GitHub Support just straight up confirmed in an email that yes, they used all public GitHub code, for Codex/Copilot regardless of license - GitHub does another retarded initiative

一个日志 said:
as with all the other explanations you got, I just like to add on. Programmers are fundamentally lazy, present company included. We want to find the best way to make sure we don't have to do something again.

That way when the alphabet homos start using GitHub pilot, I don't think they would actually double check their code. Which the AI writes from them, and given previous history of AI manipulation. (rip in peace tay tay) it would be very easy to sneak something like that code in.
View attachment 2332330
edit: adding tay tay video
Click to expand...

That image hurts my souls, this is directly painful to me. That is me for the last 2-3 months, but man my code can do some SHIT!!!, but it is overengineered as fuck to handle all kinds of stupid shit.

Now to be more on topic
Gar For Archer said:
This is pretty much my impression of ML/Neural nets. It excels at a small subset of tasks, specifically those which involve pattern recognition of a large volume of “messy” data (e.g. raw text or images) which would be exceedingly difficult or time-consuming to analyze using traditional methods. However, retards keep trying to jam it where it doesn’t belong because they’re part of the “I fucking love science!!!” troon brigade that thinks science is literal magic and that importing libraries is coding.
Click to expand...

When a thing has a very simple "Yes" or "No" answer then Neural networks are great as soon as you even have a bit of fuzziness on the answer then it becomes a problem.
Is this a CAT "YES/NO" update according to answer. Easy simple and in the long run it will work fine.

Code however is not always perfect.
During my education i had to write code in C++ for a version of Knights tour or travels or something and it needed to be scaled so it could reach a 1000*1000 chess board or something big like that.
First Version of the code works fine, but takes 4 hours to complete.
Second Version i cut the time down around 2 hours. (Why the fuck is std::pair a thing and shit)
Third version took about a min if not less (I figured out how to use pointers and queue)

All 3 codes work and that is the problem. The machine can not tell if something is good code only that a lot of people did something so it copies.
 
  • Like
Reactions: Pissmaster and Wright
ChromaticAberration said:
The Copilot Machine Learning model was trained using code from every repository available on GitHub. It has been shown that Copilot is able to reproduce code verbatim from its learning pool. That means Copilot can potentially be used to "launder" code with the goal of removing the Copyleft licenses (which make the software free and open) of the original code and use it for closed proprietary software. This is a massive legal shitshow brewing.
Click to expand...
Reverse-engineering machine learning bullshit is my favorite part of clown world. Same thing has happened recently with Apple's CP detection.
 
  • Like
Reactions: Moths
RandomTwitterGuy said:
That image hurts my souls, this is directly painful to me. That is me for the last 2-3 months, but man my code can do some SHIT!!!, but it is overengineered as fuck to handle all kinds of stupid shit.

Now to be more on topic


When a thing has a very simple "Yes" or "No" answer then Neural networks are great as soon as you even have a bit of fuzziness on the answer then it becomes a problem.
Is this a CAT "YES/NO" update according to answer. Easy simple and in the long run it will work fine.

Code however is not always perfect.
During my education i had to write code in C++ for a version of Knights tour or travels or something and it needed to be scaled so it could reach a 1000*1000 chess board or something big like that.
First Version of the code works fine, but takes 4 hours to complete.
Second Version i cut the time down around 2 hours. (Why the fuck is std::pair a thing and shit)
Third version took about a min if not less (I figured out how to use pointers and queue)

All 3 codes work and that is the problem. The machine can not tell if something is good code only that a lot of people did something so it copies.
Click to expand...
I actually think it is the other way around. Neural nets exceed at getting you mostly correct answers to very hard problems for classical techniques like image recognition and language translation. CV used to be lots of finely tuned hardcoded recognition patterns, only to be outperformed by more complex (in terms of function space) deep nets. However, if you want exact answers, that is the domain of classical AI (symbolic computation, BDDs, exact algorithms). Both are important.
 
Knight of the Rope said:
I'm fully blackpilled on the 'AI revolution' at this point and to be honest I'm all-in on another AI winter coming soon. I think people are starting to wake up from the hype.
Click to expand...
I can't fucking wait for the next winter. I so fucking sick of the second coming of the digital messiah.

Another thing to consider is since AI is just statistics it's still susceptible to all the ways that statistics can be misleading, only now at industrial scale.
Doctors berry their mistakes, carpenters sand away their mistakes, and copiers make 1000 copies of their mistakes.
Click to expand...
A while back there was this AI project that said that if you're not smiling your more likely to be a criminal; turns out when people are getting their mug shot taken they're not smiling. Of course it's sold as a way to detect criminals, so if you're frowning you're a criminal. We are at Stage Six in the United States; as technology is more accessible to the unintelligent, our future is more precariously situated to be abused by the useful idiots of the rulers.

Spasticus Autisticus said:
copilot4.png
Click to expand...
That's oxymoronic. Trannies are demons sent by Xenu to fight for the Italians for the side of evil. So trannies are not human and they are not granted any protection by the US Bill of Rights.
 
  • Like
Reactions: awoo and Knight of the Rope
  • Informative
  • Autistic
  • Semper Fidelis
Reactions: Knight of the Rope, walnut, ⠠⠠⠅⠑⠋⠋⠁⠇⠎ ⠠⠠⠊⠎ ⠠⠠⠁ ⠠⠠⠋⠁⠛ and 1 other person
stares at error messages said:
A while back there was this AI project that said that if you're not smiling your more likely to be a criminal; turns out when people are getting their mug shot taken they're not smiling. Of course it's sold as a way to detect criminals, so if you're frowning you're a criminal. We are at Stage Six in the United States; as technology is more accessible to the unintelligent, our future is more precariously situated to be abused by the useful idiots of the rulers.
Click to expand...
Imagine millions of people getting genocided because they have "the rapist forehead". LMAO
 
  • Like
Reactions: 419 and Wright
Places that offer some service for free online using the data you upload to them without consent or recourse? What's the world coming to and who would've thought?!
 
You must log in or register to reply here.
Back