Hi ,
If you wrote an original piece of code after learning how to program from another, does your code IP belong to you or them?
Last week I posed this question to readers and some
of the responses I received are quoted below:
"As a free software advocate I don’t think anyone can own code. If I write a function to calculate Fibonacci numbers, do I own that code? At the granular level code is just logic, can somebody own logic?" - Peter
Prevos
"Unless you're making a cover of their codes - interpreting their very own code, with your own interpretation of that code, to a different context - I'd argue it's yours. For them to build their codes, they needed context and previous experiences, framings, insights. That, most likely, comes from other sources too." - Rod Aparicio
"In general, code shared on Github is meant to be shared, and as long as we give the appropriate credits, etc, we should be fine. As for your writing something entirely new on your own, do our grade school teachers, and the authors of the texts books we
used, have claim over our future creations?" - Paul Rayburn
When you think about it, of course the code ethically or legally belongs to you.
All of us are
where we are today because we had the opportunity to study and learn from others. Yet, the original ideas we produce from that knowledge are still our own.
But what if, instead of a human studying and learning from the works of others, you substitute a computer instead?
This is what a California judge recently had to decide in a class action lawsuit brought against GitHub, Microsoft and OpenAI by a group of developers, and in the judge's opinion, at least, the answer is still the same.
Here's the thing...
Logically, the judge's ruling makes sense. And as a data scientist, the prospect of being able to ethically and legally use open-source data to train models is very appealing.
Yet, every decision
has consequences and as an internet user, I am deeply concerned about what those consequences might be.
The internet is one of the greatest resources of our time and has served to democratise knowledge. Yet, if the IP of those generous people who made the internet what it is today, by freely sharing their knowledge, is used to train models for the financial benefit of big tech, I
wonder whether those people will be willing to continue sharing their knowledge in the future.
The GitHub copyright case is just one of several lawsuits around the ethical and legal implications of training AI models, so a future judge may see things differently.
Whatever the future outcomes, though, for the sake of us all, hopefully, the world will soon find a solution that is fair and beneficial to everyone.
Talk again soon,
Dr Genevieve Hayes.
