Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

vladimir.varank.in

Post reply on HN
Print view View original

291–300 of 397 posts

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#291

Post by nicman23 »

overfeed wrote: post #89
dmix wrote: post #53

> Instead of continuing with the code, I spawned a fresh Pi session, and asked the agent to write a detailed specification of how the brcmfmac driver works Planning markdown files are critical for any large LLM task.

The line between AI-assisted clean-room reverse-engineeing and open-source-license-laundering is a thin one, and I think the one described in the article crosses over to laundering. In classic clean-room design, one team documents the interfaces - not the code.

i mean clean room was always license laundering and an AI agent cannot hold any copyright and it is largely not the same code

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#292

Post by nozzlegear »

majormajor wrote: post #272
shermantanktop wrote: post #250

Earlier quoted context omitted.

Can you name me another time when humanity has run out of useful work to do? Was it when we tamed fire, invented the wheel, writing, or double entry bookkeeping? All of which appear more consequential than current AI. We’ll always have something to do. And humans like doing things.

The claim of the AI true-believers is that this time it will be different because of the "general" nature of it. Fire can't build a house. The wheel can't grow crops. Writing can't set a broken bone. Double entry bookkeeping can't write a novel. If you believe that this AI+robotics wave will be able to do anything a human can do with fewer complaints, what would the humans move on to?

Fighting the clankers, of course.

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#294

Post by gck1 »

This is the third time I see pi mentioned over the last few days and pi is the first project where every writeup I've read about it is actually helpful and goes into details on how things were done and what things were built, with git repos. This is a common complaint on HN.

Now, since Claude Code is banning accounts for usage of pi (or rather, how pi is configured to use Claude models), how complicated would it be to wire pi through Anthropic's harness and treat anthropic harness as a dumb shell?

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#295

Post by jeroenhd »

Gigachad wrote: post #23
b8 wrote: post #16

It'd be nice to have drivers for newer Mac's for a better Asahi Linux experience. Good use of AI imo.

AI wouldn't work here. The OP task was converting one open source driver in to another one for FreeBSD. Since Mac doesn't have open source drivers to start with, a person still has to do the ground research. At least until you can somehow give the AI the ability to fully interact with the computer at the lowest levels and observe hardware connected to it.

Someone else here suggested having an AI write a filter driver to intercept hardware communications on Windows and try to write a driver based on that, I presume macOS can also be coerced into loading such a driver?

That approach could work, though it'll require a lot of brute-forcing from the AI and loading a lot of broken kernels to see if they work. Plus, if audio drivers are involved, you'd probably blow out the speakers at least once during testing.

Still, if you throw enough money at Claude, I think this approach is feasible to get things booting at the very least. The question then becomes how one would reverse-engineer the slop so human hands can patch things to work well afterwards, which may take as much time as having humans write the code/investigate hardware traces in the first place.

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#296

Post by lelanthran »

slopinthebag wrote: post #93
mixdup wrote: post #80

Earlier quoted context omitted.

I know people have done truly amazing things with AI lately, but I feel this in my bones. Almost every demo I see is like, uh, I don't need these extremely simple things in my life automated. I can just go to Delta and buy a plane ticket. I actually want to write my own email to my mom or wife. Of course a demo is just a demo, but also come on

I think it's a fallacy that if you make creating anything easier, more useful things will be created. In reality, you just end up with more useless things being created. Like with art, when it gets easier to create you don't end up with more good art. And with software - it's not like the quality of software has gone up as it's gotten easier to build, it's gotten much worse. A related fallacy is that great things are…

> I think it's a fallacy that if you make creating anything easier, more useful things will be created. In reality, you just end up with more useless things being created.

Well, if you lower the competence bar required to do something, then more people of lower competence will do that thing.

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#297

Post by jeroenhd »

matthewfcarlson wrote: post #186

I know this is me coming from my spoiled perspective of Linux and macOS, but the advice of running a VM that manages the WiFi hardware and passing it back to the OS seems insane to me

If I were to design an OS from the ground up today, that's probably the approach I'd take. Separate risky drivers off into virtual machines with limited attack surface and a controlled I/O system so that a hacked driver can't infect the rest of the system. Plus, if the code crashes, the kernel doesn't go down with it.

Windows and Linux have been moving drivers towards userland to deal with kernel instability and security risks but with modern virtualisation capabilities, I think going one step further only makes sense. Windows itself is already using something called "virtualisation based security" to leverage VMs to secure the kernel and this is just the logical next step.

Qubes does this stuff too, though it works with fully-featured Linux kernels rather than minimal driver interfaces, for exactly the same reasons. A hacked wifi driver may be able to inspect and redirect traffic, but it can't do much more than that, which combined with a strong VPN can protect from quite complicated attack scenarios that normal people have no other recourse against.

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#298

Post by jeroenhd »

LowLevelKernel wrote: post #92

Omg!!. Similarly, Do you know a way to interface with BIOS so that it can change the parameters?

https://man.freebsd.org/cgi/man.cgi?efivar and underlying APIs should be able to tweak some settings depending on how many variables are exported.

https://unix.stackexchange.com/questions/536436/linux-modify... suggests there may be risks involved using efivar to configure Apple hardware, as there probably isn't any kind of testing or validation present on the variables you set, but if you know what you're doing you should have similar control as you'd have on native macOS I believe.

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#299

Post by josephg »

shakna wrote: post #236
throwaway2037 wrote: post #184

Earlier quoted context omitted.

That headers looks pretty reasonable to me. I don't see anything misleading or ambiguous about it. Whenever I am heavily modifying some licensed code, I always make sure to include a similar header. > I'm going to ahead and say there are copyright law nightmares, right here. I am confused. My first thought was maybe the original Linux driver was GPL'd, but it is not. It is ISC'd. Look here: https://github.com/torvald…

It adds a contributor. To add a contributor, you need "significant" _human_ input. The output of models has so far not been deemed copyrightable. As it acknowledges the original source, it needs to show the human effort that allows it to be bound to the new contributors.

Eh. Copyright only matters if it goes to court. And you only go to court over copyright if somebody is getting sued. That only happens when a plaintiff has standing, they can show damages and the person they want to sue has enough money to make it worth their while. (And if they'll make more money than it costs them in lawyers and negative PR. Suing users and developers for interacting with the product you sold them is generally considered a bad look.)

Anyway, nobody is going to sue you because you added your name (or "project contributors") to an ISC licensed source file in your own repository. Nobody cares. And there's no damages anyway.

Especially when the line added is:

> Copyright (c) brcmfmac-freebsd contributors

If you're right, that's an empty category. Thus the inclusion has no effect.

Re: FreeBSD doesn't have Wi-Fi driver for my old MacBook, so AI built one for me

#300

Post by jeroenhd »

jeremyjh wrote: post #252
einpoklum wrote: post #79

AI didn't write a driver for him. He ported the Linux driver to FreeBSD with some assistance from an LLM. What's more interesting to me is the licensing situation when this is done. Does the use of an LLM complicate it? Or is it just a derivative work which can be published under the ISC license [1] as well? [1] : https://en.wikipedia.org/wiki/ISC_license

It wasn't a straight port, he had an LLM write a spec by reviewing the code, and then in another session another LLM did the development. That is basically a Clean-room approach. It would be unlikely there would be much - if any - code that is exactly the same so showing copyright infringement seems very unlikely.

I'd call it clean room if the AI wasn't trained on the open source drivers in the first place. The open source driver is in there, albeit in the form of lossy text compression with an external dictionary.

Now one side is collecting the necessary tokens to get the AI to output data from the training set in the second run.

Post reply on HN