (#45598446)
"Elon Musk" as a filterable option.

A++. No Notes.

Re:Algorithms for Modern Processor Architectures on Wednesday July 23, 2025 @12:48AM (#44655795)
A commercial enterprise is dropping support for older cpu architectures in their newer OSs so they can improve the average performance of the deployed software?

Don't see how that's controversial. It's something that doesn't matter to their customers or their business.

Re:Telephone Line Rural Outside Plant on Monday September 02, 2024 @11:49AM (#41426451)
Also check out https://www.youtube.com/@ConnectionsMuseum

The channel is about the fully functional Central Office turned Museum in Seattle

Re:Telephone Line Rural Outside Plant on Monday September 02, 2024 @11:44AM (#41426413)
Yes. Dynamic as the IP is renewed everytime your device resets
Re:Telephone Line Rural Outside Plant on Monday September 02, 2024 @11:43AM (#41426401)
https://www.youtube.com/@ConnectionsMuseum

There's a fully functional system deployed in Seattle as a Museum

You clearly don't use TikTok. There is quite a lot more than just dancing for 15 seconds.

> Any time one hears someone mention the phrase "content creator", one should, by default, execute s/content creator/manipulative parasite/ in their mind.

k

TikTok has one of the lowest bars to monetization out of all the major social media networks.

There is a large amount of creators on the platform that live off their content and eCommerce enabled through the platform. So it disappearing overnight would severely impact people who have a majority of their livelihood through the app.

Facebook. Twitter. YouTube. They all have short form video, but TikTok has a lot of features that just don't exist on other platforms.

You could have `./configure` print a message saying to do exactly what you said.
Re:Ask HN: Who wants to be hired? (February 2024) on Sunday February 04, 2024 @12:10PM (#39252076)
Location: Seattle, WA

Remote: Yes

Willing to relocate: Maybe, for the right gig.

Technologies: Python, Rust, Kubernetes (Helm, ArgoCD, CNPG, AWS ACK), AWS, AWS EKS, AWS CDK, Terraform, Grafana LGTM (Loki, Grafana, Tempo, Mimir), GitLab, PostgreSQL, OpenTelemetry, System Architecture, Hybrid Cloud, DevOps/SRE stuff, and more!

Resume: https://www.linkedin.com/in/volf/

Email: hn [at] m1.volf.co

Been doing DevOps/SRE for the past 7-8 years. I've found myself working well in positions where I can work closely with other engineers to build out Cloud/Distributed Infrastructure and develop software in a way that makes it not painful to run in the cloud.

I can help take small MVP applications & infrastructure, and help turn it into something that can scale up as the business does. I'm comfortable working with any stack- just depends on how much time it'll take me to brush up on the language or design.

Older »