Part 3/9:
OpenAI's terms of service clearly state that users are prohibited from copying any of its services or outputs for the purpose of creating competing models. However, while OpenAI firmly maintains this stance, the larger conversation surrounding the usage of web data raises issues of irony and fairness that cannot be overlooked.
The Double Standards of Data Use
Sasha, the speaker in the presented transcript, articulated her indignation regarding OpenAI's position, pointing out the paradox in their claims. In her view, the same terms and conditions that OpenAI enforces were disregarded when it built its own language models. By scraping massive amounts of data from various websites—everybody else's intellectual property—OpenAI has built its business on the very practices it now condemns.