Copilot works so well because it steals open source code and removes credit


The Software Freedom Conservancy (SFC), a nonprofit community of open source advocates, today announced its removal from GitHub in a scathing blog post urging members and supporters to rebuke the platform a once and for all.

In the front: The SFC issue with GitHub stems from accusations that Microsoft and OpenAI trained an AI system called Copilot on data released under an open source license.

Open-source code isn’t like a donation box where you can just take whatever you want and use it however you see fit.

Greetings, humanoids

Subscribe to our newsletter now for a weekly roundup of our favorite AI stories delivered to your inbox.

It’s more like photography. Just because a photographer doesn’t charge you for using one of their images doesn’t mean you’re still ethically and legally bound to give credit where it’s due.

According a blog post on the SFC site, Copilot does not do this when it comes to using other people’s code snippets:

This is reminiscent of long-standing issues with GitHub and the main reason why we have to abandon GitHub together. We’ve seen with Copilot, with GitHub’s main hosting service, and in almost every area of ​​business, GitHub’s behavior is significantly worse than its peers. We don’t think Amazon, Atlassian, GitLab, or any other for-profit host are perfect players. However, a relative comparison of GitHub’s behavior with that of its peers shows that GitHub’s behavior is much worse.

Background: GitHub is the de facto repository for open source code worldwide. It’s like a combination of YouTube, Twitter and Reddit, but for the programmers and the codes they produce.

Of course, there are other options. But switching from one code repository ecosystem to another is not the same as swapping Instagram for TikTok.

Microsoft acquired GitHub in 2018 for over seven billion dollars.

Since then, Microsoft has leveraged its position as OpenAI’s primary benefactor in a joint venture to create Copilot.

And the only way to access Copilot is through a special invitation from Microsoft or a paid subscription.

The SFC and other open source advocates are upset because Microsoft and OpenAI are essentially monetizing other people’s code and depriving those who use that code of the ability to give proper credit.

In other words: Microsoft takes people’s work, takes credit for it, and sells it to others via algorithms.

A solution: Kill the co-pilot. Alternatively, Microsoft and OpenAI could build a time machine, go back in time, and tag each data point in Copilot’s database so that a second model could be built that would apply appropriate credit to each output.

But it’s always easier to exploit the regulatory environment of the Wild West and take advantage of people than to worry about the ethics of the products and services you offer.

Neural’s point of view: When it comes to solid examples of AI that make human life easier, GitHub’s co-pilot tops the list. It takes some of the tedious things that can take developers hours of work and makes it as easy as pressing a button or typing a few lines of text.

And there is a bit of precedent here. GPT-3 and Dall-E use databases of human-generated media to generate new releases.

But there is an essential difference between these generators and Copilot. Drawing a duck in the style of Monet or having GPT-3 tell you the story of a happy dog ​​is one thing.

Regurgitating line-by-line code snippets from files in a database isn’t coding in someone else’s style, it’s using someone else’s code.

It’s probably a little more nuanced than that. There is, of course, sometimes more than one way to solve a coding problem. And coding is often as much an art as it is a science.

But just because you can take a picture of the setting sun with your iPhone doesn’t mean you can steal someone else’s picture of the sunset, call it your own, and sell it to other people. .

At the end of the day, it doesn’t matter. The copilot is a success. The developer community seems to absolutely love it. He got more positive press than any amount of naysaying is likely to affect.

No matter what it will end up doing to the open source community. Who needs open source repositories when you can just work for free to make money for Microsoft?

The best thing is that you have no choice. There is no opt-in or out. Microsoft and OpenAI have your data and nothing prevents them from doing what they want with it. Resistance is futile.


Comments are closed.