Slash Boxes

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 14 submissions in the queue.
posted by martyb on Friday July 09 2021, @12:52AM   Printer-friendly
from the we-violate-all-open-source-licenses-equally dept.

GitHub’s automatic coding tool rests on untested legal ground:

The Copilot tool has been trained on mountains of publicly available code

[...] When GitHub announced Copilot on June 29, the company said that the algorithm had been trained on publicly available code posted to GitHub. Nat Friedman, GitHub’s CEO, has written on forums like Hacker News and Twitter that the company is legally in the clear. “Training machine learning models on publicly available data is considered fair use across the machine learning community,” the Copilot page says.

But the legal question isn’t as settled as Friedman makes it sound — and the confusion reaches far beyond just GitHub. Artificial intelligence algorithms only function due to massive amounts of data they analyze, and much of that data comes from the open internet. An easy example would be ImageNet, perhaps the most influential AI training dataset, which is entirely made up of publicly available images that ImageNet creators do not own. If a court were to say that using this easily accessible data isn’t legal, it could make training AI systems vastly more expensive and less transparent.

Despite GitHub’s assertion, there is no direct legal precedent in the US that upholds publicly available training data as fair use, according to Mark Lemley and Bryan Casey of Stanford Law School, who published a paper last year about AI datasets and fair use in the Texas Law Review.

[...] And there are past cases to support that opinion, they say. They consider the Google Books case, in which Google downloaded and indexed more than 20 million books to create a literary search database, to be similar to training an algorithm. The Supreme Court upheld Google’s fair use claim, on the grounds that the new tool was transformative of the original work and broadly beneficial to readers and authors.

Microsoft’s GitHub Copilot Met with Backlash from Open Source Copyright Advocates:

GitHub Copilot system runs on a new AI platform developed by OpenAI known as Codex. Copilot is designed to help programmers across a wide range of languages. That includes popular scripts like JavaScript, Ruby, Go, Python, and TypeScript, but also many more languages.

“GitHub Copilot understands significantly more context than most code assistants. So, whether it’s in a docstring, comment, function name, or the code itself, GitHub Copilot uses the context you’ve provided and synthesizes code to match. Together with OpenAI, we’re designing GitHub Copilot to get smarter at producing safe and effective code as developers use it.”

One of the main criticisms regarding Copilot is it goes against the ethos of open source because it is a paid service. However, Microsoft would arguably justify this by saying the resources needed to train the AI are costly. Still, the training is problematic for some people because they argue Copilot is using snippets of code to train and then charging users.

Is it fair use to auto-suggest snippets of code that are under an open source copyright license? Does that potentially bring your code under that license by using Copilot?

One glorious day code will write itself without developers developers.

See Also:
CoPilot on GitHub
Twitter: GitHub Support just straight up confirmed in an email that yes, they used all public GitHub code, for Codex/Copilot regardless of license.
Hacker News: GitHub confirmed using all public code for training copilot regardless license
OpenAI warns AI behind GitHub’s Copilot may be susceptible to bias

Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 2) by DannyB on Friday July 09 2021, @03:33PM (3 children)

    by DannyB (5839) on Friday July 09 2021, @03:33PM (#1154331) Journal

    A good IDE lets you create your own templates. (eg, boilerplate) If you have some construction that you frequently type, you can make it a template, complete with variables. It works the same. A keystroke generates the template at the point where you are typing. You can tab through the variables to name them differently as you wish, but renaming one variable renames it everywhere within that template -- until you start filling in the body, if it has a body.

    Some people don't like the noise and complexity of modern IDEs and prefer a simple text editor.

    Some people don't like the noise and complexity of a backhoe and prefer to dig a ditch using a shovel. And much more worser is that a backhow requires a bit of learning to use. Best to stick to the shovel.

    Every day I think maybe dividing by zero will work this time.
    Starting Score:    1  point
    Karma-Bonus Modifier   +1  

    Total Score:   2  
  • (Score: 2) by hendrikboom on Friday July 09 2021, @05:37PM (1 child)

    by hendrikboom (1125) on Friday July 09 2021, @05:37PM (#1154388) Homepage Journal

    And sometime the language has features that allow the compiler to expand the boilerplate instead of the editor. Yes, complete with proper handling of bound variables.

    • (Score: 2) by DannyB on Friday July 09 2021, @08:43PM

      by DannyB (5839) on Friday July 09 2021, @08:43PM (#1154437) Journal

      The compiler and language may already have ways of hiding various scopes of variables. The problem is that if I want to change the name of variables A and B to be named X and Y, I don't want to have to go change every instance of them by hand. How would a compiler let you do that to your source code?

      In an IDE, I can click on a variable, ctrl-shift-R, then rename that variable, and the IDE precisely and exactly changes all occurrences of that variable and not any other identifiers that happen to have the same names but in other scopes or contexts. (like both a function, a variable, a class and a type all named A.) And if that variable is visible in other parts of the project, other files, it changes them there too! It is not some dumb search/replace. It is based on the compiler's understanding of the scope and visibility of that identifier throughout the entire project. The compiler and editor are deeply integrated.

      I happen to use Eclipse. No matter what language I'm editing, the editor is integrated with the proper compiler or language server.

      Every day I think maybe dividing by zero will work this time.
  • (Score: 2) by HiThere on Friday July 09 2021, @06:04PM

    by HiThere (866) on Friday July 09 2021, @06:04PM (#1154398) Journal

    If the IDE/editor ends up writing verbose text via a boiler-plate template, it's nearly as hard to read a month later as if you had written it by hand. And if you need to customize the variables...well, I've been known to miss some of those, or to do a replace that wasn't limited to the appropriate areas of text. Usually that causes an immediate error, but sometimes it's quite difficult to track down.

    This I feel to be the appropriate use-case for templated/generic classes. But, of course, those can't handle all the cases that a template substitution can. OTOH, custom macros are REALLY dangerous. Used appropriately, they're very useful. Over used, or used inappropriately, and they render the program text nearly unreadable.

    Back in the day (Fortran IV days) I once was really attracted to macro templated code. (Look at Mortran [] or DYSTAL [] though DYSTAL was really more of a library) But they rendered the code unreadable by anyone else, and after awhile unreadable by me.

    Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.