Did DeepSeek do reverse engineering and learn from other AI models like ChatGPT and Anthropic? Here's what techies are saying

DeepSeek AI has taken the AI world by storm and has also ended up giving AI stocks in Wall Street a major jolt, with the likes of Nvidia, OpenAi, Meta, and others taking a solid beating during trading hours. DeepSeek R1, the AI model launched by the China-based startup of the same name began rising through the ranks to become one of the primary competitors of ChatGPT and Gemini within a few days of launch.

Did DeepSeek reverse engineer ChatGPT's technology?

There are now strong questions about whether DeepSeek's AI model used a reverse-engineering method to learn and train from OpenAI's ChatGPT or Anthropic, which are very powerful AI models themselves. A Fortune report suggests that it is possible looking at some indications, DeepSeek may have used ChatGPT's output to initiate the AI model's training abilities.

DeepSeek putting US-based AI companies at risk

Thereafter, they may have run their own course and trained the AI model in their own manner. Using and learning from another model's output, and conducting further development from there is something that is referred to as reverse engineering. This is somehow questionable, but since the deal is now done and DeepSeek R1 is now in the market, there are further fears that this event could put OpenAI, Anthropic, and other AI companies' business models at risk.


Moreover, DeepSeek R1 is now open-sourced, which is opposite to ChatGPT, which is as closed as possible. So, if any other company were to follow and develop along DeepSeek's lines, that could pose a risk for AI companies in the US.

FAQs:

Is DeepSeek R1 free?
Yes, DeepSeek R1, an AI model developed by China-based startup DeepSeek AI, is free to use and open-source.

Is DeepSeek R1 an AI model?
Yes, DeepSeek R1 is a very powerful AI model, developed by China-based company, DeepSeek AI.