Given the current environment of amplified claims about being open source, wanted to add couple of points to the already great answers:
– Open source isn’t just the open code, it’s also anything else you need to reproduce the results that the developer’s claims it does. For example, if you take AI models for you to reproduce the results you need the code, configuration and model weights. Even if you have access to all of the same input data, without all 3 you are very unlikely to be able to reproduce the results. That’s not open source even if the code itself is available without the configuration and weights.
– There is plenty of other “open XXXXX” labels bandied about that sounds similar, like open access. Open anything else other open source is not open source and sometimes the label might read open source but it’s not – see above.
Latest Answers