这学期选修了“英语高阶阅读”，老师每周提供英文原文要求阅读，但我天天懒得读，于是先放这里以后再读(也许？)，其实在练md语法

Reading 1 (News Reading)

The week when AI changed everything

By Auzinea Bacon UPDATED FEB 28, 2026

We’ve been talking about AI changing the world for a long time. But this week felt different — as if it were right in the middle of something big.The stock market tumbled1 — three separate days — because of investors’ both bearish and bullish2 fears about AI, including Nvidia’s somewhat tepid3 outlook and a viral4 blog post that imagined a hypothetical scenario5 in which white-collar work evaporated6.

Anthropic unveiled7 new tools that could upend8 the way people do work. Then Anthropic entered into an epic9 battle over red lines with the Pentagon that risked making the AI company a pariah10. At the same time, the company loosened its safety policy as the AI market continues to charge ahead11.

And then there was Block, which laid off 4,000 people — nearly half of its staff — because of AI. The company’s CEO predicted that other companies were about to do the same.

What a week, indeed. Here’s how it unfolded12.

Investor fears

The Dow tumbled more than 800 points on Monday, in large part because of a Substack post from Citrini Research that laid out hypothetical scenarios for how developments in AI could disrupt certain parts of the economy, including AI agents13 making white-collar work superfluous14.

The post, which was published last Sunday, specifically said it was not meant to be predictive. It was a work of fiction15.

And yet stocks for companies mentioned in the report — like DoorDash and American Express — tumbled on Monday.

Tech stocks tumbled again on Thursday after leading chipmaker Nvidia released its earnings. Although the company’s profit nearly doubled in the fourth quarter and sales reached an all-time high16, Wall Street seemed disappointed by its somewhat lackluster17 outlook. It fueled AI bubble concerns — that massive investments in AI infrastructure may not translate18 to big returns.

The steep19 declines this week showed just how on edge20 investors are regarding AI — even when the news is good, or when doomsday scenarios21 are completely made out of thin air.

Changes to Claude AI

Anthropic fed into disruption fears22 earlier this month when it announced an update to its Claude Cowork agent, stoking23 fears that it could replace dozens of software tools.

On Tuesday, Anthropic announced another update to Claude to improve the tool’s performance at specific jobs, such as design, human resources and wealth management roles, and it can now work within applications like Microsoft Excel and PowerPoint apps.

Anthropic denied that its tools will replace existing software tools and jobs, noting Claude Cowork is designed to be complementary24. But its rapid-fire25 updates are giving Wall Street whiplash26.

Anthropic vs. the Pentagon

Dario Amodei, Anthropic’s chief executive, went head-to-head27 with the Pentagon in a high-stakes28 battle over AI safety.

Anthropic set two red lines for its AI: Claude will not be used in autonomous weapons, and it will not be used in the mass surveillance29 of US citizens.

Defense Secretary Pete Hegseth said the Pentagon would use the Defense Production Act to access30 Anthropic’s technologies regardless of the company’s decision. He met with Amodei on Tuesday, saying that the Pentagon wanted to use the AI model “for all lawful purposes” and set a Friday deadline for the company to agree.

Hegseth also threatened to end Anthropic’s $200 million contract and could deem31 the company a “supply chain risk,” effectively barring32 all companies with military contracts from doing business with Anthropic.

Trump amplified33 that threat on Friday, posting on Truth Social that federal agencies must “immediately” stop using Anthropic’s technology. The president said there will be a “six month phase out period34” for agencies using Anthropic’s products.

“Anthropic better get their act together, and be helpful during this phase out period, or I will use the Full Power of the Presidency to make them comply, with major civil and criminal consequences to follow,” Trump said.

But Anthropic has rejected the Pentagon’s offer.

“Threats do not change our position: we cannot in good conscience accede to their request,” Amodei said in a blog post.

Meanwhile, on Wednesday, Anthropic had loosened its core safety policy to better adapt to a fast-moving market in which competitors may not abide by the same safety standards.

Block Layoffs

Economists and investors have long been worried for years that AI adoption could result in mass layoffs35.

