重口味
PLA Navy's first-batch female shipborne pilot trainees pass daytime landing qualification tests, signaling mature pilot training system: expert_我的网站
A |
新智元报道
Anthropic把自家员工入职培训的那套课程，原样搬到了网上。 
The progress of female pilot training is an important indicator of the maturity and systemization of the shipborne aviation talent selection and training system, a military affairs expert said. Smears by some foreign media about China's pilot training capabilities are, in essence, a projection of the West's own anxieties over talent shortages, the expert added.
"Night landings are tougher than I imagined. The entire sea is pitch black. It's hard to tell where the sea ends and the sky begins. It's very easy to become disoriented," said Han Meng, a female trainee pilot at the Naval Aviation University, in the CCTV News video.
"Completing today's landing mission has given me a clearer understanding of my own abilities. I'm about to graduate, and I feel more confident about heading to my post and carrying out various maritime and far-sea missions I'll be assigned," she added.
The CCTV News report showed that the assessment featured a Z-8 helicopter landing on the aft flight deck of a vessel.
CCTV reported on Friday that the first batch of female shipborne pilot trainees had all recently passed their daytime shipborne landing qualification assessments. Shipborne landing is a critical milestone in a shipborne pilot's development, often described as the "certificate" to real naval aviation. Passing this assessment marks a significant step in the PLA Navy's efforts to cultivate female shipborne aviation talent.
All pilot trainees, regardless of gender, must meet the same rigorous physical and psychological standards required for shipborne aviation. The successful qualification of these female trainees demonstrates that the selection and training systems are capable of identifying and developing top-tier talent based on capability, not gender, Zhang Junshe, a Chinese military affairs expert, told the Global Times on Sunday.
In recent years, the PLA Navy's shipborne aviation talent development has entered a new phase of system-of-systems operations, transitioning from "able to fly and land on ships" to "able to fight and operate jointly," PLA Daily reported on April 22.
Under the synchronized training objectives of developing pilots, combatants and commanders, new pathways for talent cultivation are being explored, including growth opportunities for shipborne fighter and helicopter pilots. Training programs are being optimized across all elements, support resources coordinated across all domains, and implementation advanced across all processes, forming a systematic, integrated talent development framework.
Zhang also mentioned that foreign media outlets have hyped claims of a "pilot shortage" in China or that "talent development cannot keep pace with equipment delivery." Such claims are largely based on selective data comparisons and misunderstandings of China's aviation development, essentially reflecting the West's own anxieties about talent.
First, foreign media outlets often simply compare the total number of licensed pilots in China with that of the US, ignoring differences in population size, the prevalence of aviation culture and historical development. China's nationwide selection process draws from a vast pool of candidates, ensuring high overall pilot quality and building a strong reserve of aviation talent, Zhang noted.
In July, the PLA Navy's 2026 pilot recruitment process was successfully completed. The average college entrance examination scores of admitted pilot trainees reached a record high. In recent years, through extensive outreach and publicity campaigns, enthusiasm among young applicants has continued to grow, with the recruitment pool expanding year by year.
The Navy's recruitment authorities have established carrier-based flight aptitude assessment platforms, introduced new testing technologies including virtual reality and biofeedback, and improved the comprehensiveness and accuracy of flight aptitude evaluations, said the PLA Navy.
Meanwhile, China has established a diversified training system encompassing academy training, airline-commissioned training and civilian supplementary pathways. Through general aviation development, China is expanding its civilian talent reserve and gradually narrowing the gap with leading aviation nations, the expert said.
Moreover, modern air warfare is a complex system undertaking, relying not only on the number of pilots but also on comprehensive system-of-systems support across strategic early warning, electronic warfare, command communications and ground maintenance. The leapfrog development of China's aviation equipment is built on overall improvements in comprehensive national strength and the aviation industrial system, rather than a simple linear increase in pilot numbers, Zhang added.
