Machine learning (ML) has a long legacy of proprietary technology. As ML becomes more popular with industry-scale applications, new proprietary pushes have been entering parts of the ML application stack. We need to prepare for the challenges of keeping free technology, which come from the legacy side and the new proprietary push. Although many free software alternatives have made ways into ML research, development, and industry applications, the topic is frequently overlooked. In this session, we will look at the history of ML stack in research and development; explore ML's proprietary legacy and its status quo; look at where the proprietary army is coming (again) while we were not looking; and strategies about what we can do to keep ML research and development free. Your thoughts, questions, and further discussions are also cordially welcome!
Speakers: Helen Jiang