Anthropic Makes a New Tool for Science
Anthropic Makes a New Tool for Science
Anthropic 推出全新科學研究工具
Introduction
Anthropic has a new AI tool called Claude Science. It helps scientists find new medicines.
Anthropic 推出了一款名為 Claude Science 的新 AI 工具，旨在協助科學家研發新藥。
Main Body
Claude Science uses the Claude AI. It connects to 60 science databases. The AI has a manager. This manager gives jobs to other small AI helpers. One AI checks the facts to stop mistakes.
Claude Science 使用了 Claude AI。它連接了 60 個科學數據庫。該 AI 設有一個管理器，由該管理器將任務分配給其他小型 AI 助手。其中一個 AI 負責核對事實以防止錯誤。
Anthropic wants to be better than Google and OpenAI. They let many people pay to use the tool. The company also has a new team. This team looks for cures for rare diseases in a lab.
Anthropic 希望能超越 Google 和 OpenAI。他們允許許多人付費使用該工具。公司還成立了一個新團隊，在實驗室中尋找罕見病的治療方法。
Anthropic is a very big company. It makes a lot of money. They want to sell the tool to big medicine companies. This helps the company grow before they sell shares to the public.
Anthropic 是一家規模很大的公司，擁有豐厚的獲利。他們計劃將該工具銷售給大型製藥公司。這有助於公司在公開上市之前實現增長。
Conclusion
People who pay for Claude can use this tool now. Anthropic is now in the science and medicine business.
目前付費使用 Claude 的用戶可以使用此工具。Anthropic 正式進入科學與醫療領域。
Vocabulary Learning
💡 The 'Power' of S
In this text, we see words like helps, uses, connects, and makes.
When we talk about one person, one company, or one thing (He / She / It), we add an -s to the action word.
See the pattern:
- Anthropic (It) → makes
- The AI (It) → uses
- The manager (He/She) → gives
🛠️ Useful Word Pairs
Notice how the text connects ideas using simple pairs:
- Find → medicines
- Stop → mistakes
- Look for → cures
Tip: To reach A2, stop using only one word. Start using Action + Thing (Verb + Object) to make your sentences feel complete.
🚀 Fast Vocabulary
- Rare → Not common / Hard to find.
- Public → For everyone.
- Grow → To get bigger.
Vocabulary Learning
Anthropic Launches Claude Science to Improve Computational Research
Anthropic 推出 Claude Science 以提升計算研究能力
Introduction
Anthropic has introduced Claude Science, a specialized AI workbench created to help the scientific community with computational research and the development of new drugs.
Anthropic 推出了 Claude Science，這是一個專為科學界設計的 AI 工作站，旨在協助計算研究與新藥開發。
Main Body
The launch of Claude Science shows that Anthropic is moving from general AI models toward tools designed for specific industries. This product is an evolution of the 'Claude for Life Sciences' plug-ins from 2025, changing from a simple chatbot addition to a complete, standalone environment. Technically, the workbench uses existing models, such as Opus 4.8, and connects them to over 60 scientific databases and chemistry tools. To ensure accuracy, the system uses a project manager AI that assigns tasks to smaller assistants, while a separate verification AI checks for fake citations.
Claude Science 的推出顯示 Anthropic 正從通用 AI 模型轉向針對特定產業設計的工具。此產品是 2025 年「Claude for Life Sciences」外掛程式的演進，從簡單的聊天機器人擴充功能轉變為一個完整、獨立的環境。在技術上，該工作站使用現有模型（如 Opus 4.8），並將其連接至超過 60 個科學資料庫與化學工具。為確保準確性，系統採用一個專案經理 AI 將任務分配給較小的助手，而另一個獨立的驗證 AI 則負責檢查有無虛假引用。
Anthropic is now competing directly with Google DeepMind and OpenAI. While OpenAI uses a restricted system for its biological models and Google DeepMind relies on its own private models like AlphaFold, Anthropic has chosen a wider distribution strategy through subscriptions. This move is supported by the hiring of expert John Jumper and the scientific experience of CEO Dario Amodei. Furthermore, the company has started its own drug discovery programs for neglected diseases. Management emphasized that this allows them to test their tools in real laboratories and improve the product based on actual results.
