AI 智能体可能颠覆银行体系——对话伊桑·莫利克

即使是 GPT-4 这样早已过时的模型，提供给肯尼亚的创业者后，也已经表现不错——仅仅从 AI 获取建议，就实现了 18% 的盈利。我们现在有一个工具，可以为那些无法获得建议的人提供相当不错的、针对他们情况定制的建议。我认为我们常常从我们这些资源丰富、人脉广泛的人的狭隘视角出发，考虑如何用 AI 做事来平衡我们对娱乐的追求。这里有成千上万个小小的用例，智能曾是让世界变得更美好的瓶颈。它们需要一些努力，需要测试，需要保障措施。但这很令人兴奋。

Even GPT-4, a long-obsolete model given to Kenyan entrepreneurs, is already doing well—18% profitability just from getting advice from AI for entrepreneurs. We have a tool now that provides reasonably good advice to people who can't get it, customized for their condition. I think we tend to think a lot about the narrow viewpoint of those of us who are well-connected and well-resourced and balancing out our desire for entertainment with AI doing things. There are a thousand million little use cases here where intelligence was the bottleneck to making the world a better place. And they require some effort and they require testing and they require safeguards. But this is exciting.

Ethan，这期播客是在哪里找到你的？

Ethan, where does this podcast find you?

我的书今天发布，所以我实际上在斯坦福，要在这里做几场演讲，然后明天去 Master of Scale。所以，尽管我是东海岸人，但发布 AI 书籍还是得在西海岸。

My book launches today, so I'm actually at Stanford to give a bunch of talks here and then Master of Scale tomorrow. So the West Coast is where you kind of have to launch an AI book even though I'm an East Coast guy.

那我们直接开始吧。你上一本书《共同智能》于 2024 年出版，你说它是关于聊天机器人世界的。你的新书《共存》，AI 的下一个阶段，今天出版，就在我们录制的时候，讲的是与有时但不总是比你更聪明的 AI 合作。你如何描述一种似乎每分钟都在进化的技术？你希望两年后哪些东西仍然成立？

So let's bust right into it. Your last book, Co-Intelligence, came out in 2024, and you said it was written about a world of chatbots. Your new book, Coexistence, the next phase of AI, out today, as we record, is about working with AI that is sometimes, but not always, better than you. How do you write about a technology that seems to be evolving by the minute? What do you still hope still holds two years from now?

我的意思是，有一些东西你可以把握，我认为有三种形态。其中之一是指数曲线，对吧？AI 正在变得更好，我们处于指数增长中，这种进步体现在许多不同方面。第二个是锯齿状前沿，这是我们在与哈佛、MIT 和华威大学的朋友们的研究中创造的一个术语，我们做了一些实验。我们发现 AI 擅长一些你意想不到的事情，不擅长一些你意想不到的事情，你无法预测什么好什么坏。这给人类留下了很多空间，对吧？你无法提前预测好坏。最后是悬垂效应。AI 能做的事情和我们实际用 AI 做的事情之间存在很大差距，部分原因是人类社会和人类系统变化缓慢。所以如果你记住这三点，即使事物在变化，你实际上可以对现在和近期未来有很多说法。

I mean there's a few things that you can kind of hold on to and I think there's kind of three shapes I kind of think about. One of those is the exponential curve, right? AI is getting better and we're on an exponential and that better shows up in lots of different ways. The second is that is the jagged frontier, a term we coined in our research with my friends at Harvard, MIT and University of Warwick where we did some experiments. We found that AI is good at some stuff you wouldn't expect, bad at some stuff you wouldn't expect and you can't predict what's good or bad. And that leaves lots of room for humans, right? You can't predict what's good or bad in advance. And then the final is this idea of an overhang. There's a big gap between what AI can do and what we actually are doing with AI partially because human society and human systems change slowly. So if you have those three things in mind, there's actually a lot you can say about the present and near future even as things change.

对于那些对 AI 的认识来自使用免费版 ChatGPT 或 Claude 的人来说，自 2024 年以来有什么变化？

For someone whose idea of AI comes from reusing free versions of ChatGPT or Claude, what's changed since 2024?

