Natural selection’s trial-and-error process allows improvement without anyone understanding or guiding it. The same can apply to how we learn. There are at least three kinds of learning that foster evolution: memory-based learning (storing the information that comes in through one’s conscious mind so that we can recall it later); subconscious learning (the knowledge we take away from our experiences that never enters our conscious minds, though it affects our decision making); and “learning” that occurs without thinking at all, such as the changes in DNA that encode a species’ adaptations. I used to think that memory-based, conscious learning was the most powerful, but I’ve since come to understand that it produces less rapid progress than experimentation and adaptation. To give you an example of how nature improves without thinking, just look at the struggle that mankind (with all its thinking) has experienced in trying to outsmart viruses (which don’t even have brains). Viruses are like brilliant chess opponents. By evolving quickly (combining different genetic material across different strains), they keep the smartest minds in the global health community busy thinking up countermoves to hold them off. Understanding that is especially helpful in an era when computers can run large numbers of simulations replicating the evolutionary process to help us see what works and what doesn’t.
自然选择的试错过程可以在没有任何人理解或指导的情况下进行改进。这同样适用于我们的学习方式。至少有三种学习方式可以促进进化： 基于记忆的学习（存储通过一个人的意识思维获得的信息，以便我们以后可以回忆起来）；潜意识学习（我们从经验中获取的知识永远不会进入我们的意识，尽管它会影响我们的决策）；以及完全无需思考即可发生的“学习”，例如编码物种适应能力的 DNA 变化。我曾经认为基于记忆的有意识的学习是最强大的，但后来我开始明白，它所产生的进步不如实验和适应那么快。为了给你一个例子来说明自然如何在不思考的情况下改进，只要看看人类（及其所有的思维）在试图智胜病毒（甚至没有大脑）时所经历的斗争即可。病毒就像出色的国际象棋对手。通过快速进化（结合不同菌株的不同遗传物质），它们让全球卫生界最聪明的人忙着想出对策来阻止它们。在计算机可以运行大量模拟来复制进化过程的时代，理解这一点尤其有用，可以帮助我们了解哪些有效，哪些无效。

source:https://www.principles.com/principles/bfcc3206-e3e4-4cbf-a419-815964b739cf/