Anthropic 目前正與 Google DeepMind 和 OpenAI 直接競爭。OpenAI 在生物模型上使用限制性系統，而 Google DeepMind 依賴如 AlphaFold 等私有模型，Anthropic 則選擇透過訂閱制採取更廣泛的分發策略。此舉得到了專家 John Jumper 的加入以及執行長 Dario Amodei 科學經驗的支持。此外，公司已啟動針對被忽視疾病的藥物研發計畫。管理層強調，這使他們能在真實實驗室中測試工具，並根據實際結果改良產品。
From a financial point of view, entering the pharmaceutical market is a major step. With annual sales of $42 billion and a high company valuation, Anthropic aims to win expensive contracts with drug companies. These revenue streams are considered essential for maintaining profit and increasing the company's value before its planned public offering (IPO) later this year.
從財務角度來看，進入製藥市場是一個重大步驟。憑藉 420 億美元的年銷售額與高公司估值，Anthropic 旨在贏得製藥公司的高價合約。這些營收流被視為在今年稍後計劃公開上市（IPO）前，維持獲利並提升公司價值的關鍵。
Conclusion
Claude Science is now available in beta for paid subscribers, marking Anthropic's official entry into the market for computational biology and drug discovery.
Claude Science 目前已向付費訂閱用戶開放 Beta 版，標誌著 Anthropic 正式進入計算生物學與藥物開發市場。
Vocabulary Learning
🚀 The 'Nuance Shift': Moving from Simple to Complex Descriptions
At the A2 level, you likely say: "Anthropic made a new tool. It is for science. It is better than the old one."
To hit B2, you need to describe transitions and intentions. Look at these three linguistic upgrades found in the text:
1. Describing Evolution (The 'From To' Pattern)
Instead of saying "It changed," the article uses:
"...moving from general AI models toward tools designed for specific industries."
B2 Secret: Use this structure to show a process of change.
- A2: I studied English. Now I am better.
- B2: I am moving from basic grammar toward professional fluency.
2. The 'Power of Purpose' (Using Designed for and Aims to)
Stop using "want to" for everything. The text uses professional alternatives:
- "Designed for..." (Used for a specific group/reason)
- "Aims to..." (A professional goal)
Example from text: "...Anthropic aims to win expensive contracts."
3. Sophisticated Connectors (Beyond 'And' and 'But')
Notice how the text connects a fact to a result without using "so":
"Furthermore, the company has started its own drug discovery programs... Management emphasized that this allows them to test their tools..."
Quick Upgrade Table:
| A2 Word | B2 Upgrade (from text) | Why it's better |
|---|---|---|
| Also | Furthermore | Sounds more academic/formal |
| Let them | Allows them to | Focuses on the possibility/permission |
| Help | Support | More professional in a business context |
Vocabulary Learning
Anthropic Integration of Computational Research Workflows via Claude Science Launch
Anthropic 推出 Claude Science，將計算研究工作流整合入內
Introduction
Anthropic has introduced Claude Science, a specialized AI workbench designed to facilitate computational research and drug development for the scientific community.
Anthropic 推出了 Claude Science，這是一個專為科學界設計的 AI 工作台，旨在協助計算研究與藥物開發。
Main Body
The deployment of Claude Science represents a strategic transition from the provision of general-purpose large language models toward the establishment of vertical, industry-specific operating layers. This product evolves from the 'Claude for Life Sciences' plug-ins released in October 2025, transitioning from a chatbot augmentation to a standalone environment. Technically, the workbench does not utilize a novel model but leverages existing Claude iterations, including Opus 4.8, integrated with over 60 scientific databases and specialized toolkits for genomics and chemistry. The architecture employs a hierarchical agent system wherein a primary project manager delegates tasks to specialized sub-assistants, supported by a dedicated verification AI to mitigate the risk of fabricated citations.
Claude Science 的部署代表了公司從提供通用大型語言模型，戰略性地轉向建立垂直、針對特定行業的操作層。該產品由 2025 年 10 月發佈的 "Claude for Life Sciences" 插件演進而來，從聊天機器人的增強功能轉型為獨立環境。在技術上，該工作台並非使用新模型，而是利用現有的 Claude 版本（包括 Opus 4.8），並整合了超過 60 個科學資料庫以及基因組學和化學的專門工具包。其架構採用層級化代理系統，由一個主項目經理將任務委派給專門的子助手，並由專用的驗證 AI 支援，以降低虛構引用之風險。
Institutional positioning indicates a direct competitive challenge to Google DeepMind and OpenAI. While OpenAI has deployed a fine-tuned biological reasoning model (GPT-Rosalind) via a restricted enterprise gateway, and Google DeepMind utilizes proprietary foundational models such as AlphaFold, Anthropic has adopted a broad distribution strategy via subscription access. This maneuver is bolstered by the recruitment of John Jumper from DeepMind and the scientific background of CEO Dario Amodei. Furthermore, the company has commenced internal pre-clinical drug discovery programs targeting neglected diseases. This initiative is characterized by leadership as a method to establish tight feedback loops and validate product efficacy in real-world laboratory settings.
