General

The OpenAI - DeepSeek S**t-Storm

Posted on Fri, Jan 31, 2025

The OpenAI - DeepSeek S**t-Storm

As you may have heard, the new AI model DeepSeek R1 has been making waves with its impressive performance and free pricing model. However, a contentious issue surrounds its development.

The Legal Battle

According to OpenAI, the DeepSeek model was trained using their own AI chatbot, ChatGPT. While this might seem like a legitimate use of AI technology, OpenAI claims that ChatGPT itself was trained on a vast amount of data without permission from its original creators.

In response to this criticism, some have argued that "anything accessible on the open web is fair game." Moreover, Satya Nadella, Microsoft's CEO, echoed this sentiment, stating that "if information is available on your website, then it's in the public domain." However, I would argue that if this data was accessed illegally, it does not rightfully belong to those who used it.

On my own website’s terms and conditions page, I have a clause stating that all content belongs to Cloud Wales and cannot be used without authorization. Interestingly, ChatGPT appears to have a similar clause, yet OpenAI now claims that training AI on their data without permission is illegal.

To provide more insight into this issue, I’ve included a video from Sasha Yanshin Oh No! China Stole Data From OpenAI! that offers further context and a deeper explanation of the controversy.

What are your thoughts on this issue? Share your opinions in the comments!