The Download: threats from space mirrors and credit for AI drugs

查看原文 / View Original 编辑 / Edit

The Download: threats from space mirrors and credit for AI drugs

The Download：太空镜的威胁与人工智能药物的署名权

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. 这是我们工作日时事通讯《The Download》的今日版，为您提供每日科技界动态。

This company’s plans to deploy space mirrors could jeopardize the night sky for many

这家公司部署太空镜的计划可能会破坏许多人的夜空

A company that plans to beam sunlight from space to Earth on demand might unintentionally brighten the night sky for many more people than intended, according to a new study. Later this year, Reflect Orbital plans to launch a test satellite that will extend an 18-by-18-meter mirror in orbit. The goal is to eventually launch up to 50,000 larger satellites that can reflect sunlight to Earth on demand. Reflect Orbital says the technology could extend sunlight for solar panel charging, emergency response, and military activities. But new research suggests the giant beams could shine as bright as 10,000 full moons and scatter light over tens of kilometers, raising concerns about dark skies, aviation, and wildlife. Read the full story on the fierce debate over the space mirrors. —Jonathan O’Callaghan 一项新研究显示，一家计划按需将阳光从太空反射到地球的公司，可能会在无意中让比预期多得多的人面临夜空变亮的问题。今年晚些时候，Reflect Orbital 公司计划发射一颗测试卫星，在轨道上展开一面 18x18 米的镜子。其目标是最终发射多达 50,000 颗更大的卫星，按需将阳光反射回地球。Reflect Orbital 表示，这项技术可以延长光照时间，用于太阳能电池板充电、应急响应和军事活动。但最新研究表明，这些巨大的光束亮度可能相当于 10,000 轮满月，并将光线散射到数十公里外，引发了人们对暗夜保护、航空安全和野生动物生存环境的担忧。阅读关于太空镜激烈辩论的完整报道。——Jonathan O’Callaghan

When AI designs a drug, who gets the credit?

当人工智能设计出药物时，谁该获得功劳？

When Insilico Medicine used its computer models to propose a promising drug for pulmonary fibrosis, it enthusiastically claimed that the molecule had been “discovered by” generative AI. But when it filed a patent, the company made no mention of AI. Instead, it named five humans as the drug’s “inventors.” The discrepancy points to a fascinating wrinkle in intellectual-property law. No matter how fundamental an AI is to a discovery, when it comes to winning rights to an invention, it’s humans—and only humans—who can take the credit. As AI models become capable of generating drug designs as easily as ChatGPT can write a thank-you note, the question of who actually invented something could get much harder to answer. Here’s how AI drug discovery could force patent law to confront what it means to invent. —Antonio Regalado 当英矽智能（Insilico Medicine）利用其计算机模型提出一种治疗肺纤维化的潜力药物时，该公司热情地宣称该分子是“由生成式人工智能发现的”。但在申请专利时，该公司却只字未提人工智能，而是将五名人类列为该药物的“发明人”。这种差异揭示了知识产权法中一个引人入胜的矛盾点：无论人工智能在发现过程中发挥了多么基础的作用，在争取发明权利时，只有人类才能获得署名权。随着人工智能模型生成药物设计变得像 ChatGPT 写感谢信一样简单，关于“谁才是真正的发明者”这一问题将变得愈发难以回答。以下是人工智能药物发现如何迫使专利法重新审视“发明”定义的分析。——Antonio Regalado

The must-reads

必读内容

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology. 我为您搜罗了互联网上今天最有趣、最重要、最令人担忧和最引人入胜的科技新闻。

  1. The data center backlash is scrambling the midterm elections It’s breaking the bipartisan consensus behind America’s AI buildout. (Axios)

    • Politicians are turning against the data centers they once championed. (WSJ)
    • Across the political divide, people hate data centers. (MIT Technology Review)

  2. 数据中心引发的抵制浪潮正在扰乱中期选举 这打破了美国人工智能建设背后的两党共识。（Axios）

    • 政治家们开始反对他们曾经支持的数据中心。（《华尔街日报》）
    • 跨越政治分歧，人们普遍厌恶数据中心。（《麻省理工科技评论》）

  3. Ukraine planned to swarm Moscow airports with AI-guided drones The stalled operation aimed to send 1,000 autonomous drones a night. (Atlantic)

    • Kyiv wants Musk’s approval to use Starlink-equipped drones in Russia. (FT)
    • Ukraine’s ex-defense minister wants defense tech funding from the US. (Reuters)

  4. 乌克兰曾计划用人工智能制导无人机蜂群袭击莫斯科机场 这项搁置的行动旨在每晚派遣 1,000 架自主无人机。（《大西洋月刊》）

    • 基辅希望获得马斯克的批准，在俄罗斯境内使用配备星链的无人机。（《金融时报》）
    • 乌克兰前国防部长希望获得美国的国防技术资金。（路透社）

  5. A staggering 90% of biomedical papers show signs of AI use A new study suggests it’s far more common than previously thought. (Nature)

