Continuation of `Having "blessed" 3rd party libraries may make D more popular` DIP thread

[WraithGlade]

On Friday, 4 July 2025 at 22:29:13 UTC, monkyyy wrote:
> On Friday, 4 July 2025 at 21:23:07 UTC, WraithGlade wrote:
>> Tangentially, I seldom publish anything on GitHub because
>
> Publish everything, it doesnt matter no one actually looks

On the contrary, people *do* look at other people's source code 
on GitHub, otherwise there would be no point in the site besides 
distribution and there are better ways to do that.

Besides people though, there is also the problem of automated 
plagiarism via "AI" bots, which is motivated by large companies 
wanting the ability to steal the creative output and digital 
labor of all of the rest of humanity and the whole internet in a 
way that allows them to essentially launder the copyright of it 
all (including both open source and proprietary code). 
Microsoft's ownership of GitHub gives them more of a 
superficially/ostensibly legal excuse to train based on the data, 
even though there is no *ethical* grounds for doing so. In fact, 
I've long suspected that that was actually one of the biggest 
reasons they bought the site, if not the #1 reason in reality.

In fact, the risk could be spread even more broadly than that: 
Any software that is under the control of any such tech company 
could conceivably send any code processed by it (e.g. by a 
compiler) to the company's database to feed their AI-powered 
thievery. In fact, there have been trends of big tech companies 
in recent years doing pretty much anything to get access to more 
data to surreptitiously train their automated plagiarism systems 
(LLM AI) on. That's why I won't ever fully trust anything made by 
these companies, not even programming languages like C# 
(Microsoft) or Go (Google). I even recall reading about Google 
inserting code that tries to phone home into all executables 
compiled by Go, though that was not related to AI if my memory 
was correct. It's possible I may use C# or Go if it is the only 
practical solution for something, but it is far from ideal from 
the standpoint of someone who has learned these companies 
increasingly cannot be trusted whatsoever.

Similarly, pretty much all cloud data storage service providers 
have shown signs in recent years of changing their terms in ways 
that grant them permission to train AI based on any and all of 
your data (e.g. valuable idea documents, art assets, code, etc) 
you have on their servers. Thus, if you write a text memo to 
yourself that says "million dollar software idea" and store it on 
Google Drive or Dropbox and someone subsequently asks ChatGPT or 
similar for a million dollar idea then your idea may end up in 
the results given to the user, but in a highly rearranged and 
transformed way designed to make detecting and proving the 
plagiarism difficult or impossible.

Nonetheless, thank you for the attempted reassurance and the 
good-natured and wholesome spirit in which it was given, but I'm 
not naive enough to trust these companies nor random people 
browsing code on the internet.

Ideas are often very valuable, contrary to misguided but trending 
beliefs in recent years that "there are so many ideas that there 
is no danger in sharing them" and such. Human nature is what it 
is. Wholesome people are abundant, but so are unethical people.

I only trust evidence and logic. Those are the only true 
authorities on anything in this world that can ever exist. 
Popular assumptions and trends aren't anywhere close to 
trustworthy.

We good-natured people must learn to protect ourselves better and 
to not be naive. I wish that your reassurances were sound, but 
they aren't. Such a world would be nice though! I appreciate the 
sentiment!

Putting all that aside: I hope everyone had a wonderful 4th of 
July yesterday!