Hacker News Reader

k__on Radicle: Sovereign {code forge} built on Git

Probably less issues with account management and rate limits.

k__on Why senior developers fail to communicate their expertise

I'd say, on averaged, it's 50% what you say and 50% communication issues.

Most smart juniors have no problem with learning. Perceptual exposure and deliberate practice works almost mechanically. However, if someone can't tell you what examples you should be exposed to, you'll learn crap.

k__on Nintendo announces price increases for Nintendo Switch 2

Has this ever happened before?

k__on Agents need control flow, not more prompts

At my new job, I was assigned to improve processes with AI.

My first thought was, well agents seem nice, but I think, AI workflows are a better bet. However, I don't really understood AI or agents in depth and felt like I was just "doing things the old way" and removing flexibility from agents was a ridiculous idea.

After some research I got the impression that I was right. A well defined workflow and scope is just what's needed for AI. It's cheaper and more consistent. It probably even makes the whole thing run well with non-SOTA models.

k__on Accelerating Gemma 4: faster inference with multi-token prediction drafters

That matches my experience.

GPT and Claude would work much better than Gemini, even if the direct feedback was sparse or diffuse.

However, the moment I gave Gemini a fast testing framework that gave it instant feedback, it would mill through all kind of problems.

Claude and GPT are seniors.

Gemini is a very motivated mid level.

k__on Vibe coding and agentic engineering are getting closer than I'd like

To be fair, MS send a world class engineer to make JavaScript usable for codebases at that scale.

k__on Computer Use is 45x more expensive than structured APIs

Agentic engineers can build well defined, very deterministic middleware on top of OpenRouter.

Anthropic even says, that an agent based solution should only be your last resort and that most problems are well served with a one-shot.

https://www.anthropic.com/engineering/building-effective-age...

k__on Dutch central bank ditches AWS and chooses Lidl for European Cloud

European central bank will probably go for Aldi.

k__on Eden AI – European Alternative to OpenRouter

How come?

You can use it with just European models if you want.

k__on Saying goodbye to Agile

"Agile was always aiming to solve the wrong problem (that code is the bottleneck)"

No, it aimed to solve the "out specs are bad and we need to iterate faster" problem.

"a massive lie exposed by LLMs"

No. LLMs add no insight about the problem and they expose nothing. They just help to engage this well-known problem with another tool.

k__on Book review: There Is No Antimemetics Division

Writing good endings is hard.

I liked Ra, but I liked Fine Structures more.

k__on Book review: There Is No Antimemetics Division

I liked piecing the story together in the SCP wiki.

Later I read the first version of the book and it was okay, but the vibes were a bit lost.

The new version of the book I didn't even finish.

k__on Why are we still using Markdown?

Because there's only one reasonable implementation of AsciiDoc

k__on Proton Meet Isn't What They Told You It Was

Sadly this is not binary.

k__on Ukrainian drone holds position for 6 weeks

Maybe drones will make human soldiers unacceptable in the future.

k__on Slop is not necessarily the future

"No one has ever made a purchasing decision based on how good your code is."

Because to determine good code, you need to see it and I'd presume most open source code is free.

k__on We haven't seen the worst of what gambling and prediction markets will do

Half-OT: when does gambling become safe?

I mean, insurance is basically basically betting that bad things happen to you.

k__on Goodbye to Sora

Fair.

I just have the feeling that it doesn't get the job done anymore.

I hope we will see the rise of alternatives.

k__on Goodbye to Sora

"coding is the prime use case for this where you can make money"

Is it?

I have the impression GenAI deteriorates the internet both from a content and tech perspective.

Bots that waste your time because they don't work well or because they are pushing an agenda, and low quality content that floods social media from people who want to make a quick buck.

GitHub and AWS became increasingly unstable. X, Instagram, and WhatsApp are suddenly sprinkled with subtle bugs.

Everything just got faster and we got more of it, but nothing of it is good anymore because everyone tries to replace 90% of their work with GenAI instead ofmaybe starting at 10-20% and then add more when you're sure it works.

k__on The three pillars of JavaScript bloat

Doesn't this go against the credo of not building your own crypto?

k__on Silicon Valley's "Pronatalists" Killed WFH. The Strait of Hormuz Brought It Back

Yes, I was surprised that my girlfriend found a job in a big company where she only has one office say a week.

k__on Silicon Valley's "Pronatalists" Killed WFH. The Strait of Hormuz Brought It Back

As a German living in Germany, I had the impression most companies here are anti-WFH.

It got better with COVID, but you still have to dig, to find something 100% remote.

k__on Beyond has dropped “meat” from its name and expanded its high-protein drink line

In my experience, the pea-based products are pretty good.

I'm a huge burger fan and stopped eating meat at home, thanks to this wave of vegan alternatives.

k__on 1M context is now generally available for Opus 4.6 and Sonnet 4.6

I heard, the middle of the context is often ignored.

Do long context windows make much sense then or is this just a way of getting people to use more tokens?

k__on The Bitter Lesson Has No Utility Function

Yes, the issues is that LLMs sometimes fill semantic gaps.

When I write an article and I notice that there are steps missing for a conclusion then I do some research and fill them in.

These gaps often only show up when drafting, so if the LLM drafts, it might see a need to fill them in. You read the draft and see no steps are missing, but since you didn't do the research for the filled in steps, and they "look good to you", you might miss errors.

k__on The Gervais Principle, or the Office According to “The Office” (2009)

I'm consulting or otherwise feral for about 12 years.

Didn't make me less clueless, I think...

k__on The Gervais Principle, or the Office According to “The Office” (2009)

I liked that model a lot, but it made me a bit sad too.

All my life I was bad at being a loser, somehow I never really felt I fit in. I thought this was because of psychopathic tendencies or something. However, after reading this I realized there was another option and I was just clueless.

k__on Grammarly is offering ‘expert’ AI reviews from famous dead and living writers

I offer my expertise in tech writing to review your AI articles and docs.

k__on Relicensing with AI-Assisted Rewrite

How do you prove the training data didn't contain the code?

I'd assume an LLM trained on the original would also be contaminated.