。就在昨天，Claude Academy正式上线。
289个资源：课程、教程、真实用例全在里面，从「AI到底是什么东西」一路到Claude Code、API、MCP、agents和生产级部署。最重要的是，注册一个免费Claude账号就能看，学完发结业徽章，可以公开验证，也能挂到LinkedIn上。按照博客上的说法，本来这些内容原本是给新员工入职培训上的。
B |
从零通关Claude CodeAnthropic 表示每天有数百万人跑到他们的官网上学习怎么使用 AI，公司为此还专门养了一支教育团队。Claude Academy就是这支教育团队的总入口。零基础的可以先搞清楚 AI 能干什么，然后上手对话、Cowork这些。
C | 想写代码的更直接。Claude Code从入门到实战课都摆在那儿，再往后就是API、MCP、agents，一路到把东西部署上线。
课程还按人群分了岔路，写代码的、老师的、搞公益的、开店的，各走各的。
D | 底下有一门通识课大概14节，相当于整个体系的地基，不管你是谁都建议先过一遍。
装上Claude Academy Skill，它还会照着你的习惯和进度，直接告诉你下一步该看哪个。
最硬的一块：入职第一天学的4D框架整个Academy最核心的东西，是Anthropic和Rick Dakan、Joseph Feller两位教授一起开发的 AI Fluency 4D框架。
E | 简单说就四件事：Delegation——哪些活交给AI，哪些自己干。Description——把需求和过程说清楚。
F | Discernment——挑剔地检查AI交回来的东西。Diligence——用得负责任，并且告诉别人AI在里面做了什么。
G | Anthropic 的新员工入职第一天就要学习这个框架。它练的不是怎么写 Prompt，而是怎么判断哪些活该给 AI，哪些要自己干。还要学习 AI 最容易在哪些地方犯蠢，免得被带到沟里。
入职之后这些培训也不会停，他们管这叫「ever-boarding」。
H | 说白了就是没有「结业」这回事，因为AI的能力边界一直在变，人的认知也得跟着刷新。活到老，学到老。教育团队讲了一件挺反常识的事：具体技巧正在快速贬值。以前你得先告诉Claude这份东西是写给法务同事看的，现在它自己就会来问你了。模型在进化，昨天的最佳实践今天可能就是多余动作。
所以他们把重心从教动作挪到了「养心智」，翻来覆去就这两句话——今天的AI，是你这辈子会用到的最差的AI。按风险高低决定验证力度。课程还专门提醒，别把自己在意的本事全交出去，那会使技能萎缩。比如，敏感的内容还是要自己写，汇总幻灯片丢给AI；探索性的数据分析交给Claude跑，但最后的核对还是要自己来。
这已经不是教你怎么使用工具那么简单了。这是在学习如何跟一个越来越强大的非人类队友相处、配合。
I |
竞争已经换了赛道官方@claudeai那条帖子发出去几小时，点赞过万，转帖上千，浏览量冲到百万级。
网友@hiro44_pino连夜写了长帖，说289个资源不用全啃，每天看一点，立刻拿来实践。
J | 先用官方免费教材就够了，比买高价课划算得多。
K | 实用派给了更短的路线：先看7分钟的The 4 Properties of AI，再拿手上的活试一个具体的场景，卡住了再回来补长课程。有网友表示课程还是很实用的，自己连续 6 个月每天使用 Claude，还是能在高级课程中学习到新的技巧。
开发者 Kadam 认为，这种免费资源是一种巨大的胜利。像 Claude Code 这样的工具教程对优化很有用。
当然也少不了阴阳怪气的声音，说这是教你如何更高效地被AI取代自己。模型在往前跑，人得跟上这件事真正值得琢磨的地方在于：它动摇的不只是AI培训市场，而是整个「学习」这件事的逻辑。传统教育体系是先花四年打基础，再用经验慢慢长。
L | 但AI的能力边界每几个月就刷新一次，你半年前学的prompt技巧可能下个版本就内置了。知识的保质期从没这么短过。
Anthropic的解法是，不教你记招式，教你练心法。
M | 不教你按哪个按钮，教你判断什么时候该按、什么时候该自己来。4D框架的底层逻辑说穿了就一句话：人要永远比工具多想一步。Anthropic把新人第一天的课免费放出来，其实是在说一件事：光是模型变强没用，用的人跟不上，白搭。那句「今天的AI是你这辈子会用到的最差的AI」，反过来读更扎心：它明天一定会更强，而大多数人的用法，还停在昨天。
Current article:http://ih8w.cheshenzhuaduanza.buzz/news/20260826_3442336.html
Published on:20:11:39