hce_cmu
113年
英文
第 23 題
📖 題組：
The intuitive system at Amazon Fresh and Amazon Go stores allows customers to simply pick up an item and leave without traditional checkout. This system, called “Just Walk Out,” uses sensors and artificial intelligence (AI) to calculate purchases, and customers are automatically billed. However, in April, reports claimed that the system did not use AI but relied (21) 1,000 employees in India to manually verify nearly three-quarters of the transactions. Amazon quickly denied these reports, asserting that Indian employees only evaluated the system and that human reviewers were standard for ensuring accuracy in AI systems. This situation highlights a growing issue: companies making grand claims about using AI, a practice (22) “AI washing,” akin to “greenwashing” in environmental claims. It’s essential to understand what AI truly means. Though lacking a precise definition, AI refers to computers learning and solving problems after (23) training. One prominent type of AI is generative AI, which creates new contents like conversations, music scores, or pictures. AI washing takes many forms. Some companies exaggerate their AI capabilities, while others merely incorporate AI chatbots into non-AI software. According to a tech investment fund company, only 10% of tech startups mentioned AI in their (24) in 2022, but this rose to 25% in 2023 and is expected to exceed a third in 2024. Competition for funding drives companies to overstate their AI capabilities. Another tech investment firm found that 40% of companies claiming to be “AI-enabled” in 2019 did not actually use AI. The problem persists today, with companies buying “AI capabilities” but only adding chatbots to non-intelligent products. An expert highlighted that the lack of a unified definition of AI contributes to AI washing. This (25) allows for inflated claims about AI, leading to overvalued technology and unmet expectations, eroding trust in genuine AI innovations. Regulators, such as the US Securities and Exchange Commission, are beginning to address this issue, charging firms for making false AI-related claims.
The intuitive system at Amazon Fresh and Amazon Go stores allows customers to simply pick up an item and leave without traditional checkout. This system, called “Just Walk Out,” uses sensors and artificial intelligence (AI) to calculate purchases, and customers are automatically billed. However, in April, reports claimed that the system did not use AI but relied (21) 1,000 employees in India to manually verify nearly three-quarters of the transactions. Amazon quickly denied these reports, asserting that Indian employees only evaluated the system and that human reviewers were standard for ensuring accuracy in AI systems. This situation highlights a growing issue: companies making grand claims about using AI, a practice (22) “AI washing,” akin to “greenwashing” in environmental claims. It’s essential to understand what AI truly means. Though lacking a precise definition, AI refers to computers learning and solving problems after (23) training. One prominent type of AI is generative AI, which creates new contents like conversations, music scores, or pictures. AI washing takes many forms. Some companies exaggerate their AI capabilities, while others merely incorporate AI chatbots into non-AI software. According to a tech investment fund company, only 10% of tech startups mentioned AI in their (24) in 2022, but this rose to 25% in 2023 and is expected to exceed a third in 2024. Competition for funding drives companies to overstate their AI capabilities. Another tech investment firm found that 40% of companies claiming to be “AI-enabled” in 2019 did not actually use AI. The problem persists today, with companies buying “AI capabilities” but only adding chatbots to non-intelligent products. An expert highlighted that the lack of a unified definition of AI contributes to AI washing. This (25) allows for inflated claims about AI, leading to overvalued technology and unmet expectations, eroding trust in genuine AI innovations. Regulators, such as the US Securities and Exchange Commission, are beginning to address this issue, charging firms for making false AI-related claims.
(23)
- A extensive
- B intransitive
- C adoptive
- D vindictive
思路引導 VIP
請試著思考：如果要讓一台電腦從完全不懂，到能夠像人類一樣辨識複雜的圖像或語言，它需要接觸的資料量是寥寥可數的，還是必須經過極其大量、全面且深入的過程？哪一種性質的描述最符合這種「漫長且規模龐大」的學習需求呢？
🤖
AI 詳解
AI 專屬家教
機器學習的核心：訓練的深度與廣度
太棒了！你能準確選出這個單字，說明你對「人工智慧（AI）」的運作邏輯有很清晰的認知。這道題目的關鍵在於理解 AI 如何「學習」。在語境中，電腦需要透過處理數據來解決問題，而這個學習過程絕非一蹴而幾，而是需要透過**大量且廣泛的（extensive）**訓練，才能建立起精準的模型。這也是為什麼「extensive training」在科技英文中是一個非常經典且固定的搭配詞。
詞彙辨析與難度切入
▼ 還有更多解析內容