Those fears seemed to come to fruition on Thursday when Block — the company behind Square, Cash App and Afterpay — said it would cut its staff by 40%. Co-founder Jack Dorsey plainly cited “intelligence tools” as the reason in a letter to shareholders.

The company would lay off more than 4,000 people, reducing its workforce to under 6,000. And Dorsey said in his note that “most companies are late” and he thinks “the majority of companies will reach the same conclusion.”

That sparked a whole host of36 new fears about a jobs apocalypse37 — one that many economists still doubt is coming en masse38. But Dorsey’s prediction of mass AI-induced layoffs sure seems closer to reality this week than at any time in history. (899 words)

Supplementary Reading

What the Anthropic AI safety saga39 is really all about

Analysis by Lisa_Eadicicco Lisa Eadicicco David_Goldman David Goldman

Anthropic has reached a familiar crossroads for a growing tech company: how to scale40 without compromising the principles that set it apart.

The AI company has made safety its guiding principle. It advocated for AI regulation and called for worker protections as AI replaces some human tasks. Anthropic has worked hard to send a specific message to customers: We’re the good guys.

Yet the self-imposed guardrails41 the company laid down to build that brand may now be forming obstacles to its success.

This week, the Pentagon gave Anthropic an ultimatum42: Drop your AI ethical43 restrictions or lose your $200 million contract and face a blacklisting44. Separately, also this week, Anthropic loosened its core safety policy to better adapt to a fast-moving market in which competitors may not abide by45 the same safety standards.

It’s unclear how this week will play out46 for Anthropic’s business and its reputation, but its decisions will be consequential47.

We know that, because Anthropic’s dilemma48 is a familiar one in the tech industry. Many companies tout49 their values and morality50, only to be confronted with tough decisions that force them to choose between growth and maintaining those ideals.

Anthropic may want to take note.

OpenAI and the weekend of Sam

Just over two years ago, Anthropic’s biggest rival grappled with dissent51 over growth at the cost of safety.

In one of the most bizarre52 boardroom dramas in corporate history, Anthropic’s chief rival OpenAI abruptly fired its founder and CEO Sam Altman on a November Friday in 2023, only to rehire him the following Tuesday.

The saga involved a unique corporate structure that placed the fast-growing, for-profit company behind ChatGPT under the auspices53 of a nonprofit board. Four years earlier, the company had written into its charter54 that OpenAI remained “concerned” about AI’s potential to “cause rapid change” for humanity. The company’s overseers55 feared that Altman was moving so fast that he risked undermining the safety the company pledged56 to provide.

But firing Altman led to threats of a mass exodus57 of employees – an untenable58 situation that could have led to the destruction of the company. So the board just days later rehired Altman. The board dissolved soon after, and Altman changed the corporate structure last year to free itself of its nonprofit overseer.

OpenAI has since struggled to balance speed and safety, facing several lawsuits that claim its products convinced young people to harm themselves. OpenAI denies those claims.

What’s next for Anthropic

These case studies offer a cautionary59 roadmap for Anthropic.

Now, the biggest near-term consequence for Anthropic is likely how clients and potential customers value and trust the company, said Owen Daniels, associate director of analysis at Georgetown’s Center for Security and Emerging Technology.

Anthropic said its self-imposed safety measures were always meant to be flexible and subject to change as AI evolves. It pledged to be transparent60 about safety in the future and said it really didn’t have a choice: If it stopped growing, rivals that don’t value safety as much could push ahead and make AI “less safe” overall.

It’s unclear what will come of Anthropic’s change, because AI’s existential61 risks are still largely “conceptual62,” noted Sundararajan.

He said he’d be skeptical63 of any expert who called this an important moment in AI safety. But it could be an important moment for the company.

“Pulling back from a particular safety promise here by Anthropic, to me, is more about Anthropic and less about the future of AI,” he said.

This story has been updated to include more details about Anthropic’s safety policy changes.