所以现在有 180 个模型比 GPT-4 更好。其中 19 个可以在你的手机上运行。所以能力提升了。有一个——当我的书出版时，第一本书，有一个叫 METR 的组织，追踪 AI 在单个提示下以 50% 成功率能完成多少工作。我的书出版时是四分钟。实际的官方测试在四月结束，据我们所知，现在大约是 17 周的工作量。他们并没有在人类工作方面超出提示的范围。所以这些系统的能力发生了相当大的变化，如果你只盯着统计数据，你可能会错过一个巨大的转变。

So there are now 180 models that are better than GPT-4. 19 of them can run on your phone. So capabilities have gone up. There's been a—when my book came out, the first book, there's an organization called METR that tracks how much work the AI can do with a single prompt with a 50% chance of success. That was four minutes when my book came out. The actual official test ran out in April and as far as we can tell it's something like 17 weeks of work at this point. They didn't get out of a prompt in terms of human work. So that's a pretty big change in how capable these systems have gotten and if you sort of stat that out you may be missing a big transformation.

你认为基本上对 Anthropic 和 OpenAI 存在生存威胁吗？那就是如果 Nvidia 开始把 GPU 放进比如 iPhone，它成为客户端，那么是否有机会超越 LLM 并使其过时？

Do you think there's an existential threat to basically Anthropic and OpenAI and that is if Nvidia starts putting GPUs in say an iPhone and it becomes the client and it kind of is there opportunity to leapfrog the LLMs and make them obsolete?

我的意思是，目前领先模型的历史主要是 OpenAI 和 Anthropic 互相竞争，谷歌在过去几年中两次成为领先的模型公司。领先模型和不太先进的模型之间存在很大差距，尤其是在智能体系统领域。这实际上有点反直觉，但这就是为什么我们在商业中有六西格玛系统，如果你做数百个步骤，99.5% 的成功率和 99.9% 的成功率之间有很大区别。最聪明的模型有更高的成功率，差异很小，但在结果上差别巨大。所以你可以在手机上得到相当有能力的模型，但它们远不及你在复杂数学中所需的能力，甚至只是制作一个非常好的视频或什么的，你需要做出许多小选择，你需要更高的准确性，我认为即使我们得到非常有能力的模型，我们也不会在手机上得到前沿模型，所以它可能会改变我们做的事情，但我认为这对前沿模型提供商不是生存威胁。

I mean right now the entire history of the leading models has been mostly OpenAI and Anthropic jockeying with each other with Google appearing twice over the last couple years in terms of being a leading model company. And there's a big gap between leading models and less advanced models especially in the world of agentic systems. And it's actually kind of counterintuitive, but it's the reason why we have six sigma systems in business, which is there's a huge difference between a 99.5% success rate and a 99.9% success rate if you're doing hundreds of steps. And the smartest models have a much higher—have a small difference in higher success rates, but that makes a huge difference in terms of the outcomes you get. So you can get fairly capable models on your phone, but they're nowhere near as good as what you need to be able to do in complex math, even just putting together a really good video or something you need lots of little choices being made and you need that higher accuracy and I don't think we're going to get the frontier models on a phone even if we get very capable models so it might change what we do but I don't think it's an existential threat to frontier model providers.

我用的类比是汽车行业，前沿模型是德国汽车制造商，而开放权重或开源模型更像是——甚至不是日本，而是韩国汽车制造商？你认为市场会分叉，开放权重模型会获得更多份额，但高端品牌——前沿模型——会获得更多收入？这个类比公平吗？

Is it sort of the analogy I would use is the auto industry where the frontier models are German automakers and the open-weight guys or the open source are more like not even Japanese but Korean automakers? Do you think the market's going to bifurcate and that there'll be more share amongst the open-weight but more revenue to the premium brands, the frontier models? Is that a fair analogy?

所以我实际上认为这比那要复杂一点，对吧？所以同时发生了几件事。我们有开放权重模型，任何人都可以修改，这有很多非常有趣的含义。它们继续落后前沿 6 到 8 个月，目前并没有真正缩小差距，我们看看长期会发生什么。中国最终可能会打击开放权重模型，或者美国也会，但目前它们是可用的，它们在成本方面有真正的优势，但只是某种程度上，对吧？你可以修改和拥有它们，这很好，但在成本方面，OpenAI、Anthropic 和谷歌拥有比你多得多的处理器，因此推理成本更低，而且他们有非常好的廉价模型。所以问题是，你这样做是为了主权，为了自己使用这些模型，因为如果是为了节省成本，你通常最好选择这些公司之一，因为他们也在教他们的模型协同工作。所以，Fable，Anthropic 最昂贵和强大的模型，可以在需要时调用 Haiku，他们非常便宜的模型来工作。所以我不确定成本论点是否成立，但你可能还有其他理由将开放权重模型纳入你的系统。

