$MYSTMYST ราคา

According to Beating AI news, OpenAI has finally identified the specific reasons for the rapid depletion of Codex quotas. Tibo Sottiaux, the responsible person, stated that the team has discovered three areas of excess consumption: inefficiency in compressing context during lengthy conversations with images; excessive usage in certain high-consumption scenarios related to Computer History; and the automatic generation of conversation titles consuming more quota than expected. Previously, Tibo mentioned that no overall anomalies were found and directed some affected users to sub2api and subscription sharing. He further acknowledged yesterday that the cache hit rate for some users had indeed worsened, potentially leading to accelerated quota consumption. The investigation has now pinpointed several specific issues within Codex itself. Among them, 'context compression' is designed to allow Codex to continuously compress old content during lengthy tasks to free up context for ongoing work. OpenAI's official documentation also confirms that Codex automatically uses compaction to maintain long-duration tasks. When there are many images and multiple compressions occur, this process currently incurs additional waste. Computer History is a recently launched feature that allows users to bring recorded actions from selected apps and web pages on Mac into ChatGPT and Codex. OpenAI will push a fix on Sunday, U.S. time, along with a complete reset of Codex usage for all paid subscribers. Tibo mentioned that the reset is expected to go live around 2 PM Pacific Time, corresponding to around 5 AM Beijing time on August 24. The team has also found an unrelated new optimization plan that is said to significantly improve efficiency and will continue to be advanced next week.

According to Beating AI news, OpenAI has finally identified the specific reasons for the rapid depletion of Codex quotas. Tibo Sottiaux, the person in charge, stated that the team has discovered three areas of excess consumption: the inefficiency of compressing context repeatedly during long conversations with images; excessive usage in certain high-consumption scenarios with Computer History; and the automatic generation of conversation titles, which has consumed more quota than expected. Previously, Tibo claimed that no overall anomalies had been found and directed some affected users to sub2api and subscription sharing. He further admitted yesterday that the cache hit rate for some users had indeed worsened, potentially leading to accelerated quota consumption. The investigation has now pinpointed several specific issues within Codex itself. The 'context compression' process is designed to allow Codex to continuously compress old content during long tasks to free up context for ongoing work. Official OpenAI documentation also confirms that Codex automatically uses compaction to maintain long-duration tasks. However, when there are many images and multiple compressions, this process currently incurs additional waste. Computer History is a recently launched feature that allows users to bring recorded actions from selected apps and web pages on their Mac into ChatGPT and Codex. OpenAI plans to push a fix on Sunday, U.S. time, along with a complete reset of Codex usage for all paid subscribers. Tibo indicated that the reset is expected to go live around 2 PM Pacific Time, corresponding to approximately 5 AM Beijing time on August 24. The team has also identified an unrelated new optimization plan that is said to significantly improve efficiency and will continue to be pursued next week.

On August 23, CertiK reported that Term Labs experienced a governance attack, resulting in asset losses of approximately $8.5 million. The attacker's address currently holds 2,843 ETH (approximately $7.1 million) and about 1.6 million DAI. Term Labs stated that it has identified a governance vulnerability affecting Term Vaults and is currently conducting further investigations.