  1. tumbled /ˈtʌmbld/ 动词 (tumble的过去式及过去分词) 1. (价格、数字等)暴跌，骤降 2. 跌倒，摔落，滚落 3. 倒塌，坍塌 4. 倾泻，滚流；乱哄哄地涌出 5. (政府、政权等)倒台 ↩︎

  2. bearish /ˈberɪʃ/ adj. (无比较级) 1. (股市)行情看跌的 2. (态度)悲观的，消极的; bullish /ˈbʊlɪʃ/ adj. (无比较级) 1. (股市)行情看涨的 2. (态度)乐观的，积极的 ↩︎

  3. tepid /ˈtepɪd/ adj. (比较级 more tepid，最高级 most tepid) 1. (液体)微温的，温热的 2. (反应、热情等)不冷不热的，不热烈的，冷淡的 ↩︎

  4. viral /ˈvaɪrəl/ adj. (无比较级) 1. (医学)病毒的，病毒性的 2. (网络、信息)病毒式传播的，迅速流行的 ↩︎

  5. hypothetical scenario /ˌhaɪpəˈθetɪkəl səˈnærioʊ/ n. phr. 假设性情景，假想情况 ↩︎

  6. evaporated /ɪˈvæpəreɪtɪd/ 动词 (evaporate的过去式及过去分词) 1. (液体)蒸发，挥发 2. (情感、感觉、活动、工作等)消失，消散，逐渐减少 ↩︎

  7. unveiled /ʌnˈveɪld/ 动词 (unveil的过去式及过去分词) 1. (首次)展示，推出，公布 2. 为…揭幕，揭开…上的覆盖物 3. 揭露，披露 ↩︎

  8. upend /ʌpˈend/ v. (第三人称单数 upends，现在分词 upending，过去式 upended，过去分词 upended) 1. 颠倒，倒立 2. 颠覆，彻底改变 3. 推翻，打乱 ↩︎

  9. epic /ˈepɪk/ adj. (无固定比较级，可用 more epic, most epic) 1. 史诗般的，宏大的，壮丽的 2. 漫长而艰难的，艰苦卓绝的 3. (场面、战斗等)气势恢宏的，极具规模的 n. 1. 史诗，叙事诗 2. 史诗般的作品(电影、小说等) ↩︎

  10. pariah /pəˈraɪə/ n. 1. 被社会遗弃者，流浪者 2. (团体、组织中被)排斥的人(或事物)，不受欢迎的人(或事物) ↩︎

  11. charge ahead /tʃɑːrdʒ əˈhed/ v. phr. (常为现在分词或过去式) 1. 猛冲向前，向前冲刺 2. 继续全速前进，不顾阻碍地推进 3. (商业、技术等)快速发展，高歌猛进 ↩︎

  12. unfolded /ʌnˈfoʊldɪd/ 动词 (unfold的过去式及过去分词) 1. 展开，打开 2. 展现，呈现 3. (故事、事件等)发展，逐渐明朗 ↩︎

  13. AI agents /ˌeɪ ˈaɪ ˈeɪdʒənts/ n. pl. 人工智能代理，AI代理（指能够自主执行任务、做出决策并与环境交互的智能程序或系统） ↩︎

  14. superfluous /suːˈpɜːrfluəs/ /sjuːˈpɜːfluəs/ adj. (无比较级) 过多的，过剩的；多余的，不必要的 ↩︎

  15. fiction /ˈfɪkʃən/ n. (复数 fictions) 1. 小说 (文学作品) 2. 虚构的事，想象的事 3. 杜撰，谎话 ↩︎

  16. reached an all-time high /riːtʃt ən ˈɔːlˌtaɪm haɪ/ v. phr. (词组，reach an all-time high 的过去式) 创下历史新高，达到前所未有的最高水平 ↩︎

  17. lackluster (英式拼写 lacklustre) /ˈlækˌlʌstər/ adj. 无光泽的，暗淡的；无活力的，平淡无奇的，不出色的（文中指业绩展望乏善可陈） ↩︎

  18. translate /trænsˈleɪt/ /trænzˈleɪt/ v. (第三人称单数 translates，现在分词 translating，过去式 translated，过去分词 translated) 1. 翻译，译 2. (使)转变，(使)变为 3. 解释，说明 4. (以某种方式)理解，领会 5. (被)转化为，导致 ↩︎

