这个周末我真的花时间对几款大家一直在吹的新AI工具做了压力测试。 有些在快速起草方面很惊艳，但在一些奇怪的边缘情况下仍然会出现幻觉。我很好奇大家在实际工作中（不是演示）都在用什么。
Original
Spent the weekend actually stress-testing a few new AI tools people keep hyping. Some are impressive for quick drafts, but hallucinations still pop up in weird edge cases. Curious what others are using for real work, not demos.