OpenAI Secretly Uses Google Data To Train ChatGPT
They say it takes a village to raise an AI, and in the case of ChatGPT, that village apparently includes its biggest competitor.
A Rival's Helping Hand
Many in the tech world view OpenAI’s ChatGPT as a significant, perhaps even existential, threat to Google’s foundational Search business. As users increasingly turn to chatbots for answers, the dominance of traditional search engines is being challenged. However, it appears that OpenAI has been leveraging its rival's resources to fuel its competition.
The Methods Behind the Model
According to a report from The Information, OpenAI has been using Google Search results to power ChatGPT's responses for timely topics like news, sports, and stock market data. While Google’s terms of service restrict direct access, OpenAI reportedly utilized a third-party firm, SerpApi, which specializes in scraping this public data from the web.
This isn't the first instance of OpenAI using Google-owned data without explicit permission. The company has also previously used data from YouTube to train its sophisticated AI models, highlighting the aggressive data acquisition strategies employed in the race for AI supremacy.
