「真實威脅」還是「科幻情節」？
Artificial intelligence poses an existential threat to humanity and must be stopped “before it is too late,” at least according to some experts.
人工智慧對人類構成生存危機，且必須在「為時已晚」之前加以阻止，至少某些專家是這樣認為的。
exist (v.)生存、存在existential (adj.)存在的、存在主義的
Amid the recent AI craze, this is one of the clear warnings about the potential risks of new, “intelligent” technologies.
在最近人工智慧熱潮中，有人對於新興「智慧」技術的潛在風險提出這樣的明確警告。
craze (n.)風潮、熱潮warning (n.)警告potential (adj.)可能的、潛在的potential problems 潛在問題potential customers 潛在的客戶
But is this threat real or simply science fiction?
但這種威脅是真實存在還是純粹只是科幻情節呢？
simply (adv.)僅僅
In dystopian sci-fi movies, the plots are almost always the same: computers or robots evolve beyond human control, becoming super-intelligent beings that end up destroying humanity.
在反烏托邦科幻電影中，情節幾乎千篇一律：電腦或機器人超越人類控制，演變成超級智慧生物，最終摧毀人類。
utopia (n.)烏托邦dystopia (n.)反烏托邦dystopian (adj.)反烏托邦的evolve (v.)演化、進化revolve (v.)旋轉being (n.)生命、生物體
Although some great minds like Stephen Hawking have bought into various aspects of the idea that super-developed machines could “spell the end of the human race,” others believe this outcome is unlikely.
儘管有些像史蒂芬．霍金那樣的聰明人也相信超先進機器人可能會「招致人類滅絕」的觀點，但其他人認為這種結果不太可能發生。
Great minds think alike. 英雄所見略同。buy into 相信spell (v.)招致（不幸、災難等）
However, they do believe that there are very real dangers to AI.
然而，他們確實認為人工智慧存在非常真實的危險。
According to the Stop Killer Robots campaign, one of the more realistic AI scenarios that could harm humans is “giving machines the power to make decisions on life and death.”
根據「制止殺手機器人」運動的說法，現實中人工智慧比較可能傷害人類的情況之一是「讓機器擁有生死決策的權力」。
campaign (n.)宣傳活動marketing campaign 行銷活動election campaign 競選活動scenario (n.)情況、局面
Furthermore, Kerstin Dautenhahn, a robot expert at Canada’s Waterloo University, says that robots themselves are not evil, and it is not very plausible that AI will give them greater capabilities to reason or the “desire to kill all humans.”
此外，加拿大滑鐵盧大學的機器人專家多滕漢表示，機器人本身並不邪惡，人工智慧也不大可能賦予它們更強的推理能力或「殺光全部人類的慾望」。
evil (adj.)邪惡的plausible (adj.)貌似可能的、貌似有理的capability (n.)能力reason (v.)推理、推論
However, she notes that since programming dictates the actions of robots and AI systems, it is entirely possible that “bad actors” could utilize technology to do horrible things.
然而，她指出，由於機器人和人工智慧系統的行動受程式編寫的指令所影響，「壞人」完全有可能利用科技做出可怕的事情。
dictate (v.)命令、規定
Therefore, perhaps the warnings are justified, though not how they are intended.
因此，也許這些警告是有道理的，只是意義不同而已。
justified (adj.) 合理的，有道理的
In reality, it may be humans’ interactions with AI, not the technology itself, that poses the real danger.
實際上，人類與人工智慧的互動可能才是真正構成危險的因素，而不是科技本身。