Venture Beat just published a new piece of the practice of data scraping, in the wake of OpenAI's 2 lawsuits brought by parties affected by this. Data scraping is what is used to train AI models: large amounts of data are fed into the model to train it; data such as books and other works that fall under copyright laws.(Stay up to date with all AI news and developments at our special News tab in the top left!)
It's a controversial topic. Every time a new wave of innovation in data comes in, questions about privacy arise. Old laws and assumptions are put to the test. For example, Google services work on many different sources of data that have been "taken" from other places; data itself, images, physical places for Maps, etc.The challenge is to enable the advent of new technologies and the movement of data they rely on, while protecting the rights of copyright holders and authors of original work.Is data scraping in AI being used fairly?