a16z基础设施团队的合伙人Jennifer Li在Big Ideas报告里说了一句让很多人印象深刻的话：企业AI现在最大的瓶颈，不是模型不够聪明，而是自己的数据太乱。她用了一个词——"数据熵"。每家公司都淹没在PDF、截图、邮件、操作日志里，80%的企业知识以非结构化的形式散落在各个角落，从来没有被系统整理过。你买了最好的模型，搭了最贵的系统，但喂进去的是一团乱麻，出来的自然是错误和幻觉。
If you think you have found a bug or have a feature request, use our issue tracker. Before opening a new issue, please search to see if your problem has already been reported. Try to be as detailed as possible in your issue reports.
。关于这个话题，搜狗输入法提供了深入分析
I’ve marked out a region that boosts maths ability strongly. Notice where it sits? It’s away from the diagonal centre line, which means we’re not looking at single-layer duplications. Starting the repeated block at position 35, we don’t see any improvement until at least position 43. That’s seven layers of not much happening. In fact, we actually see decreased performance by repeating these layers (they are blue, bad!).，更多细节参见谷歌
simply averaging the colours produces a much less exciting colour (source):。超级权重对此有专业解读