大家觉得Data Scientist在未来10-20年还有前途吗？

匿名用户 发表于 2024-05-10 06:35:31
Anything that calls itself a science, isn't one.

Any short cool sentence that summarized a phenomenon involves human behavior, is almost surely BS.      —— a wiser man than cenzhu

Chat GPT:
That statement is pretty sweeping and doesn't hold up under scrutiny. Science is defined by its methodology: observation, hypothesis formation, testing, and peer review. It's a rigorous process that's led to countless advancements. Dismissing everything that calls itself a science would mean rejecting fields like physics, chemistry, biology, and many others, which have all demonstrated their validity through experimentation and evidence. While there are fringe or pseudoscientific claims out there, it's not fair to discredit the entire field of science because of them.

匿名用户 发表于 2024-05-11 10:54:13
DS这块牌子已经烂了。即使没有AI，也只会向着DA甚至BA的方向去。

本来就是DA，还不是FB当年创造的title

没有，我们公司da和ds都被撤了，干的活被mle和sde代替。。有时候我也不能理解我一个sde要去干部分ds的活，但确实ds可有可无的样子。

Data science 在很多公司和部门确实不够science，体现（1）从业人员行事和思维方式并不依从严谨科学研究规范（2）目标聚焦单一企业日常业务运营，而非专注普遍适用性理论与方法的创新。

先定义啥是ds

匿名用户 发表于 2024-05-11 10:54:13
DS这块牌子已经烂了。即使没有AI，也只会向着DA甚至BA的方向去。

以前就是叫da或者bi，这些年带了个高帽子

感觉DA和BA会合并

匿名用户 发表于 2024-5-12 10:27
Any short cool sentence that summarized a phenomenon involves human behavior, is almost surely BS.  ...

https://duncan.hull.name/2011/07/01/but-is-it-science/

"probably"

匿名用户 发表于 2024-5-12 07:56
以前就是叫da或者bi，这些年带了个高帽子

错误的。以前的DS是现在modeling MLE和AS的混合体。

说白了还是看个人能力，重复繁琐的劳动只会被AI取代，大牛会利用AI提高效率