Stories
Slash Boxes
Comments

SoylentNews is people

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: 1) by shrewdsheep on Friday July 09 2021, @03:54PM (1 child)

    by shrewdsheep (5215) on Friday July 09 2021, @03:54PM (#1154342)

    To some degree, it is a matter of opinion, so nothing really to disagree about.

    As the topic seems to be somewhat sensitive, a definition would be appropriate as a starting point: if the information expressed in N lines of code can be given in N/K of code, I call the code boilerplate. For me, K is somewhere between 5 and 10. With this definition, your for loop would not count for me. My programming style focuses very much around the do not repeat yourself principle and small units of code. For me, a function containing 20 lines of code is long (well in high-level languages like R/python/perl, I do not manage in C++). The examples given for the Copilot, I would have factored out into smaller functions in most cases.

    The Copilot goes the wrong way around, IMO. Instead of suggesting boilerplate code, the boilerplate code should be avoided altogether. One very concrete example is packaging where many tools exist to create skeletons (be it R/python/perl/whatever) for you. This is the wrong way round. The information needed is just the code (being inline documented) and a single dictionary containing required meta-information. Basically adding 10-20 lines of description to an existing code directory should allow you to create a package. From this the entire packaging can proceed. The actual package is always temporary code.

  • (Score: 2) by DannyB on Friday July 09 2021, @08:50PM

    by DannyB (5839) Subscriber Badge on Friday July 09 2021, @08:50PM (#1154441) Journal

    I think everyone is in favor of making things simple and less verbose. As simple as possible, but not any simpler.

    Now how simple it should be depends on how big your projects are. Java is used for very large source code bases. Many diverse teams may write many different modules or libraries that end up in a single executable.

    I think most languages could benefit from some form of IDE assistance to help you type out repetitive templates of code. I used the for() loop for an example, because the typical pattern of a for loop is to have a single variable that is referenced three times. You should only have to type in that variable name once, not three times. When you change the first occurrence of the variable, it should change the others, keystroke by keystroke.

    The DRY principle is something I strongly embrace. But in a for() loop, you typically repeat the variable name three times. Or more times if you reference that variable within the body of the loop and not just the initialization, increment, and test of the loop construct. It sure is nice if I can change that variable name one time and have it change everywhere it is used within that loop construct.

    I'm not against something like CoPilot, if it works well. But the copyright an license issues are a genuine concern.

    --
    People today are educated enough to repeat what they are taught but not to question what they are taught.