  19. steep /stiːp/ adj. (比较级 steeper，最高级 steepest) 1. 陡峭的，险峻的 2. (价格、要求等)过高的，不合理的 3. (上升或下降)急剧的，大起大落的 v. (steeped，steeped，steeping) 1. 浸泡，浸渍 2. 使沉浸于，使深陷于 (be steeped in) ↩︎

  20. on edge /ɒn edʒ/ /ɑːn edʒ/ adj. phr. 1. (人)紧张不安的，烦躁的，如坐针毡的 2. (神经)紧绷的，敏锐的 3. 竖着的，立在边缘上的(较少用于字面义) ↩︎

  21. doomsday scenarios /ˈduːmzdeɪ sɪˈnɑːrioʊz/ n. phr. 1. 末日情景，世界末日设想 2. (经济、科技等领域的)灾难性后果预测，最坏情况假设 ↩︎

  22. fed into disruption fears /fed ˈɪntuː dɪsˈrʌpʃən fɪəz/ v. phr. (feed into 的过去式，后接名词) 1. 加剧了(人们对行业)颠覆的担忧 2. 助长了(对混乱、失业等)恐惧情绪 3. 使(原本存在的)担忧进一步恶化 ↩︎

  23. stoking /ˈstoʊkɪŋ/ 动词 (stoke 的现在分词) 1. 给(炉子等)添燃料，拨旺(炉火) 2. 煽动，激起，加剧(情绪、恐惧、冲突等) ↩︎

  24. complementary /ˌkɑːmplɪˈmentri/ adj. (无比较级) 1. 互补的，补充的，相辅相成的 2. (色彩)互补的(混合后产生白色或灰色) ↩︎

  25. rapid-fire /ˈræpɪd faɪər/ adj. 1. (枪炮)速射的，快射的 2. (问题、话语等)连珠炮似的，快速连续的 3. (更新、变化等)迅速的，接连不断的 ↩︎

  26. whiplash /ˈwɪplæʃ/ n. 1. 鞭绳，鞭梢 2. （车辆碰撞等引起的）挥鞭式颈部扭伤，颈椎过度屈伸损伤 v. 1. 像鞭子一样猛抽，突然甩动 2. 使…受到猛烈冲击；使…陷入混乱(或震惊) 引申义（文中用法）: 指(市场、心态等)因快速变化而遭受的剧烈震荡、措手不及的强烈反应。 ↩︎

  27. head-to-head /ˌhed tə ˈhed/ adj./adv. 1. (竞争、比赛等)正面交锋的(地)，直接对抗的(地) 2. 势均力敌的(地) n. (尤指体育比赛中的)直接较量，对决 ↩︎

  28. high-stakes /ˌhaɪ ˈsteɪks/ adj. (无比较级) 1. 高赌注的，高风险高回报的 2. (形势、竞争等)意义重大的，后果严重的 ↩︎

  29. surveillance /sərˈveɪləns/ n. (不可数) 1. (对嫌疑人、场所等的)监视，监控 2. (政府、机构等的)监察，盯梢 ↩︎

  30. access /ˈækses/ n. (不可数) 1. (进入某地、使用某物或见到某人的)机会，权利，通道 2. 进入，到达 3. 查阅权，使用权限 v. (第三人称单数 accesses，现在分词 accessing，过去式 accessed，过去分词 accessed) 1. 访问，存取(计算机数据) 2. 进入，到达 3. 使用，获取 ↩︎

  31. deem /diːm/ v. (正式用法，通常不用进行时，第三人称单数 deems，过去式 deemed，过去分词 deemed) 认为，视为，相信 ↩︎

  32. barring /ˈbɑːrɪŋ/ 介词 1. 除非，如果没有 2. 除…之外 动词 (bar的现在分词) 1. 禁止，阻止，阻挡 2. 封锁，阻挡 3. 把…排除在外 ↩︎

  33. amplified /ˈæmplɪfaɪd/ 动词 (amplify的过去式及过去分词) 1. 放大，增强(声音、信号等) 2. 扩大，加强，强化(效果、影响等) 3. 详述，进一步阐述 ↩︎