機構定位顯示，這是對 Google DeepMind 和 OpenAI 的直接競爭挑戰。OpenAI 雖然透過受限的企業門戶部署了微調後的生物推理模型 (GPT-Rosalind)，而 Google DeepMind 則使用如 AlphaFold 等專有基礎模型，但 Anthropic 採取了透過訂閱獲取的廣泛分發策略。此舉得到了從 DeepMind 招募的 John Jumper 以及執行長 Dario Amodei 科學背景的助力。此外，公司已啟動針對被忽視疾病的內部臨床前藥物研發計畫。領導層將此舉定義為建立緊密反饋迴路並在現實實驗室環境中驗證產品功效的方法。
From a fiscal perspective, the expansion into the pharmaceutical sector is significant. With reported annualized sales of $42 billion and a valuation of $965 billion, Anthropic is positioning itself to secure high-value contracts with pharmaceutical entities. Such revenue streams are viewed as critical for maintaining profitability and enhancing institutional valuation ahead of a projected initial public offering later this year.
從財務角度來看，擴展至製藥領域具有顯著意義。Anthropic 報告的年化銷售額為 420 億美元，估值達 9650 億美元，正致力於與製藥實體簽署高價值合約。此類收入流被視為在預計今年稍晚進行首次公開募股 (IPO) 之前，維持獲利能力並提升機構估值的關鍵。
Conclusion
Claude Science is currently available in beta for paid subscribers, marking Anthropic's formal entry into the computational biology and drug discovery market.
Claude Science 目前向付費訂閱用戶提供 beta 測試，標誌著 Anthropic 正式進入計算生物學與藥物研發市場。
Vocabulary Learning
The Architecture of Nominalization and 'Heavy' Noun Phrases
To bridge the gap from B2 to C2, a student must move beyond the subject-verb-object linearity and embrace Nominalization. The provided text is a masterclass in transforming actions (verbs) into concepts (nouns) to create a high-density, authoritative academic tone.
⚡ The 'Weight' Shift: From Process to Entity
Observe the phrase: "The deployment of Claude Science represents a strategic transition from the provision of general-purpose large language models toward the establishment of vertical, industry-specific operating layers."
In a B2 context, a student might write: "Anthropic deployed Claude Science and is now moving from giving out general models to creating specific layers for industries."
The C2 Transformation:
- Deploy The deployment
- Provide The provision
- Establish The establishment
By turning these verbs into nouns, the writer strips away the 'human' actor and focuses on the institutional phenomenon. This allows for the insertion of precise adjectives (strategic, vertical, industry-specific) that modify the concept rather than the action.
🧩 Syntactic Sophistication: The 'Hierarchical' Modifier
C2 mastery requires the ability to stack modifiers without losing grammatical cohesion. Consider this construction:
"...a dedicated verification AI to mitigate the risk of fabricated citations."
Analysis of the 'Semantic Chain':
Dedicated (Adj) Verification (Noun-as-Adj) AI (Head Noun) to mitigate (Infinitive of Purpose) the risk (Abstract Object) of fabricated citations (Qualifying Genitive).
This is not merely 'complex' English; it is compressed English. It conveys a specific technical function with zero wasted words.
🖋️ Lexical Precision: The 'Institutional' Register
The text utilizes verbs that denote systemic movement rather than simple change. Note the use of:
- Bolstered: Not just 'helped', but reinforced structurally.
- Mitigate: Not 'stop' or 'reduce', but to make a risk less severe.
- Leverages: Not 'uses', but utilizes an existing asset for a new advantage.
C2 Takeaway: To achieve this level, stop searching for synonyms and start searching for registers. Don't just replace 'use' with 'utilize'; replace the entire sentence structure with a nominalized phrase that frames the action as a strategic asset.