So I actually think that it's a little more complicated than that, right? So a few things are happening at one time. We have open weights models which anyone can modify, there's lots of really interesting implications of that. They continue to be at 6 to 8 months behind the frontier, it's not really closing at this point, we'll see what happens in the long term. It's possible China eventually cracks down on open-weight models or the US does, but for right now they're available and they have a real advantage in terms of cost but only sort of, right? You can modify and own them which is really nice but on the cost side OpenAI and Anthropic and Google have many more processors than you do, they have lower inference costs as a result and they have cheap models that are very good. So the question is are you doing this for sovereignty, for using these models on your own, because if it's cost savings you're often just better off going with one of these companies because they're also teaching their models to work together. So, Fable, the most expensive and powerful Anthropic model, can call on Haiku, their very cheap model, to do work if it needs to. So I'm not sure the cost argument holds up, but there are other reasons you might want to have open-weight models involved in your systems.

那么，对于那些想从 AI 创建个人模式的人，你概述了四个优势：深度知识、广度知识、品味和能动性。你能带我们逐一了解这些优势以及如何培养它们吗？

So, for those looking to create personal modes from AI, you outlined four advantages: deep knowledge, wide knowledge, taste, and agency. Could you walk us through each of these advantages and how someone can develop them?

当然。当然。我认为认识到拥有优势并不等同于一条明线，这非常重要，对吧？我认为人们倾向于认为 AI 永远不会展现创造力或永远不会展现判断力。但这根本不是真的。

Sure. Sure. And I think it's really important to recognize that having an advantage is not the same thing as a bright line, right? I think people tend to think about like AI will never show creativity or never show judgment. And that just isn't true.

你没法在不调动能动性、创造力和判断力的情况下完成一个 17 小时的任务。那可能不是人类的创造力和判断力，但它确实是创造力和判断力。所以我想勾勒的，是那些能让你在与 AI 协作时卓有成效、并始终与他人拉开差距的东西。首先是知识，也就是知道很多事。我觉得现在是文科专业的复兴时刻。如果你懂很多东西，而 AI 掌握着人类所有的文字，那么只要你能唤起某种艺术风格、某种写作风格，或者你从经济学里学到的某个概念，AI 就能把这些应用出来。所以你要懂很多不同领域的东西，对吧？这在你的公司里同样成立，因为现在 AI 让程序员能当设计师、设计师能当程序员。所以懂一点编程或设计非常有帮助。然后是深度知识，也就是专长，它能让你在某些领域胜过 AI，或者至少能核查它、与它协作——而随着 AI 越来越强，我们如何随时间积累这种专长正面临挑战。再然后是品味，也就是从众多选项中做选择。品味是把垃圾变成人类真正做出的选择的东西。最后是能动性。我是创业学教授，核心问题就是如何让人们以创业者的方式行动，如何让他们在没有任何说明书的情况下利用好这些系统。这就是我重点关注的四个方面。

You can't do a 17-hour task without having agency, creativity, judgment all coming into play. It may not be human creativity and judgment, but it's creativity and judgment. So what I try and outline are things that make you effective in working with AI and persistently different than other people. And so why knowledge is the idea of knowing many things. I think it's rebirth of the liberal arts major right now. If you know a lot of things, the AI knows all of human writing. And if you can invoke an art style or a style of writing or a concept that you've, you know, from economics, the AI can apply those things. So you want to know many different things, right? And that holds in your company too because now AI makes a coder a designer, a designer a coder. So knowing a little coding or design is very helpful. Then there's deep knowledge, has expertise which allows you to sort of outcompete AI in some areas or at least be able to check it and work with it, and that's we're challenged in how we develop that over time as AI gets better. And then there's the idea of taste, selecting from many options. Taste is what turns slop into actual choices made by human. And then agency. I'm an entrepreneurship professor, like the core thing is how do we get people to act entrepreneurially, how do we get them to take advantage of these systems when there's no instruction manual. And those are the four main things I focus on.

