今天读了一篇来自于纽约时报的文章,讨论AI的局限性,起码可以暂时有助于缓解当前人们的AI焦虑 (说真的,如今人们对AI的追捧简直一窝蜂,太疯了!)
以下是我对文章的核心内容做的简单摘录,偶尔加了几句我自己的碎碎念。
After decades of dismissing liberal arts and humanities studies as useless and insisting that the mastery of science, engineering, math and tech is essential to future success, the tech world is coming around to the idea that learning about human nature could be a valuable asset in the coming AI revolution.在数十年间将文科和人文学科视为无用之物,并坚称掌握科学、工程、数学和技术是未来成功的关键之后,科技界如今开始意识到,了解人性或许会在即将到来的人工智能革命中成为一项宝贵的资产。
长久以来我一直都不太清楚所谓的“liberal arts”和“humanities studies”到底指的是什么,包括了哪些具体科目,这次就彻底查清楚:
1)the Liberal Arts are a broad field of study designed to develop general intellectual capabilities, critical thinking, and a well-rounded perspective.The goal is not just to learn what to think, but how to think. 博雅教育是一门广泛的学科领域,旨在培养通识能力、批判性思维以及全面的视野。其目标不仅在于学习“思考什么”,更在于学习“如何思考”。它主要包含三大类别:
1) Humanities,which explore the human condition, culture, and expression, include a number of specific disciplines including Philosophy, History, Literature, Languages, Art History, Musicology, Ethics, etc. 人文学科致力于探索人类的生存状态、文化及表达方式,涵盖多个具体学科,包括哲学、历史、文学、语言学、艺术史、音乐学、伦理学等。
2) Social Sciences, which study human society and social relationships using scientific methods, contains the following disciplines: Psychology, Sociology, Anthropology,Economics, Political Science, and Geography. 社会科学运用科学方法研究人类社会及社会关系,包含以下学科:心理学、社会学、人类学、经济学、政治学和地理学。
3) Natural Sciences, which aim to understand the physical and natural world through observation and experiment, involves Biology, Chemistry, Physics, Astronomy,and Earth Science.自然科学(or 理科) 旨在通过观察和实验来理解物理和自然世界,涵盖生物学、化学、物理学、天文学和地球科学。
所以,“liberal arts”严格来说不能简单翻译成“文科”,而是范围更广的“博雅教育”,而中文里的“文科”和“理科”都归属于“the Liberal Arts”这个大范围里。
继续读文章:
As it turns out, tech jobs may be drying up after years of students rushing to computer science. Who needs to code? A.I. does that for you. What A.I. can’t do(yet) is the stuff that makes us human: empathy, emotion, psychology, critical thinking. In a way, A.I. is a false mirror. It reflects back answers to black-or-white questions, but it does little to help explain the human experience the way art or philosophy can.事实证明,在学生们争相投身计算机科学领域多年后,科技类岗位可能正在逐渐减少。谁还需要编程?人工智能可以代劳。但人工智能(目前)还做不到的是那些让我们成为人类的特质:同理心、情感、心理活动以及批判性思维。从某种意义上说,人工智能是一面虚假的镜子。它能对非黑即白的问题给出答案,却无法像艺术或哲学那样,帮助我们阐释人类的体验。
"the black-or-white questions" 这个说法即简洁又精准,值得记一下
Daniela Amodei, a founder of Anthropic, told ABC News that “the things that make us human will become much more important instead of much less important.” She said that at Anthropic is looking to hire people who are “compassionate and curious”about other people.
最近Claude code火的不得了,貌似就是由Anthropic这家公司研发出来的。特意查一下:
Anthropic是一家以“AI安全”为核心使命的美国人工智能公司。它由前OpenAI高管达里奥·阿莫迪(Dario Amodei)和丹妮拉·阿莫迪(Daniela Amodei)兄妹于2021年创立,是全球最知名的大模型企业之一,被视为OpenAI的最强劲对手。Anthropic开发了名为Claude的大语言模型系列。这个系列包括Opus、Sonnet,以及Haiku等版本,分别应对不同的复杂任务场景。除了通用的Claude对话模型,公司还推出了Claude Code——一款专门为程序员设计的AI编程工具,能直接在终端里协助编写和调试代码,已成为其重要的收入来源。
最近我还考虑要不要系统学习一下python编程,现在看来估计也没啥必要了,会用Claude Code就行了。
Other billionaires and execs – Jamie Dimon at JPMorgan Chase, Ginni Rometty at IBM,Satya Nadella at Microsoft, Mike Novogratz at Fortress Investment Group andJack Clark at Anthropic – have warned of the need for emotional intelligence and storytelling in a world dominated by A.I. 其他亿万富翁和高管——包括摩根大通的杰米·戴蒙、IBM的吉妮·罗梅蒂、微软的萨蒂亚·纳德拉、堡垒投资集团的迈克·诺沃格拉茨以及Anthropic的杰克·克拉克——都曾警告称,在人工智能主导的世界中,情商和讲故事的能力至关重要。
Some people are beginning to realize you have to avoid sauteing your brain in AI slop if you want to keep it fit. 有些人开始意识到,如果想保持头脑清醒,就必须避免让大脑在人工智能的垃圾信息中“煸炒”。
这里面有两个词值得记一下:
1) sauté vt. to cook something in a little hot oil or fat 想一下每次炒菜的时候,先热油加葱姜蒜翻炒炝锅,就是这个动作
2) AI slop: low-quality AI-generated content designed to farm views 俗称AI垃圾