    • A third of new web pages also involve AI authorship. (TechCrunch)
    • AI for science needs reasoning, not just data. (MIT Technology Review)

  6. 惊人的 90% 生物医学论文显示出人工智能使用的迹象 一项新研究表明，这种情况比之前想象的要普遍得多。（《自然》）

    • 三分之一的新网页也涉及人工智能创作。（TechCrunch）
    • 科学人工智能需要推理能力，而不仅仅是数据。（《麻省理工科技评论》）

  7. China plans to make the first landing at the moon’s south pole this year The Chang’e 7 mission will hunt for lunar ice. (Scientific American)

    • It will deploy robotic landers, rovers, and “hopping” probes. (Reuters)

  8. 中国计划今年首次在月球南极着陆 嫦娥七号任务将寻找月球冰。（《科学美国人》）

    • 该任务将部署机器人着陆器、月球车和“跳跃”探测器。（路透社）

  9. We need new measures of AI intelligence Better experiments could reveal what AI actually understands. (Quanta)

    • And human-centered, context-specific metrics. (MIT Technology Review)

  10. 我们需要衡量人工智能智能的新标准 更好的实验可以揭示人工智能到底理解了什么。（Quanta）

    • 以及以人为本、针对特定情境的指标。（《麻省理工科技评论》）

  11. A tech billionaire wants to automate the restaurant industry Marc Lore aims to control everything from production to delivery. (FT)

  12. 一位科技亿万富翁想要实现餐饮业自动化 Marc Lore 旨在控制从生产到配送的每一个环节。（《金融时报》）

  13. An English county is rejecting Palantir for a homegrown platform Greater Manchester insists it can do a better job itself. (Wired)

  14. 英国一个郡拒绝使用 Palantir，转而选择本土平台 大曼彻斯特地区坚持认为自己能做得更好。（《连线》）

  15. New research estimates humans could live for 194 years If we cure all the theoretically curable causes of aging. (New Scientist)

    • Longevity enthusiasts are gaining influence. (MIT Technology Review)

  16. 最新研究估计人类寿命可达 194 岁 前提是我们治愈所有理论上可治愈的衰老原因。（《新科学家》）

    • 长寿爱好者们的影响力正在扩大。（《麻省理工科技评论》）

  17. The galaxy’s fastest star could show how a black hole warps spacetime Its orbit could reveal the speed of Sagittarius A*’s spin. (Wired)

  18. 银河系最快的恒星可能揭示黑洞如何扭曲时空 它的轨道可以揭示人马座 A* 的自转速度。（《连线》）

  19. Mark Zuckerberg has bought an Irish castle The 440-acre estate reportedly cost up to €30 million. (Verge)

  20. 马克·扎克伯格买下了一座爱尔兰城堡 据报道，这座 440 英亩的庄园耗资高达 3000 万欧元。（The Verge）

Quote of the day

今日名言

“The emperor is a fan of Flock, and we must continue utilizing Flock technologies so that we can follow and surveil the rebel scum.” —Darth Vader (or, at least, a man in his costume), passionately defends Flock at a San Diego public safety committee meeting, 404 Media reports. “皇帝是 Flock 的粉丝，我们必须继续使用 Flock 技术，以便我们能够追踪和监视那些叛军渣滓。” ——达斯·维达（或者至少是一个穿着他戏服的人）在圣地亚哥公共安全委员会会议上热情地为 Flock 辩护，404 Media 报道。

One More Thing

额外关注

Meet the new biologists treating LLMs like aliens 认识一下那些把大语言模型当作外星生物研究的新生物学家

LLMs are such vast and complicated systems that nobody quite understands what they are, how they work, or what they can really do—not even the people who build them. That makes it hard to get a grip on their hallucinations, set up effective guardrails, and know when to trust them. So researchers are trying something different: studying them as if they were doing biology or neuroscience on vast living creatures—city-size xenomorphs that have appeared in our midst. They’re discovering that large language models are even weirder than they thought. But they also now have a clearer sense than ever of what these models can and can’t do—and what’s going on under the hood when they do unexpected things. Step inside the strange new science of understanding LLMs. —Will Douglas Heaven 大语言模型（LLM）是如此庞大而复杂的系统，以至于没有人能完全理解它们是什么、如何运作，或者它们真正能做什么——即使是构建它们的人也不例外。这使得人们很难掌控它们的“幻觉”、设置有效的护栏，以及判断何时该信任它们。因此，研究人员正在尝试一种不同的方法：像研究生物学或神经科学一样，将它们视为巨大的生物——即出现在我们中间的城市规模的异形。他们发现，大语言模型比他们想象的还要怪异。但他们现在也比以往任何时候都更清楚这些模型能做什么、不能做什么，以及当它们做出意想不到的行为时，其内部究竟发生了什么。走进理解大语言模型的奇妙新科学。——Will Douglas Heaven

← 返回新闻列表 / Back to news list