Salesforce leaders, including SVP Sanjna Parulekar, have admitted that confidence in large language models has dropped sharply over the past year due to issues like failing on complex instructions, AI “drift,” and inconsistent task execution, such as missing customer surveys.
Read more
Share with:
