推荐微信扫码登录以获得更好的体验!另:🎉建议更新到最新版,以体验新功能!🎉 版本更新记录→

《经济学人》前美国国安高官:AI怀疑论者看错了威胁,它其实更

共 1 集

剧集目录

集标题时长
P1《经济学人》前美国国安高官:AI怀疑论者看错了威胁,它其实更8:04

P1 · 《经济学人》前美国国安高官:AI怀疑论者看错了威胁,它其实更能保卫脆弱的现状 [C1] p01 原声 (P1)

字幕摘录

时间英文中文
0:07Anne Neuberger安妮·诺伊伯格
0:09AI sceptics overlook how vulnerable the status quo is and how much the technology can help defend it.AI怀疑论者忽略了现状的脆弱程度以及技术能帮助捍卫现状的程度.
0:17Anne Neuberger is a general partner and the head of global affairs at Andreessen Horowitz.Anne Neuberger是安德烈森·霍罗维茨的总合伙人和全球事务负责人.
0:23She previously served as Deputy National Security Advisor of the United States.她此前曾担任美国国家安全副顾问.
0:29Much of the debate about artificial intelligence, including a recent leader in The Economist,许多关于人工智能的辩论, 包括最近一位《经济学家》的领袖,
0:35asks how to prevent a powerful new technology from making the world more dangerous.问如何防止强大的新技术使世界更加危险。
0:41There is an equally important, more urgent question.有一个同样重要、更紧迫的问题。
0:46How quickly can we use it to make an already dangerous world safer?我们如何能利用它使已经危险的世界更安全?
0:52The world that large language models entered was very vulnerable.大语言模型进入的世界非常脆弱。
0:57In 2017, NotPetya, a Russian cyber attack aimed at Ukraine,2017年,NotPetya,俄罗斯网络攻击针对乌克兰,
展开字幕全文(90 条)
序号英文中文
1Anne Neuberger安妮·诺伊伯格
2AI sceptics overlook how vulnerable the status quo is and how much the technology can help defend it.AI怀疑论者忽略了现状的脆弱程度以及技术能帮助捍卫现状的程度.
3Anne Neuberger is a general partner and the head of global affairs at Andreessen Horowitz.Anne Neuberger是安德烈森·霍罗维茨的总合伙人和全球事务负责人.
4She previously served as Deputy National Security Advisor of the United States.她此前曾担任美国国家安全副顾问.
5Much of the debate about artificial intelligence, including a recent leader in The Economist,许多关于人工智能的辩论, 包括最近一位《经济学家》的领袖,
6asks how to prevent a powerful new technology from making the world more dangerous.问如何防止强大的新技术使世界更加危险。
7There is an equally important, more urgent question.有一个同样重要、更紧迫的问题。
8How quickly can we use it to make an already dangerous world safer?我们如何能利用它使已经危险的世界更安全?
9The world that large language models entered was very vulnerable.大语言模型进入的世界非常脆弱。
10In 2017, NotPetya, a Russian cyber attack aimed at Ukraine,2017年,NotPetya,俄罗斯网络攻击针对乌克兰,
11spread worldwide within hours, deleting data and paralyzing systems,在数小时内向全世界传播,删除数据和瘫痪系统,
12including those of Maersk, which carries about a fifth of global shipping.包括梅尔斯克的船,它承载着大约五分之一的全球航运。
13In 2023, Iranian-linked hackers broke into American water utilities,2023年,与伊朗有联系的黑客闯入美国供水公司,
14forcing operators to switch to manual control.强制操作员切换到手动控制.
15More worrying still are the attacks not yet triggered.更令人担忧的是,袭击尚未启动。
16Chinese state hackers have planted footholds in American power, water and transport systems,中国的州黑客在美国的电力、水和交通系统 扎下了根基
17ready to strike silently during a conflict.准备在冲突期间无声攻击。
18Few of these attacks relied on brilliant new exploits,这些攻击中很少有依靠 辉煌的新利用,
19code that weaponizes systems flaws to cause harm.将系统缺陷武器化以造成伤害的代码。
20They succeeded because the defenders could not find, fix or fence off他们成功是因为维权者找不到、修补或封锁
21known weaknesses fast enough across sprawling aging systems.已知的弱点 足够快地跨过扩展的老化系统。
22The cost of that status quo belongs in any honest accounting of AI risk.维持现状的代价在于对AI风险的任何诚实的核算。
23I have worked on both sides of cyber conflict,我为网络冲突双方工作过
24conducting intelligence operations at America's National Security Agency,在美国国家安全局进行情报活动,
25then running its cyber security directorate to turn those lessons into defenses.然后管理网络安全局 把这些教训变成防御
26Later at the White House, I pushed through the first mandatory cyber security rules后来在白宫 我推了第一个强制性的网络安全规则
27for America's pipelines, railways and airports.为了美国的管道,铁路和机场。
28The experience taught me something important.经历教会了我一些重要的事情.
29AI can help defenders even more than it can help attackers.大赦国际能够帮助维权者比帮助攻击者还要多。
30True, AI makes the challenge harder in some ways.的确,大赦国际在某些方面使挑战变得更加困难。
31It lowers the skill needed by attackers它降低了攻击者所需的技能
32and lets them jiggle every digital doorknob continuously and at scale.让他们不停地和大规模地 抖动每一个数字门把手
33It is already speeding up the discovery of new flaws,它已经在加速发现新的缺陷,
34including in software foundational to the Internet.包括互联网基础软件。
35But the old rule of cyber conflict, that defenders must be strong everywhere但网络冲突的旧规则 维权者必须强壮
36while attackers need to find only one weakness,攻击者只需要找到一个弱点
37rested on defenders lacking the time, people and money to cover everything.维权者没有时间、人和金钱来应付一切。
38AI alleviates that problem.大赦国际缓解了这一问题。
39AI is already making defenders' response to these threats stronger.大赦国际已经在加强维权者对这些威胁的反应。
40They own their code and can task AI to preemptively analyze it line by line他们拥有自己的代码,可以让AI先发制人地逐行分析
41from all possible angles of attack.从所有可能的角度攻击。
42Rather than only probing it from the outside, as attackers must do.而不是像攻击者必须做的那样,只从外部探测。
43They see their own network traffic,他们看到自己的网络流量,
44so AI can spot the subtle signs of an intruder using stolen credentials这样AI就能发现入侵者的微妙迹象 用偷来的证件
45and legitimate tools, as China's hackers do, which older tools miss.和合法工具,像中国的黑客所做的, 旧的工具错过了。
46Using AI available today, utilities can turn operational data使用今天可用的AI,公用设施可以转换操作数据
47into virtual replicas of power grids and water systems.变成电网和水系的虚拟复制品。
48These digital twins let operators test defenses这些数码双胞胎让操作员测试防御
49without putting essential services at risk.而不危及基本服务。
50Attackers working from outside can build only crude approximations.从外部工作的攻击者只能建立粗糙的近似.
51Crutely, AI can fix flaws after finding them.谨慎地说,AI在找到缺陷后可以弥补缺陷.
52In DARPA's AI Cyber Challenge last year,去年在DARPA的AI网络挑战中,
53AI systems scanning open-source software of the kind that underpins critical infrastructureAI系统扫描支持关键基础设施的开源软件
54patched 43 planted vulnerabilities and uncovered 18 previously unknown flaws,修复了43个埋设的弱点,并发现了18个以前未知的缺陷,
55fixing 11 of them.修理11个
56For aging equipment that cannot safely be patched,对于无法安全补齐的老旧设备,
57AI can give each flaw a specific workaround,AI可以给每个缺陷一个具体的工作,
58such as a firewall rule or a settings change,例如防火墙规则或设置更改,
59tested on a digital twin before it goes live.测试在数字双胞胎开始前。
60And AI is cheap enough that small, understaffed utilities can use it too.而AI也足够便宜,小的,人手不足的公用事业也可以使用.
61None of this happens by default.这一切都不是默认发生的。
62AI favors defenders only if they move with urgency.只有在维权者急于行动时,大赦国际才给予他们优惠。
63Government should start with the infrastructure operators and shared suppliers,政府应从基础设施运营商和共享供应商开始,
64whose failure would cause the greatest disruption,他们的失败将造成最大的破坏,
65enlisting AI labs to find flaws in their code or firmware招募人工智能实验室寻找其代码或固件中的缺陷
66and produce fixes by set deadlines.并按规定的期限作出修正。
67Government should share the cost for smaller operators.政府应分担较小经营者的费用。
68Success should be measured by how quickly threats are contained成功与否应当以控制威胁的速度来衡量
69and whether essential services survive an attack.以及基本服务是否在攻击中幸存。
70Critical infrastructure operators could address the highest priority systems within a year,关键的基础设施运营商可在一年内处理最优先的系统,
71judging by how quickly they met the cybersecurity standards issued by the Biden administration.从他们达到拜登政府发布的网络安全标准的速度来看
72Findings should be shared with trusted partners using the coordinated disclosure practices利用协调的披露做法,与信任的伙伴分享调查结果
73that keep fixes from tipping off attackers before systems are patched.在系统被补丁之前 避免向攻击者告密
74The aim should be for a discovery in one country to strengthen every country running the same technology.目标应该是在一个国家进行探索,以加强每一个使用相同技术的国家。
75Such practical partnerships are more urgent than AI treaties, and the foundations exist.这种实际的伙伴关系比大赦国际的条约更为紧迫,而且存在各种基础。
76NATO's annual Locked Shields exercise brings together thousands of participants from dozens of countries北约每年的锁盾演习 聚集了来自数十个国家的数千名参与者
77to defend power grids, water systems and other networks against live attacks.以防御电网、供水系统和其他网络的现场攻击。
78The exercises could be expanded to include using AI to generate patches and controls,这些练习可以扩大,包括利用AI生成补丁和控制,
79sharing the results across the alliance.分享整个联盟的成果。
80Here, too, there is precedent.这里也有先例。
81In 2021, America formed the International Counter Ransomware Initiative, now comprising nearly 70 countries.2021年,美国组建了国际反兰素软件倡议,现由近70个国家组成.
82Members share threat warnings and forensic techniques,成员分享威胁警告和法医技术,
83including how to trace ransoms laundered through the crypto ecosystem.包括如何追踪通过密码生态系统清洗的赎金。
84They coordinate policy, for example by pledging not to pay ransoms.它们协调政策,例如保证不支付赎金。
85Alongside law enforcement takedowns of gangs and their money laundering networks,在执法部门打击帮派及其洗钱网络的同时,
86the effort has squeezed the business.努力挤压了企业。
87After peaking at over $1 billion in 2023, ransom payments have fallen each year since.在2023年达到顶峰,超过10亿美元之后,赎金支付自此逐年下降.
88The time has come to use AI as aggressively on defence as adversaries will use it in attack.现在是时候用AI来防御了 因为对手会用它来攻击
89A serious international agenda for AI safety should explain not only how it will contain the risks of a new technology,一项严肃的AI安全国际议程不仅应解释它将如何遏制新技术的风险,
90but how it will keep the lights on for the world's citizens, and how soon.但它将如何保持 灯为世界的公民, 以及多久。
该视频共有字幕 90 条。解锁更多字幕为会员功能,请移动到 价格