Hacker Newsnew | past | comments | ask | show | jobs | submit | Vegenoid's commentslogin

No, because the function of code is distinct from the implementation of the code. With software, something that is functionally identical can be created with a different underlying implementation. This is not the case with media.

Will you elaborate on the “indefensibly bad legislation”?

This is just unrealistic day dreaming. You can go be in a field picking produce for work - we have a shortage of these laborers. Most people don’t actually want to do that, they want a cushy office job that doesn’t wear down their body and that offers them the ability to increase their skill and value over time.

The software engineer who thinks they’d be happier working in a field is largely just a grass is always greener phenomenon. It turns out that for most people, they don’t like work whatever it is, because work is done not by choice but by necessity.


As a teenager I worked at a company that rented rafts for a short trip down a river. We’d take the rafts from the customers at the end and truck them back up to the start. As they became bigger and busier, it became more important to keep track of the status of rafts and know when they were going to be getting back to the top.

They paid tens of thousands to have software made for this purpose. It sucked and was totally unable to handle the simultaneous realtime access and modification that was required.

They knew I was good with computers, so asked me if I had any ideas. In about an hour I made them a Google Sheet that worked great for the next several years until I left.


I pay for Kagi to get better search results. Lately, I’ve felt that Kagi’s search has been just as full of low-information and AI generated results as Google. I’ve been wondering why I’m still paying for it. This seemed like a good litmus test. Unfortunately, Kagi displays pretty much the same results as Google for nanoclaw.


Yeah that's increasingly been my feeling as well. I have to keep prefacing my Kagi recommendations with, "web search is less and less useful every year, but..."

I still appreciate being able to customize rankings, bangs, and redirects. But with how utterly shit the web is overall, any web search is basically only good if you know the site(s) the answer(s) will be on. When you're searching for something novel-to-you, even Kagi is just going to show you a full page of unregulated slop on the dumbest, just-registered-this-year domains. Real information is increasingly limited to small islands of trust.


Isn't Kagi basically just using a blocklist? In which case it's whack a mole as new sites spring up or bubble up to the top of other results. I keep my own blocklist and intermittently search key phrases to blanket block new sites, and there's often new sites popping up.


This sounds very interesting, could you elaborate on your methods and tools?


https://github.com/iorate/ublacklist

Basically,

- whenever a spam site is in my results, click to block - click through some and check out pages like "About" "Terms" "Disclaimer" "Privacy". often common boilerplate phrases abound. affiliate marketing spam, and similar, for example - exact string search those phrases - (get often many pages of results) - using a userscript, block all domains in each result page

my results have gotten so much better for common types of queries. even then, new sites pop up almost daily it seems. add to the list.


I don’t like any search engines now :(


It's because the search engine is being eaten by the LLM. I'm not suggesting that it's a perfect substitute. It's just what I feel is happening.


more like LLM garbage are rotting search engines from the inside out


Naw this is a pre LLM problem.


Yep. SEO spam has been a thing for decades.

LLMs have supercharged it though, it's so much easier to create dozens or hundreds or thousands of ultra low effort LLM written webpages and websites that it ever was before LLMs.


I'm not talking about LLMs diluting search. I'm saying users are using LLMs to search more than search itself, including in search engines.


I hadn't really noticed anything like this until you pointed it out. My main use for Kagi is to pin Wikipedia results... I just tried searching for "nanoclaw" on Kagi (I'm in the UK so results biased towards there) and got:

1. nanoclaw[dot]net (!)

2. github.com/qwibitai/nanoclaw which looks like a ripoff?

3. Three videos, at least one of which looks like slop with crypto ads

4. github.com/gavrielc/nanoclaw which I presume is the real repo judging from the name?

5. Three "interesting finds" the top one of which is nanoclaw.dev, but with the title "Don't trust AI agents" because it's a blog post from that site

6. A fork of the qwibitai/nanoclaw repo


> 2. github.com/qwibitai/nanoclaw which looks like a ripoff?

That is literally the GitHub repo the original article shows as being "real".


> it remains structurally constrained by separation of powers, federalism, and independent courts and media, features that fascist regimes systematically dismantle.

But the Trump administration (a.k.a. the executive branch) is trying to systematically dismantle these things. When people refer to fascism in the U.S. government, they don’t mean the entire government. They mean the Trump administration, which is the face of the U.S. government, has a great deal of the share of power, and is seeking more. The brand of far right nationalism, that the nation is in decline and violence must be used to restore it, along with the economic policies and deference to corporations and the wealthy, are things that make them more fascist than just authoritarian.

Saying that we cannot call it fascism until the transformation is complete doesn’t make sense - if a group of people have beliefs and goals that align with fascism, and are taking steps to impose them, you can call them fascist, even if they have not yet realized the full set of conditions that make the government fully fascist.


