Translated data: OpenAI began developing a new AI model called Arrakis last fall, aiming to improve operational efficiency and reduce training costs, but the test results fell short of expectations. OpenAI decided to halt the project after a month of testing. The failure of Arrakis has led OpenAI to shift focus to other model developments, and it may also result in commercial losses for Microsoft. The incident highlights the uncertainties in the field of artificial intelligence, and sparse models could be the trend for future cost reduction.