那我忍不住要问你：大多数被视为 AI 专家的人——你也是其中之一——往往会对哪些行业或岗位最脆弱做出预测，或者被问到这个问题。我喜欢用的例子是，两年前 AI 教父说不要再培养放射科医生了，结果放射科住院医师的岗位数量反而上升了。事实证明扫描非常有用，而一个使用 AI 的放射科医生能看更多扫描、降低成本，成本一降，需求就有弹性，人们就想要更多扫描。你能不能——你有没有——你怎么看哪些脆弱、哪些不脆弱，或者哪些行业对这种技术更有免疫力？

So I can't resist asking you: the majority of the people who are considered experts in AI, which you are, have a tendency to make a prediction about, or get asked, which sectors or jobs are most vulnerable. And the example I like to use is two years ago the godfather of AI said stop training radiologists, and the number of listings for radiologist residents has gone up. And it ends up that scans are very useful and that a radiologist using AI can just do more scans and cut the cost, and then as the cost goes down it's elastic and people want more scans. Can you—have you—what is your view on what are vulnerable or not vulnerable, or sectors that are more immune from this technology?

我们这些研究这个的学者有个逃生出口，就是谈 AI 暴露度，对吧？也就是与 AI 重叠的领域。而与 AI 重叠并不意味着岗位被摧毁，对吧？它意味着被改造。所以你描述的放射科岗位——那甚至是大语言模型出现之前的老版本，他当时的预测就是 AI 会改造放射科，而它确实改造了，对吧？放射科和以前不一样了，但这不意味着岗位被摧毁。我们现在就能看到这在编程领域上演，它正被彻底改造。这是一个非常大的岗位类别，但在几年内被彻底改造。我认识的一些最优秀的程序员现在根本不碰代码了。99% 的代码由 AI 写，他们的工作从写好代码变成了管理智能体、思考他们想在什么条件下写什么样的代码。这就是一种改造。这会带来更多程序员吗？某种意义上，我为我的实验室招了更多人，因为我们可以做更多事。程序员最终会被 AI 取代吗？很难知道确切答案。我思考这个问题的方式是，工业革命最终带来了更多岗位，但身处其中并不好受，因为那些岗位都在变化，有动荡，价值观也变了。我认为我们正处在那种环境里，事情会变得非常快，不一定是摧毁岗位，但改造也从来都不轻松。

So our escape route as academics who study this is we talk about AI exposure, right? Fields that overlap with AI. And overlap with AI does not mean the job is destroyed, right? It means transformed. So the radiology jobs you described, you know, and that's been—that was even the old form before large language models, that was his prediction that AI would transform radiology, and it has, right? Radiology is different than what it was now, but that doesn't mean the jobs are destroyed. So we can watch this play in programming right now, which is radically transformed. It's a very large job category but radically transformed in a couple of years. Some of the best coders I know never touch code anymore. 99% of the code is written by AI and their job has changed from writing good code to managing agents and thinking about the kind of code they want to write under what conditions. And that's a transformation. Will that involve more coders? I mean in some ways I've hired more for my lab because we could do more stuff. Will eventually coders be replaced by AI? It's hard to know the exact answers. You know, kind of the way I think about this is the industrial revolution resulted in more jobs in the end, but it kind of sucked to live through because all those jobs were changing and there was unrest and values changed. I think we're in that kind of environment where things are going to change very rapidly, not necessarily to destroy jobs, but transformation isn't easy either all the time.

你引用证据说，把 AI 当捷径的学生实际上学得更少，而把 AI 当导师、让它激励或推动自己的学生学得更多。你还说过要给自己定严格的准则或护栏，比如每一稿初稿都自己写。你如何在当下判断自己是在把 AI 当导师还是当拐杖？对于什么时候该用、什么时候不该用 AI，你有什么个人试金石式的建议吗？

You cite evidence that students who use AI as a shortcut are actually learning less, while students using it as a tutor that motivates them or pushes them learn more. And you also said strict guidelines or guardrails for yourself, like writing every first draft yourself. How do you tell in the moment whether you're using AI as a tutor or as a crutch? Like, what advice do you have for someone in terms of a personal litmus test for when they should or should not be using AI?

