In 2023, Google attempted to improve its chatbot Bard—later renamed Gemini—by systematically utilizing ChatGPT's output. Documents reveal that contractors from Scale AI compared thousands of ChatGPT responses with Bard's to refine their answers, prompting an internal review of performance. The aim was to create responses that at least matched or surpassed those of ChatGPT. Scale AI's structured approach included data analysis and performance bonuses for superior rewrites. Despite these advancements, questions arose regarding compliance with OpenAI's terms of service concerning competition analysis.
In 2023, Google sought to enhance its chatbot Bard, later named Gemini, by using ChatGPT responses to improve and refine its own outputs.
Scale AI contractors iteratively rewritten Bard's responses to meet or exceed ChatGPT's quality, resulting in a competitive analysis of chatbot performance.
Collection
[
|
...
]