  34. six month phase out period /sɪks mʌnθ feɪz aʊt ˈpɪriəd/ n. phr. 六个月的逐步淘汰期，为期六个月的阶段性停止使用期（指在规定时间内分阶段终止对某项产品或技术的使用） ↩︎

  35. mass layoffs /mæs ˈleɪɔːfs/ n. phr. (复数，常作 mass layoffs) 大规模裁员，集体解雇 ↩︎

  36. a whole host of /ə hoʊl hoʊst əv/ det. phr. 大批，大量，众多（后接可数名词复数） ↩︎

  37. apocalypse /əˈpɑːkəlɪps/ n. (复数 apocalypses) 1. (尤指世界末日的)大灾难，大动乱 2. 启示，天启 (尤指基督教《圣经》中的《启示录》) 3. (在本文组合词 job apocalypse 中) 大规模失业灾难，工作末日 ↩︎

  38. en masse /ɑːn ˈmæs/ /ɒn ˈmæs/ adv. (法语借词，英语中作副词使用) 1. 全体地，一同地，集体地 2. 大量地，大批地 ↩︎

  39. saga /ˈsɑːɡə/ n. (复数 sagas) 1. (中世纪冰岛或挪威的)英雄传奇，萨迦 2. 长篇故事，(尤指)长篇冒险故事 3. (通常指经历一系列事件、持续很长时间的)一长串事件，一整套经历，家世小说 4. (新闻报道中常指)一连串的麻烦事，风波 ↩︎

  40. scale /skeɪl/ n. 1. 规模，范围 2. 等级，级别 3. 刻度，标度 4. 鳞片 5. 秤，磅秤 v. (第三人称单数 scales，现在分词 scaling，过去式 scaled，过去分词 scaled) 1. 攀登，攀爬 2. 按比例放大(或缩小) 3. (企业、业务等)扩大规模，增长，扩张(文中用法) ↩︎

  41. self-imposed guardrails /ˌself ɪmˈpoʊzd ˈɡɑːrdreɪlz/ n. phr. 1. 自行设定的护栏，自我施加的限制或约束 2. (在AI安全语境下)企业或组织主动制定的安全准则与道德红线 ↩︎

  42. ultimatum /ˌʌltɪˈmeɪtəm/ n. (复数 ultimatums 或 ultimata) 最后通牒（指一方向另一方提出的、要求在指定期限内接受其条件否则将面临惩罚或行动的最终要求） ↩︎

  43. ethical /ˈeθɪkl/ adj. (比较级 more ethical，最高级 most ethical) 1. 伦理的，道德的（关于是非善恶原则的） 2. 合乎道德的，有道德操守的（行为、决策等符合伦理标准） n. (常用复数 ethics) 伦理学，道德标准，行为准则 ↩︎

  44. blacklisting /ˈblæklɪstɪŋ/ 名词 (不可数，也作动词的现在分词，blacklist的动名词形式) 1. 黑名单制度，列入黑名单的行为 2. (商业、政治等领域)将…列入黑名单，排斥，拒绝与…往来 文中用法：指五角大楼威胁将Anthropic认定为“供应链风险”，从而迫使所有与军方有合同的公司不得与Anthropic合作。 ↩︎

  45. abide by /əˈbaɪd baɪ/ v. phr. (第三人称单数 abides by，现在分词 abiding by，过去式 abided by，过去分词 abided by) 1. 遵守，遵循(法律、规则、协议、决定等) 2. 信守，履行(诺言) 3. 接受，服从(某人的裁决等) ↩︎

  46. play out /pleɪ aʊt/ v. phr. (第三人称单数 plays out，现在分词 playing out，过去式 played out，过去分词 played out) 1. (事件、局势等)逐渐发生，展开，发展 2. (把剧本等)演完，将(戏剧)演到底 3. 耗尽，用完(供应、精力等) (文中用法为第1义) ↩︎

