Only if you can reasonably argue that the output is the input (even with exact matches over a certain size being auto-rejected), and that it is enough to qualify as a copyrightable work. I’d argue line completions can never be enough to be copyrightable, and even a short function barely meets the bar unless it is considered creative in some way.
Every open source license grants permission for AI training, and GitHub copilot by default rejects completions that exactly match code from its training. You can’t pretend to be pro-open source or pro-free software but at the same time be upset that people are using licensed software within its license terms.
Nowadays most of them are created by land developers, which often maintain complete control for many years after the first houses are sold in the community. Nothing grassroots about them when they are founded and run by a private company.
Extrapolating the specific use case of pfsense routers to all other applications seems misleading, especially when arm laptops and servers exist that prove their utility and power savings for their use cases. It seems to be the case that there isn’t a particular arm device that is priced well and well suited for pfsense, which does have somewhat unusual requirements compared to other applications.
Usually a boneless wing is a chunk of meat, with identifiable meat fibers and such. Just a breaded and fried chunk of breast. Whereas chicken nuggets are usually made from ground chicken, often molded into a few different shapes.
What do you mean “no side update channels”? There are lots of software that update outside of a distro repo and lots of software that pulls metadata from the internet that could cause an error in the parser.
This is sorta how dark reader and such works. It turns out that implementing dark mode for most websites is more complicated than inverting all the css colors. For example, some gray on white text might have enough contrast to be easily read, but when inverted the text is hard to discern or nearly invisible. Images too, they might have a white background but not look good when inverted. Native support is better
It’s pretty hypocritical that the US gets mad about Chinese overseas bases when the US has bases in every corner of the earth, including an entire ring around China.
As someone who works with and knows several military contractors, I’ve never heard of the US taking ownership of any code written. In fact, most of what they’re paying for is for companies to extend software they’ve already written to better fit the governments use case, such that even if the government owned the new improvements, that code wouldn’t function without the base application that pre-dates a government contract.
For one particular example, the military is currently funding a program to research techniques to use AI for automated vulnerability discovery and exploitation. The info is public here: https://aicyberchallenge.com/
The military is one of the larger stakeholders in AI advancement, and generally doesn’t care about the environment. There will not be a probe that could stifle future advancement in military applications.
Only if you can reasonably argue that the output is the input (even with exact matches over a certain size being auto-rejected), and that it is enough to qualify as a copyrightable work. I’d argue line completions can never be enough to be copyrightable, and even a short function barely meets the bar unless it is considered creative in some way.