DeepSeek, a Chinese AI company, is under scrutiny from OpenAI and Microsoft for allegedly building competitive AI models using data sourced from OpenAI's API. Concerns emerged from security findings indicating abnormal data exfiltration linked with DeepSeek accounts. OpenAI suspects that DeepSeek utilized a distillation technique, infringing on OpenAI's terms of service, to train its AI cheaply compared to the expensive training of its own models. This controversy highlights ongoing tensions related to intellectual property theft in AI development.
We know PRC (China) based companies - and others - are constantly trying to distill the models of leading US AI companies. This shows persistent efforts to exploit intellectual property.
It's an efficient way to train smaller models at a fraction of the more than $100 million that OpenAI spent to train GPT-4, expressing concerns about competition through potentially illegal means.
Collection
[
|
...
]