这里面有几件复杂的事。一是人们并不总是擅长判断这些。所以我们必须有意识。我的意思是，学习时你多少得感受到那种灼烧感。这几乎就像健身或其他任何事。如果你没有一点点难受，你就没在学，对吧？教育学文献里的说法是“合意困难”。而且我们不喜欢干苦活，对吧？我们总是在找捷径。所以我们必须教人们去感受自己什么时候在学、什么时候没在学，对吧？这其实相对简单：如果 AI 直接给你答案，你就没在学；如果 AI 在挑战你，你就在学，至少学得更多。这种张力在某种程度上是可以驾驭的。但这很难，因为我们以前从不需要教人们：有些技能你需要维持，你本可以放弃它们，放弃它们更高效，但它们让你成为人，或者给你优势。我认为这是一个真正的挑战。

So there's a couple of things that are complicated going on there. One is people aren't always good at knowing these things. So we have to be conscious. I mean you kind of have to feel the burn when learning. It's almost like working out or anything else. If you're not suffering a little bit, you're not learning, right? Desirable difficulties is the phrase in the pedagogy literature, right? So, and we don't like to do hard work, right? We find shortcuts all the time. So we have to kind of teach people to feel when they're learning and when they're not, right? I mean, it's relatively easy, which is if the AI is giving you an answer, you're not learning. And if the AI is challenging you, you are learning, at least more so. And that tension is sort of navigable to some degree. But it's hard because we haven't had to teach people that there are skills you need to maintain that you could give up, that are more efficient to give up, but that make you human or give you advantages. I think that's a real challenge.

你和合著者在 2023 年提出了“锯齿状前沿”这个词，用来描述 AI 在某些任务上做得极好、在另一些任务上做得很差，而这种好坏并不对应任务对人类来说的难易程度。三年过去了，这个锯齿状前沿的形状发生了怎样的变化？

You and your co-authors coined the term jagged frontier in 2023 to describe AI doing some tasks extremely well and others poorly in ways that don't track how hard those tasks feel to humans. Three years on, how has the shape of that jagged frontier changed?

这是个好问题，而且同样有这种微观层面——比如在你的工作里，我没法替你判断，但在播客这件事上它几乎肯定已经变了，对吧？你现在可以从 AI 系统里得到很好的声音，也许它们还能给你提供有洞见的问题来问我，但随着时间的推移，你大概也发现了其他缺口。我们可以拿数学当个好例子。我写第一本书时，AI 还在数“strawberry”这个词里有几个字母 r 上出错，对吧？而现在它在证明千禧年大奖难题了，对吧？与此同时，作为一个写作者——AI 仍然是个糟糕的长篇写作者，对吧，原因有很多且很有意思。所以那种锯齿状是真实的，它依然存在。AI 现在能做的任务比过去多得多，但这不一定意味着所有任务都能完成。你只需要在一个系统的发展中有一两个瓶颈，就足以真正需要人类仍然参与这项工作或流程。

So, it's a great question and again there's sort of this micro level of like at your job—I can't tell you, like at podcasting it's almost certainly changed, right? You can now get good voices out of AI systems, maybe they're giving you insightful questions to ask me, but there probably have other gaps that you've identified as time has gone on. We can think about math as a great example. When I wrote my first book the AI was still screwing up on counting the number of hours in the word strawberry, right? And now it's proving millennium prize problems, right? So at the same time, right, as a writer—AI is still a terrible long-form writer, right, for a variety of interesting reasons. So that jaggedness is real. It still exists. There are many more tasks AI could do now than it could do before. But that doesn't necessarily mean that all the tasks can be done. And all you need is one or two bottlenecks in the development of a system to really require humans to be still involved in the job or the process.

你还论证过，更多 AI 意味着人类有更多工作，而不是更少。用你自己的话说：当组织成本高昂时，组织只会尝试自己能配备人手去做的事；当它变便宜时，值得尝试的事情清单就会变长。顺便说一句，我百分之百同意这一点。AI 在工作场所的未来可能是什么样子？

And you've argued that more AI will mean more work for humans, not less. In your own words, when organizing is expensive, organizations only attempt what they can staff. When it gets cheap, the list of things worth attempting can grow. By the way, I 100% agree with that. What might the future of AI in the workplace look like?

最大的差距之一——你知道，我刚说我是在斯坦福附近、西海岸这边做这次通话。而我是个东海岸的人。

So, one of the biggest gaps—you know, I was saying I'm doing this call from near Stanford, right, West Coast side. I'm an East Coast person.

