Comment by tanto

I agree that proprietary data will become more valuable. It is, even today, mostly not accessible for AI training and holds so much value. We are working on Flower (https://flower.dev), which enables training AI on private data without the data owner having to share it.