Intellectual Property Maximization
I really dislike the modern era of intellectual property maximization where all text or imagery has provenance traced back to the source of where it could possibly have entered someone's mind. I honestly couldn't care less in most cases.
Most Plagiarism Is Fine
[edit | edit source]I really can't find it in me to care about a lot of plagiarism. There is a time where it's bad: one supplies one's Ph.D. thesis or exam or homework results as a certificate of work. There the credential is only valid because the certificate is valid. The point of being anti-plagiarism there is to prove that you deserve the certificate. There, if the majority of one's work is the same as another's then it doesn't function as a valid certificate.
But if you were to find something novel and it were to be built on a bunch of work by other people, I think the least interesting thing in the world is to pore over it to perform some kind of longest-common substring match in order to find what things were 'cited' and what things weren't. It's really just not that interesting to me that some novel discovery did not use quotes and the appropriate literature search.
This kind of formulaic objection to work is quite common, though:
Only 16 citations, half of which aren’t even from this century. There’s only one citation from this decade. They don’t cite any of works on blowup for 3D Euler, like Elgindi, Chen-Hou, Cordoba-Martinez-Zoroa, etc. At best, this is terrible citation practice and at worst it’s blatant plagiarism.
Edit: seems they’ve updated the paper to include citations of Cordoba-Martinez-Zoroa (now has 22 references) which is a far cry from properly citing the ideas the paper is built upon
Because they dont care for proper scientific practices
And even when I agree that people should be pilloried for what is, in effect, theft of public money on fraudulent pretences, I don't actually think that it makes sense to pillory them for this kind of ML-detected plagiarism. Jason Arday was a fraud beyond belief and it's right that he was stripped of his post through resignation, but the objections all focused around allegations of plagiarism - none of which I found particularly compelling. A few strings here and there in common are inevitable when one's been reading some work and it sticks in one's mind. I can't say that all my ideas are truly unique and I don't actually retain a graph that takes me back there.
For all I know, if someone were to take all my blog posts, removed the quoted sections, and scanned it for plagiarism they'd find it. I'd hoped for a sufficiently idiosyncratic writing style that I'd be noticeable to AIs, but given that I'm not, I'm likely just Yet Another Software Engineer With A Blog and inevitably all my ideas will be either passé or from other sources - none of which I can recall!
All Things Are Owned
[edit | edit source]This same prudishness around idea ownership has also spread wickedly through other fields. Now, supposedly we are to believe that UIs are automatically copyrighted productions of the people who owned them. Make a two level menu that reaches a page with a red outline and you're "copying" everyone else who did that.
One truly does wonder what would come of Linux distributions that attempted to mimic MacOS or Windows these days. Back in the day, those were common, but also common were games like SuperTuxKart which were just open-source clones of commercial games. These days, the zeitgeist has shifted, even on so-called 'hacker' forums:
We are all familiar with the Corgi event: https://x.com/mfts0/status/2070080422482977095
With the barrier to create new apps having dropped significantly thanks to LLMs, I am seeing more cases about copyright and unfair competition.
I've seen and participated in some of these cases. Usually expert witnesses are required.
Curious to hear the community stance on this one.
"Now software developers are feeling what authors and artist felt". https://x.com/PriyRanjan96/status/2070204156703568377
There are several claims of: Copying UI is Ok, your product is not undifferentiated enough.
Here is a legal assessment of the situation: https://x.com/jessebradner/status/2070492879718350986
I honestly could not give less of a damn that two pieces of software involve a 'danger zone' that someone can delete something in or whatever. Github, as far as I know, popularized this lingo and the fact that two document holding companies have some information pertaining to document holding using that terminology is about the least-interesting thing of all time.
Generative AI
[edit | edit source]The other unusual thing is watching a large number of people bemoan the advent of generative AI. With modern LLMs and Diffusion Models we all get the power of creation in our hands. But communities that would once eagerly embrace the idea of software or media piracy as Information wants to be free suddenly find themselves eager to defend the idea of intellectual property and who owns what and so on.
Conclusion
[edit | edit source]I think all of this is nonsense. Obviously, where the significance is a claim of work being done then that's a lie. But if you're just saying "We now know that X is true" and you have a proof then I couldn't care less about who wrote the techniques prior. This is similar to that nonsense about how science done on participants unethically is not science. This is wholly not true. You could have eaten an orphan's heart between each test of MMR-2 and that wouldn't have changed that it worked.