  47. consequential /ˌkɑːnsɪˈkwenʃl/ adj. 1. 随之发生的，作为结果的 2. 重要的，重大的，影响深远的（文中用法） ↩︎

  48. dilemma /dɪˈlemə/ /daɪˈlemə/ n. (复数 dilemmas) 1. 进退两难的境地，窘境，困境 2. (逻辑学)双难推理，两刀论法 ↩︎

  49. tout /taʊt/ v. (第三人称单数 touts，现在分词 touting，过去式 touted，过去分词 touted) 1. 兜售，招揽(生意) 2. 吹嘘，吹捧，标榜 3. (尤指在体育比赛或赛马中)兜售内部信息，卖赌马信息 n. 兜售者，招揽者，票贩子 ↩︎

  50. morality /məˈræləti/ n. (复数 moralities) 1. 道德，伦理 (关于是非善恶的原则或体系) 2. 德行，道德行为 (符合道德规范的程度) 3. 道德观，道德准则 (特定社会、群体或个人信奉的道德标准) ↩︎

  51. rival /ˈraɪvəl/ n. 竞争对手，敌手 adj. 竞争的，对抗的 v. 与…相匹敌，比得上; grappled with /ˈɡræpld wɪð/ v. phr. (grapple with 的过去式) 1. 与…扭打，搏斗 2. 努力对付，竭力解决(难题、异议等)，应对，处理; dissent /dɪˈsent/ n. (正式) 异议，不同意见 v. 持异议，不同意 ↩︎

  52. bizarre /bɪˈzɑːr/ adj. (比较级 more bizarre，最高级 most bizarre) 怪诞的，奇异的，极其古怪的（常带有令人意想不到或反常的意味） ↩︎

  53. auspices /ˈɔːspɪsɪz/ n. pl. (复数形式，常用 “under the auspices of”) 在…的支持(或保护、赞助)下；由…主办(或主持) ↩︎

  54. charter /ˈtʃɑːrtər/ n. 1. 宪章，章程（阐述宗旨、权利、原则的正式文件） 2. 特许状，许可证 3. 包租，租赁 v. 1. 特许成立，发给许可证 2. 包租（飞机、船等） ↩︎

  55. overseers /ˈoʊvərsɪrz/ n. pl. (单数 overseer) 监督者，监管者，监察人 ↩︎

  56. pledged /pledʒd/ 动词 (pledge的过去式及过去分词) 1. 发誓，保证，承诺 2. 许诺给予，抵押 3. 使(某人)保证，使宣誓 ↩︎

  57. exodus /ˈeksədəs/ n. (通常作单数，大写 Exodus 指《出埃及记》) 1. (大批人同时)离开，出走，外流，迁移 2. (在文中) 员工的大规模离职 ↩︎

  58. untenable /ʌnˈtenəbl/ adj. (无比较级) 1. (论点、理论、立场等)站不住脚的，不堪一击的 2. (局面、处境等)难以维持的，无法继续的，不能长久的 3. (地方)无法防守的 ↩︎

  59. cautionary /ˈkɔːʃəneri/ adj. (无比较级) 警告性的，告诫的，警世的 ↩︎

  60. transparent /trænsˈpærənt/ /trænzˈpærənt/ adj. (比较级 more transparent，最高级 most transparent) 1. 透明的，清澈的 2. 显而易见的，一目了然的，易识破的 3. 坦诚的，公开透明的(指言行、过程等没有隐瞒) 4. (文章、风格等)明晰的，易懂的 ↩︎

  61. existential /ˌeɡzɪˈstenʃl/ adj. (无比较级) 1. (哲学)存在的，关于存在的，存在主义的 2. (风险、威胁、危机等)关乎生存的，致命的(可能导致毁灭或终结) ↩︎

  62. conceptual /kənˈseptʃuəl/ adj. (无比较级) 1. 概念的，观念上的 2. 理论层面的，非实证的 ↩︎

  63. skeptical (英式拼写 sceptical) /ˈskeptɪkl/ adj. (比较级 more skeptical，最高级 most skeptical) 1. 怀疑的，持怀疑态度的 2. 不相信的，不轻信的（常与 about / of 连用） ↩︎