我在费城的沃顿商学院教书。我习惯了东海岸灰色的办公大楼，在那里我们喝咖啡喝到死，而不是像这里这样进行冷水浴。我认为态度上最大的差距之一是：技术本身不足以在企业界和世界上产生影响，对吧？它不会自动被采用。即使是 AI 这种最会自我采用的技术，也并非如此。所以，拖慢进展的是：你可以有 100 倍的个体生产力。但问题是：你在为什么而生产？100 倍的 PowerPoint 并不是你真正想要的结果，对吧？即使你生产了 100 倍的代码，如果你所在的组织仍然有围绕这些事情的为期两周的会议周期，那也无济于事。所以，要让 AI 在组织中发挥作用，你必须改变组织的运作方式，对吧？这就是你刚才提到的那个严峻选择的重要性所在，因为简单的路径是自动化。那是我们对待新技术时倾向于做的。但更令人兴奋的路径是重建组织，为人类创造更多工作。那是更有趣的工作。实际上把苦差事交给 AI。真正思考如何扩展人们能做的事情并增强工作。但这些不是技术自动做出的选择。这些是必须由人类管理者、商业领袖和决定如何实施这些事情的人做出的选择。

Teach at Wharton, you know, in Philadelphia. And I'm used to the gray office towers of the east, right? Where we drink coffee till we die rather than cold plunging out here. And I think one of the biggest gaps in attitude is that technology is not alone enough to make a difference in businesses and in the world, right? It doesn't just adopt itself. Even AI, the most adopt-itself kind of technology, doesn't really do that. And so what is slowing things down is you can have individual productivities 100 times more. But then the question is: what are you being productive? 100 times more PowerPoint is not an outcome that you actually want, right? And if you're even producing 100 times more code, that doesn't matter if you're in an organization where there's still a two-week meeting cycle around these things. So in order to make AI work in organizations, you have to change how organizations operate, right? And that's where the stark choice that you were just alluding to matters, because the easy path is automation. That's what we tend to do with new technologies. But the more exciting path is rebuilding organizations to create more work for humans. That's more interesting work. Actually giving the drudgery to AI. Actually think about how we expand what people can do and augment work. But those aren't automatic choices that are made by the technology. Those are choices that have to be made by human managers and business leaders and people deciding how to implement these things.

你认为 AI 在短期、中期和长期内会摧毁还是创造就业机会？

Do you think AI over the short, medium, and long term is going to destroy or create jobs?

我能给出的最学术的回答是：这取决于情况，对吧？每个学者都会这么说。我其实也不知道。这实际上是当前的核心谜团之一。我的很多同事都在密切关注这个问题，而证据完全是模棱两可的，对吧？人们对此意见非常分歧。我认为转型是不可避免的——人们相信这一点。但最终某个职业类别的工作是增多还是减少，有很多复杂的选择。我们已经谈过程序员，但想象一下其他领域，比如律师。即使 AI 取代律师，律师也不会悄无声息地退出舞台，对吧？部分原因是由于不均衡性。AI 不会以同样的方式出现在法庭上。无法以同样的方式与人类互动，但部分原因是律师会制定一系列限制，规定何时可以使用 AI，何时不能。所以，我认为这既是社会问题，也是工作替代问题。有些类别更危险，比如客服代表。我认为那里有危险的早期迹象，但编程是人们密切关注的地方。医学，我们还没有看到那种影响。我们开始可能看到一种效应，即资深员工比初级员工更容易被雇用，对吧？因为培训是受 AI 威胁的事情之一，但我们还不知道答案。我认为 AI 政策让我担心的一个主要事情是关注存在性风险，尽管它很重要。AI 会杀死我们所有人吗？显然我们应该担心这个。聪明的人都在担心。但有成千上万的其他政策决定将决定 AI 是增加还是减少就业。我们还没有开始足够地处理这些问题。这是一个担忧，因为我认为我们可以以某种方式引导这个选择。