The program perhaps most responsible for the renewed interest in the terminal, Claude Code, is made with React. Life moves pretty fast.


What models are you all using with Pi? How much are you paying on monthly or weekly basis for your usage? I'm very interested in it, but my budget is constrained and the usage granted by a $20/month Claude plan seems much more affordable than when I've tried API-based access with other agents. Unfortunately, this locks me in to Claude Code.


Qwen3 coder next works very well. I started with 20usd/month ollama cloud and switch to 200usd then decided host my own llm with a gigabyte ai atop since now i'm using llms agents everywhere on my home lab


Use OpenRouter, lots of great open-weights models like MiniMax, Kimi K2, Mistral, Qwen, ...


so are people pairing OpenRouter with open agent harnesses like Pi or OpenCode?


I really wish they were called lessons instead of skills. It makes way more sense and prevents the overloading of the term "skill".


There is some papers [0] showing that the skill and agent files reduce the reasoning effectiveness in some use cases (e.g. autogenerated)

[0] https://arxiv.org/abs/2602.11988

reference: https://news.ycombinator.com/item?id=47034087


We'd just be overloading "lessons" as well, and even more so because it takes more work to ground the concept, given its larger semantic distance from what we're describing.


You haven't addressed the parent's concern at all, which is that what the LLM was trained on, not what was fed into its context window. The Linux driver is almost certainly in the LLM's training data.

Also, the "spec" that the LLM wrote to simulate the "clean-room" technique is full of C code from the Linux driver.


This is speculation, but I suspect the training data argument is going to be a real loser in the courtroom. We’re getting out of the region where memorization is a big failure mode for frontier models. They are also increasingly trained on synthetic text, whose copyright is very difficult to determine.

We also so far have yet to see anyone successfully sue over software copyright with LLMs—-this is a bit redundant, but we’ve also not seen a user of one of these models be sued for output.

Maybe we converge on the view of the US copyright office which is that none of this can be protected.

I kind of like that one as a future for software engineers, because it forces them all at long last to become rules lawyers. If we disallow all copyright protection for machine generated code, there might be a cottage industry of folks who provide a reliably human layer that is copyrightable. Like Boeing, they will have to write to the regulator and not to the spec. I feel that’s a suitable destination for a discipline. That’s had it too good for too long.


Okay, so will companies now vibe-code a Linux-like license-washed kernel, to get rid of the GPL?

> The Linux driver is almost certainly in the LLM's training data.

Yes, and? Isn't Stallmans first freedom the "freedom to study the source code" (FSF Freedom I)? Where does it say I have to be a human to study it? If you argue "oh but you may only read / train on the source code if you are intending to write / generate GPL code", then you're admitting that the GPL effectively is only meant for "libre" programmers in their "libre" universe and it might as well be closed-source. If a human may study the code to extract the logic (the "idea") without infringing on the expression, why is it called "laundering" if a machine does it?

Let's say I look (as a human) at some GPL source code. And then I close the browser tab and roughly re-implement from memory what I saw. Am I now required to release my own code as GPL? More extreme: If I read some GPL code and a year later I implement a program that roughly resembles what I saw back then, then I can, in your universe, be sued because only "libre programmers" may read "libre source code".

In German copyright law, there is a concept of a "fading formula": if the creative features of the original work "fade away" behind the independent content of the new work to the point of being unrecognizable, it constitutes a new work, not a derivative, so the input license doesn't matter. So, for LLMs, even if the input is GPL, proprietary, whatever: if the output is unrecognizable from the input, it does not matter.


> Let's say I look (as a human) at some GPL source code. And then I close the browser tab and roughly re-implement from memory what I saw. Am I now required to release my own code as GPL? More extrtsembles what I saw back then, then I can, in your universe, be sued because only "libre programmers" may read "libre source code".

It's entirely dependent on how similar the code you write is to the licensed code that you saw, and what could be proved about what you saw, but potentially yes: if you study GPL code, and then write code that is very uniquely similar to it, you may have infringed on the author's copyright. US courts have made some rulings which say that the substantial similarity standard does apply to software, although pretty much every ruling for these cases ends up in the defendant's favor (the one who allegedly "copied" some software).

> So, for LLMs, even if the input is GPL, proprietary, whatever: if the output is unrecognizable from the input, it does not matter.

Sure, but that doesn't apply to this instance. This is implementing a BSD driver based on a Linux driver for that hardware. I'm not making the general case that LLMs are committing copyright infringement on a grand scale. I'm saying that giving GPL code to an LLM (in this case the GPL code was input to the model, which seems much more egregious than it being in the training data) and having the LLM generate that code ported to a new platform feels slimy. If we can do this, then copyleft licenses will become pretty much meaningless. I gather some people would consider that a win.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: