《对话》下一个前沿不是人工智能，而是“人工社会” [C1]
共 2 集
剧集目录
| 集 | 标题 | 时长 |
|---|---|---|
| P1 | 《对话》下一个前沿不是人工智能，而是“人工社会” [C1] | 7:09 |
| P2 | 《对话》下一个前沿不是人工智能，而是“人工社会” [C1] | 14:08 |
P1 · 《对话》下一个前沿不是人工智能，而是“人工社会” [C1] p01 原声 (P1)
![《对话》下一个前沿不是人工智能，而是“人工社会” [C1] p01 原声 (P1)](https://1zimu.com/subtitle/imgs/20261007/uztzhti2zplp7s09vhz888lo_1791352995747.jpg)
字幕摘录
| 时间 | 英文 | 中文 |
|---|---|---|
| 0:05 | 你在听《对话》 2026年9月10日 Nick Jennings | |
| 0:11 | 英国Loughborough大学副校长兼校长写道: | |
| 0:16 | The next frontier is not artificial intelligence, it's artificial societies. | 接下来的前沿不是人工智能,而是人工社会. |
| 0:22 | There is a temptation to divide the future of AI into two possibilities, utopia or catastrophe. | 将AI的未来分为乌托邦或灾难两种可能性是一种诱惑. |
| 0:29 | Neither extreme is particularly helpful. | 这两种极端都特别有用。 |
| 0:33 | The more interesting possibility is messier and requires a step shift in our thinking, | 更有趣的可能性是 更混乱 并且需要一步一步地改变我们的想法, |
| 0:38 | from artificial intelligence to AI societies. | 从人工智能到人工智能 |
| 0:41 | When most people hear AI, they typically think of chat GPT, | 当大多数人听到AI, 他们通常想到聊天GPT, |
| 0:46 | co-pilot or another conversational system. | 副驾驶或其他谈话系统 |
| 0:49 | You ask a question, that system generates an answer. | 你问一个问题, 这个系统产生一个答案。 |
展开字幕全文（95 条）
| 序号 | 英文 | 中文 |
|---|---|---|
| 1 | 你在听《对话》 2026年9月10日 Nick Jennings | |
| 2 | 英国Loughborough大学副校长兼校长写道: | |
| 3 | The next frontier is not artificial intelligence, it's artificial societies. | 接下来的前沿不是人工智能,而是人工社会. |
| 4 | There is a temptation to divide the future of AI into two possibilities, utopia or catastrophe. | 将AI的未来分为乌托邦或灾难两种可能性是一种诱惑. |
| 5 | Neither extreme is particularly helpful. | 这两种极端都特别有用。 |
| 6 | The more interesting possibility is messier and requires a step shift in our thinking, | 更有趣的可能性是 更混乱 并且需要一步一步地改变我们的想法, |
| 7 | from artificial intelligence to AI societies. | 从人工智能到人工智能 |
| 8 | When most people hear AI, they typically think of chat GPT, | 当大多数人听到AI, 他们通常想到聊天GPT, |
| 9 | co-pilot or another conversational system. | 副驾驶或其他谈话系统 |
| 10 | You ask a question, that system generates an answer. | 你问一个问题, 这个系统产生一个答案。 |
| 11 | But AI is rapidly moving beyond this. | 但AI正在迅速超越这个范围. |
| 12 | Systems can now monitor the world, make decisions, negotiate transactions | 系统现在可以监测世界、作出决定、谈判交易 |
| 13 | and carry out tasks over extended periods of time. | 并长期执行任务。 |
| 14 | AI is no longer just generating an answer, it is doing something about it. | 大赦国际不再只是提出一个答案,它正在做一些事情。 |
| 15 | That points to something much bigger than a better chat bot, | 这说明比更好的聊天机器人要大得多 |
| 16 | a world in which AI agents act on our behalf and increasingly interact with other AI agents. | 一个AI代理代表我们行事,并越来越多地与其他AI代理互动的世界. |
| 17 | An agent perceives what is happening, decides what to do, then takes actions to achieve this goal. | 一个代理人能察觉到正在发生的事情,决定该怎么办,然后采取行动来实现这一目标。 |
| 18 | It might book a journey, monitor a supply chain, | 它可能预定一个旅程, 监测一个供应链, |
| 19 | coordinate a team or manage a household's finances. | 协调团队或管理家庭财务。 |
| 20 | Now imagine not one agent, but millions of them. | 现在想象一下 不是一个特工 而是数百万个特工 |
| 21 | Your AI agent can negotiate a mortgage with your bank's agent, | 你的人工智能经纪人可以和你银行的经纪人谈抵押贷款 |
| 22 | schedule surgery with a hospital's agent | 安排医院的手术 |
| 23 | and rearrange your travel plans by dealing directly with the agents of airlines, hotels and insurers. | 与航空公司、旅馆和保险公司的代理人直接打交道,重新安排旅行计划。 |
| 24 | This future is much closer than it sounds and this should change the questions we are asking about AI. | 这一未来比听起来要近得多,这应该会改变我们所问的关于大赦国际的问题。 |
| 25 | Until now, the tendency has been to focus on how intelligent a single agent might become. | 直到现在,人们一直倾向于关注单一的毒剂的智能化程度。 |
| 26 | The more profound challenge is what happens when millions of them interact with one another at scale. | 更深刻的挑战是,当数百万人进行大规模互动时,会发生什么事。 |
| 27 | The intellectual foundations of today's AI systems were laid long before chat GPT. | 今日AI系统的知识基础早在聊天GPT之前就已经奠定. |
| 28 | For decades, I and other researchers of multi-agent networks have studied how autonomous agents can cooperate, | 几十年来,我和多代理网络的其他研究人员 研究了自主代理如何合作, |
| 29 | coordinate and negotiate when nobody has complete information and nobody controls everything. | 当没有人掌握完整的信息,也没有人控制一切时,协调和谈判。 |
| 30 | The earlier systems that emerged focused on how the distinct AI sub-areas of reasoning, | 早期出现的系统侧重于AI不同的子推理领域, |
| 31 | planning and acting could be combined into an effective goal-oriented agent | B. 规划和行动可以合并为一个面向目标的有效代理人 |
| 32 | and how tens of these agents could communicate and cooperate to solve a common objective. | 以及其中的数十名特工能够如何沟通与合作以解决一个共同目标。 |
| 33 | As these interactions became more complex and involved more agents, | 随着这些互动变得更加复杂和涉及更多的代理人, |
| 34 | there was a shift from cooperation between agents that all belonged to a single organization | 各机构之间的合作发生了转变,它们都属于一个组织。 |
| 35 | to agents with different owners and sometimes competing aims. | 给有不同主人的代理人,有时还有相互竞争的目的。 |
| 36 | This focused attention on building algorithms that could form agent teams, automate negotiation | 这让注意力集中在构建可以组成代理团队的算法,自动谈判 |
| 37 | and determine agent trustworthiness. | 并且确定代理人的可信度。 |
| 38 | Today, the pieces needed to build large-scale multi-agent AI systems are falling into place. | 今天,建设大规模多代理AI系统所需的棋子正在落地. |
| 39 | Modern AI agents can call software tools, access information, write and execute code, | 现代AI代理可以调用软件工具,访问信息,写和执行代码, |
| 40 | communicate with other systems and operate for extended periods. | 与其他系统通信并长期运作。 |
| 41 | Consider a supply chain. | 考虑一个供应链。 |
| 42 | One AI agent could represent a manufacturer trying to secure components, | 一个人工智能代理可以代表一个制造商试图保护部件, |
| 43 | another a supplier trying to maximize its revenue. | 另一个供应商试图最大限度地增加收入。 |
| 44 | Yet more could manage transport, inventory and warehouses. | 然而,还可以管理更多的运输、库存和仓库。 |
| 45 | Each agent might be doing exactly what it is designed to do. | 每位特工可能都在做自己要做的事 |
| 46 | But the important question is whether the system they create behaves sensibly. | 但重要的问题是,他们创造的系统是否明智。 |
| 47 | This shift offers enormous potential benefits but also increases the risks. | 这种转变带来了巨大的潜在好处,但也增加了风险。 |
| 48 | 在最近涉及OpenAI和技术平台Hugging Face的实验中, | |
| 49 | thousands of collaborating agents exchanged tens of thousands of messages | 数千名合作人员交换了数万条信息 |
| 50 | and were able to get around the deliberately weakened security controls designed to contain them. | 并得以绕过蓄意削弱的旨在控制这些装置的安全控制。 |
| 51 | The details of one experiment matter less than the broader warning. | 一个实验的细节比更广泛的警告要少。 |
| 52 | When AI systems interact, the behavior of the collective can be harder to predict | 当AI系统相互作用时,集体的行为可能更难预测 |
| 53 | than the behavior of any individual system. | 而不是任何个体系统的行为。 |
| 54 | That should make us cautious but not cause us to down tools. | 这应当使我们谨慎,但不会使我们放弃工具。 |
| 55 | Instead, we need to shift our mindset from building intelligent machines | 相反,我们需要转变我们的心态 从建设智能机器 |
| 56 | to building intelligent societies. | 建立智慧社会 |
| 57 | Once agents can cooperate, compete and resolve conflicts with one another, | 一旦特工可以合作 互相竞争和解决冲突 |
| 58 | we are no longer dealing with isolated machines, we are dealing with a society. | 我们不再处理孤立的机器, 我们处理一个社会。 |
| 59 | Thus, the next frontier is not artificial intelligence, it is artificial societies. | 因此,下一个前沿不是人工智能,而是人工社会. |
| 60 | We already know that intelligence alone does not make a society work. | 我们已经知道,光靠智慧不能使社会发挥作用。 |
| 61 | Human societies depend on rules, institutions, incentives, norms and mechanisms | 人类社会依赖规则、体制、奖励措施、规范和机制 |
| 62 | for resolving disagreements. | 解决分歧。 |
| 63 | AI societies will need their equivalents. | AI社会需要它们的同等条件。 |
| 64 | Who is responsible when two agents make a bad decision? | 当两个特工做出错误的决定时,谁负责呢? |
| 65 | What happens when the interests of different agents conflict? | 不同代理人的利益发生冲突时会怎样? |
| 66 | Who sets the rules? | 谁制定规则? |
| 67 | And who has the power to change them? | 谁能改变它们? |
| 68 | These are not just technical issues. | 这些不仅仅是技术问题。 |
| 69 | They are questions about economics, law, politics and society. | 它们涉及经济、法律、政治和社会。 |
| 70 | They also point to an important role for humans. | 它们还指出了人类的重要作用。 |
| 71 | The most useful future is unlikely to be one in which AI simply replaces people. | 最有用的未来不可能是AI简单地取代人的未来. |
| 72 | While replacement will undoubtedly happen in some cases, | 虽然在某些情况下无疑会更换, |
| 73 | I believe a more common scenario will involve people and agents working together, | 我相信一个更常见的情况 将涉及人和特工合作, |
| 74 | with each doing what it does best. | 每个人尽其所能 |
| 75 | Humans bring judgement, experience, values, contextual understanding and accountability. | 人类带来判断、经验、价值观、背景理解和问责制。 |
| 76 | Agents bring speed, persistence, scale and the ability to process enormous amounts of information. | 特工带来速度、持久性、规模和处理大量信息的能力。 |
| 77 | The goal should not be to create machines that make humans irrelevant. | 目标不应是创造机器,使人类变得无关紧要. |
| 78 | It should be to create systems in which humans and machines | 应该是创造人类和机器的系统 |
| 79 | can achieve things neither can achieve alone. | 无法单独实现的东西。 |
| 80 | But such a future requires more than just better AI models. | 但这样的未来需要的不仅仅是更好的AI模型. |
| 81 | It needs trust and transparency about what agents are doing, | 它需要信任和透明地了解代理人正在做什么, |
| 82 | strong privacy protections and clear lines of accountability. | 强有力的隐私保护和明确的问责制。 |
| 83 | It will also require societies and governments to decide how these systems should be regulated | 它还要求社会和政府决定如何管理这些系统。 |
| 84 | when the most important behaviour may emerge not from one AI developer | 当最重要的行为不是来自一个AI开发者时 |
| 85 | but from interactions between systems built by many different organisations. | 但来自许多不同组织建立的系统之间的相互作用。 |
| 86 | AI's past decade has been defined by a race to build smarter systems. | AI的过去10年被一个建立更智能系统的种族所定义. |
| 87 | I believe the next decade will be defined by a different challenge, | 我认为下一个十年将受到不同的挑战, |
| 88 | ensuring that millions of autonomous systems can work together safely, fairly and effectively. | 确保数百万个自主系统能够安全、公平和有效地合作。 |
| 89 | The future of AI will not be determined solely by the intelligence of individual agents. | AI的未来不会完全取决于个别特工的情报. |
| 90 | It will be determined by the societies they create. | 它将由它们创造的社会决定。 |
| 91 | And societies, as humans know all too well, | 社会,正如人类所知道的, |
| 92 | are much harder to govern than individuals. | 比个人更难治理 |
| 93 | You are listening to The Conversation where Nick Jennings writes, | 你在听尼克・詹宁斯写的谈话 |
| 94 | The next frontier is not artificial intelligence, it's artificial societies. | 接下来的前沿不是人工智能,而是人工社会. |
| 95 | This article was published on 10th September 2026 and was read by Martin Buchanan for NOAA. | 这篇文章发表于2026年9月10日,由马丁·布坎南为NOAA阅读. |
该视频共有字幕 95 条。解锁更多字幕为会员功能，请移动到 价格
P2 · 《对话》下一个前沿不是人工智能，而是“人工社会” [C1] p02 原声+中配 (P2)
字幕摘录
| 时间 | 英文 | 中文 |
|---|---|---|
| 0:05 | 你在听《对话》 2026年9月10日 Nick Jennings | |
| 0:11 | 英国Loughborough大学副校长兼校长写道: | |
| 0:16 | The next frontier is not artificial intelligence, it's artificial society. | 下一个前沿不是人工智能,而是人工社会. |
| 0:21 | 你在听《对话》 2026年9月10日 Nick Jennings | |
| 0:28 | 英国Loughborough大学副校长兼校长写道: | |
| 0:29 | The next frontier is not artificial intelligence, it's artificial society. | 下一个前沿不是人工智能,而是人工社会. |
| 0:35 | There is a temptation to divide the future of AI into two possibilities, utopia or catastrophe. | 将AI的未来分为乌托邦或灾难两种可能性是一种诱惑. |
| 0:42 | Neither extreme is particularly helpful. | 这两种极端都特别有用。 |
| 0:45 | It's easy to divide the future of artificial intelligence into two possibilities. | 将人工智能的未来分为两种可能性很容易. |
| 0:55 | The more interesting possibility is messier and requires a step shift in our thinking from artificial intelligence to AI societies. | 更有趣的可能性是更混乱,需要我们的思想从人工智能向AI社会逐步转变. |
展开字幕全文（92 条）
| 序号 | 英文 | 中文 |
|---|---|---|
| 1 | 你在听《对话》 2026年9月10日 Nick Jennings | |
| 2 | 英国Loughborough大学副校长兼校长写道: | |
| 3 | The next frontier is not artificial intelligence, it's artificial society. | 下一个前沿不是人工智能,而是人工社会. |
| 4 | 你在听《对话》 2026年9月10日 Nick Jennings | |
| 5 | 英国Loughborough大学副校长兼校长写道: | |
| 6 | The next frontier is not artificial intelligence, it's artificial society. | 下一个前沿不是人工智能,而是人工社会. |
| 7 | There is a temptation to divide the future of AI into two possibilities, utopia or catastrophe. | 将AI的未来分为乌托邦或灾难两种可能性是一种诱惑. |
| 8 | Neither extreme is particularly helpful. | 这两种极端都特别有用。 |
| 9 | It's easy to divide the future of artificial intelligence into two possibilities. | 将人工智能的未来分为两种可能性很容易. |
| 10 | The more interesting possibility is messier and requires a step shift in our thinking from artificial intelligence to AI societies. | 更有趣的可能性是更混乱,需要我们的思想从人工智能向AI社会逐步转变. |
| 11 | When most people hear AI, they typically think of chat GPT, co-pilot or another conversational system. | 当大多数人听到AI时,他们一般会想到聊天GPT,副驾驶或者另一个对话系统. |
| 12 | You ask a question, that system generates an answer. | 你问一个问题, 这个系统产生一个答案。 |
| 13 | But AI is rapidly moving beyond this. | 但AI正在迅速超越这个范围. |
| 14 | Systems can now monitor the world, make decisions, negotiate transactions and carry out tasks over extended periods of time. | 系统现在可以监测世界,作出决定,谈判交易,并在很长一段时间内执行任务. |
| 15 | AI is no longer just generating an answer, it is doing something about it. | 大赦国际不再只是提出一个答案,它正在做一些事情。 |
| 16 | That points to something much bigger than a better chatbot, | 这说明比一个更好的聊天人要大得多 |
| 17 | a world in which AI agents act on our behalf and increasingly interact with other AI agents. | 一个AI代理代表我们行事,并越来越多地与其他AI代理互动的世界. |
| 18 | That points to something much bigger than a better chatbot, | 这说明比一个更好的聊天人要大得多 |
| 19 | a world in which AI agents act on our behalf and increasingly interact with other AI agents. | 一个AI代理代表我们行事,并越来越多地与其他AI代理互动的世界. |
| 20 | An agent perceives what is happening, decides what to do, then takes actions to achieve this goal. | 一个代理人能察觉到正在发生的事情,决定该怎么办,然后采取行动来实现这一目标。 |
| 21 | Coordinate a team or manage a household's finances. | 协调一个团队或管理一个家庭的财务。 |
| 22 | Now imagine not one agent, but millions of them. | 现在想象一下 不是一个特工 而是数百万个特工 |
| 23 | Your AI agent can negotiate a mortgage with your bank's agent, schedule surgery with a hospital's agent, | 你的人工智能探员可以和你银行的经纪人谈房贷 安排医院的经纪人做手术 |
| 24 | and rearrange your travel plans by dealing directly with the agents of airlines, hotels and insurers. | 与航空公司、旅馆和保险公司的代理人直接打交道,重新安排旅行计划。 |
| 25 | This future is much closer than it sounds, and this should change the questions we are asking about AI. | 这个未来比听起来要近得多,这应该能改变我们所问的AI的问题. |
| 26 | Until now, the tendency has been to focus on how intelligent a single agent might become. | 直到现在,人们一直倾向于关注单一的毒剂的智能化程度。 |
| 27 | The more profound challenge is what happens when millions of them interact with one another at scale. | 更深刻的挑战是,当数百万人进行大规模互动时,会发生什么事。 |
| 28 | The theoretical foundation of today's AI system has been established long before chatgpt. | 今日AI系统的理论基础早在聊天前就已经建立. |
| 29 | For decades, I have been discussing with researchers on the multi-intelligent network. | 几十年来,我一直在与多智能网络的研究人员进行讨论。 |
| 30 | In a situation where information is incomplete and no one can control everything, how can a self-controlled intelligent body achieve cooperation, coordination and communication? | 在信息不完整,无人能控制一切的情况下,自我控制的智能体如何实现合作,协调和交流? |
| 31 | The earlier systems that emerged focused on how the distinct AI sub-areas of reasoning, planning and acting could be combined into an effective goal-oriented agent, | 早期出现的系统侧重于如何将AI的推理、规划和行动等不同的分领域合并为一个面向目标的有效代理人, |
| 32 | and how tens of these agents could communicate and cooperate to solve a common objective. | 以及其中的数十名特工能够如何沟通与合作以解决一个共同目标。 |
| 33 | As these interactions became more complex and involved more agents, there was a shift from cooperation between agents, | 随着这些互动变得更加复杂,涉及更多的代理人,从代理人之间的合作发生了转变, |
| 34 | that all belonged to a single organization, to agents with different owners and sometimes competing aims. | 它们都属于一个单一组织,属于具有不同所有人和有时相互竞争目的的代理人。 |
| 35 | As these interactions became more complex and involved more and more agents, | 随着这些互动变得更加复杂,并越来越多地涉及到代理人, |
| 36 | cooperation also shifted from agents that all belonged to a single organization, | 合作也从所有行为者都属于一个组织, |
| 37 | to agents with different owners and sometimes competing aims. | 给有不同主人的代理人,有时还有相互竞争的目的。 |
| 38 | This focused attention on building algorithms that could form agent teams, automate negotiation and determine agent trustworthiness. | 这集中了对构建算法的关注,这些算法可以组成代理团队,使谈判自动化,并确定代理的可信赖性. |
| 39 | Today, the pieces needed to build large-scale multi-agent AI systems are falling into place. | 今天,建设大规模多代理AI系统所需的棋子正在落地. |
| 40 | Modern AI agents can call software tools, access information, write and execute code, communicate with other systems and operate for extended periods. | 现代AI代理可以调用软件工具,访问信息,编写和执行代码,与其他系统通信并长时间运行. |
| 41 | Consider a supply chain. | 考虑一个供应链。 |
| 42 | One AI agent could represent a manufacturer trying to secure components. | 一个AI代理可以代表一个试图保护部件的制造商. |
| 43 | Another, a supplier trying to maximize its revenue. | 另一个是供应商试图最大限度地增加收入。 |
| 44 | Yet more could manage transport, inventory and warehouses. | 然而,还可以管理更多的运输、库存和仓库。 |
| 45 | Each agent might be doing exactly what it is designed to do. | 每位特工可能都在做自己要做的事 |
| 46 | But the important question is whether the system they create behaves sensibly. | 但重要的问题是,他们创造的系统是否明智。 |
| 47 | This shift offers enormous potential benefits, but also increases the risks. | 这一转变带来了巨大的潜在好处,但也增加了风险。 |
| 48 | 在最近涉及OpenAI和技术平台Hugging Face的实验中, | |
| 49 | thousands of collaborating agents exchanged tens of thousands of messages | 数千名合作人员交换了数万条信息 |
| 50 | and were able to get around the deliberately weakened security controls designed to contain them. | 并得以绕过蓄意削弱的旨在控制这些装置的安全控制。 |
| 51 | 在最近由OpenAI和技术平台Hugging Face进行的实验中, | |
| 52 | thousands of smart bodies that worked together exchanged tens of thousands of messages | 数以千计的聪明人一起工作 交换了数万条信息 |
| 53 | and were able to get around the deliberately weakened security controls designed to contain them. | 并得以绕过蓄意削弱的旨在控制这些装置的安全控制。 |
| 54 | One of the details of the experiment is far less important than the extensive warning behind it. | 实验的细节之一远不如背后的广泛警告重要. |
| 55 | When AI systems interact with each other, | 当AI系统互相影响时, |
| 56 | the behavior of the entire system group is often harder to predict than the behavior of any single system. | 整个系统组的行为往往比任何单一系统的行为更难预测. |
| 57 | This should keep us cautious, but we don't have to give up our efforts. | 这应该让我们保持谨慎,但我们不必放弃努力. |
| 58 | On the contrary, we need to transform the thinking model from creating intelligent machines to building a smart society. | 相反,我们需要将思维模式从创造智能机器转变为建设智能社会. |
| 59 | We are no longer dealing with isolated machines, we are dealing with a society. | 我们不再处理孤立的机器,而是处理一个社会。 |
| 60 | Thus, the next frontier is not artificial intelligence, it is artificial societies. | 因此,下一个前沿不是人工智能,而是人工社会. |
| 61 | We already know that intelligence alone does not make a society work. | 我们已经知道,光靠智慧不能使社会发挥作用。 |
| 62 | Human societies depend on rules, institutions, incentives, norms, and mechanisms for resolving disagreements. | 人类社会依赖规则、体制、激励措施、规范和机制来解决分歧。 |
| 63 | AI societies will need their equivalents. | AI社会需要它们的同等条件。 |
| 64 | Who is responsible when two agents make a bad decision? | 当两个特工做出错误的决定时,谁负责呢? |
| 65 | What happens when the interests of different agents conflict? | 不同代理人的利益发生冲突时会怎样? |
| 66 | Who sets the rules? | 谁制定规则? |
| 67 | And who has the power to change them? | 谁能改变它们? |
| 68 | These are not just technical issues, they are questions about economics, law, politics, and society. | 这些不仅仅是技术问题,而是经济、法律、政治和社会问题。 |
| 69 | They also point to an important role for humans. | 它们还指出了人类的重要作用。 |
| 70 | The most useful future is unlikely to be one in which AI simply replaces people. | 最有用的未来不可能是AI简单地取代人的未来. |
| 71 | While replacement will undoubtedly happen in some cases, | 虽然在某些情况下无疑会更换, |
| 72 | I believe a more common scenario will involve people and agents working together, with each doing what it does best. | 我认为,更常见的情况是,人们和代理人将携手合作,各尽所能。 |
| 73 | Humans bring judgment, experience, values, contextual understanding, and accountability. | 人类带来判断、经验、价值观、背景理解和问责制。 |
| 74 | Agents bring speed, persistence, scale, and the ability to process enormous amounts of information. | 特工带来了速度、持久性、规模和处理大量信息的能力。 |
| 75 | The goal should not be to create machines that make humans irrelevant. | 目标不应是创造机器,使人类变得无关紧要. |
| 76 | It should be to create systems in which humans and machines can achieve things neither can achieve alone. | 应该是建立各种系统,使人类和机器能够实现无法单独实现的东西。 |
| 77 | But such a future requires more than just better AI models. | 但这样的未来需要的不仅仅是更好的AI模型. |
| 78 | It needs trust and transparency about what agents are doing, strong privacy protections, and clear lines of accountability. | 它需要信任和透明地了解代理人正在做什么,需要强有力的隐私保护,以及明确的问责制。 |
| 79 | It will also require societies and governments to decide how these systems should be regulated | 它还要求社会和政府决定如何管理这些系统。 |
| 80 | when the most important behavior may emerge not from one AI developer, | 当最重要的行为可能 不从一个AI开发者, |
| 81 | but from interactions between systems built by many different organizations. | 但来自许多不同组织建立的系统之间的相互作用。 |
| 82 | The most important behavior may not emerge from one AI developer, | 一个AI开发者可能不会出现最重要的行为, |
| 83 | but from interactions between systems built by many different organizations. | 但来自许多不同组织建立的系统之间的相互作用。 |
| 84 | Societies and governments also need to decide how these systems should be regulated. | 社会和政府也需要决定如何管理这些系统。 |
| 85 | In the past decade, the core trend in the field of AI is to create smarter systems in a closed environment. | 在过去十年中,AI领域的核心趋势是在封闭的环境下建立更聪明的系统. |
| 86 | can work together safely, fairly, and effectively. | 能够安全、公平和有效地合作。 |
| 87 | The future of AI will not be determined solely by the intelligence of individual agents. | AI的未来不会完全取决于个别特工的情报. |
| 88 | It will be determined by the societies they create. | 它将由它们创造的社会决定。 |
| 89 | And societies, as humans know all too well, are much harder to govern than individuals. | 人类非常清楚,社会比个人更难治理。 |
| 90 | You are listening to The Conversation, where Nick Jennings writes, | 你在听尼克·詹宁斯写的《对话》 |
| 91 | The next frontier is not artificial intelligence, it's artificial societies. | 接下来的前沿不是人工智能,而是人工社会. |
| 92 | This article was published on the 10th of September, 2026, and was read by Martin Buchanan for NOAA. | 这篇文章发表于2026年9月10日,由马丁·布坎南为NOAA阅读. |
该视频共有字幕 105 条。解锁更多字幕为会员功能，请移动到 价格