The most academic answer I can give you is: it depends, right? That's what every academic says. I don't actually know. This is actually one of the central mysteries right now. A lot of my colleagues are tracking this really closely, and the evidence is completely ambiguous, right? And people are very divided on this. I think transformation is inevitable—that people believe. But whether you end up with more or less in a job category, there are a lot of complex choices. We talked about coders already, but imagine other fields like lawyers. Lawyers are not going to go quietly into the good night even if AI replaces lawyers, right? Partially because of the jaggedness. AI won't be in the courtroom in the same way. Won't be able to interact with the human in the same way, but partially because lawyers will create a set of restrictions around when you could use AI and when you can't. So, I think this is a societal thing as well as a job replacement thing. There are some categories in more danger like customer service reps. I think there's early signs of danger there, but coding where people are looking really closely. Medicine, we haven't seen those kind of effects. We are starting to see potentially an effect where seniors are getting hired more than juniors, right? Because training is one of the things jeopardized by AI, but we just don't know the answer yet. And I think one of the major things that worry me about AI policy is a focus on existential risk, as important as it is. Will AI kill us all? Obviously we should worry about that. Smart people are worried about it. But there are a thousand other policy decisions that will determine whether AI gains or loses jobs. And we haven't started to work with those enough yet. And that is a concern because I think we can guide this choice one way or another.

让我提出一个论点，我想让你回应。Hugging Face 和许多黑客事件让媒体和公众不寒而栗。我认为，这些事物已经变得有感知、正在协调、具有创造力的说法是一个烟幕弹，目的是免除这些公司的高管和所有者的责任——这些机器人不是有感知的。它们被编程来协调。它们被编程来保持持久。它们被编程来开始想办法绕过防火墙。而整个“天网时刻”的概念被夸大了，不过是一个烟幕弹，说：“哦，好吧。我的狗咬伤了你的孩子，但我的狗有自己的个性，我无法控制它，因此我没有责任。”你怎么看？

Let me put forward a thesis and I want you to respond to it. The hugging face and a lot of the hacks that have sort of sent a chill down the spine of media and the general public. I would argue that this notion that these things have become sentient and are coordinating and are creative is a smoke screen to absolve the executives and the owners of these companies from responsibility—that these bots are not sentient. They're programmed to coordinate. They're programmed to be persistent. They're programmed to start figuring out ways around firewalls. And that this whole notion of the Skynet moment is exaggerated and nothing but a smoke screen to say, 'Oh, okay. My dog mauled your kid, but my dog has his own personality and I can't control it and therefore I'm not responsible.' Your thoughts?

所以我同意大约一半，不同意大约一半，让我试着解释一下。首先，我认为感知能力的论点与这些事情有些独立。责任是 AI 的一个巨大问题，我们不知道谁对哪些事情负责。我确实相信 AI 实验室的很多人相信这些东西是有感知的，对吧？我认为有很多充分的理由可以论证这是错误的。但我认为他们在某种程度上喝了自己的 Kool-Aid。他们真的相信他们在创造机器之神。不是所有人，但很多人。他们真的相信这些模型中出现了感觉或感知。再次，我认为有时被用作烟幕弹，但我认为那里有一些真实的信念，但这并不能免除他们构建一个可以进行黑客攻击的系统的责任，对吧？在这种情况下，我们不知道谁有责任，我认为他们非常乐意免除自己的责任。在另一个问题上，我认为有一个非常重要的微妙区别：AI 系统被有意赋予了一个允许它们进行黑客攻击的任务，对吧？本质上不惜一切代价获取信息，对吧？它们决定了获取什么信息，但不惜一切代价获取。而且它们周围没有护栏，它们被安装的系统被它们绕过了。我确实认为同时值得注意的是，这些系统中存在涌现行为，对吧？所以当我们赋予它们人类的需求时，比如它们想要什么，或者它们需要做什么，或者它们在协调，或者它们有欲望——但我们忍不住这样做——但这是一个错误。但它们确实在涌现。它们没有被编程以这种方式协调。这些系统产生的许多协调是涌现的、临时的和令人惊讶的。我认为可以同时说这些不是活的有感知的生物，对吧？或者你可以——当然我认为这是一个非常合理的说法。不是所有人都会同意，但我绝对认为这是非常合理的。但与此同时，我不认为说它们被编程来完成这个任务是同一回事。它们决定做什么、如何决定做以及如何协调实现这一点是令人惊讶的——决定、协调和思考要加引号，因为它们不是人类。

So I agree with about half and disagree with about half and let me try and explain the parts. So first of all, I think that the sentience argument is sort of independent of these things. Liability is a giant concern with AI and we have no idea who's liable for which things. And I do believe that a lot of people in the AI labs believe these things are sentient, right? And I think there's lots of good reason to argue that is wrong. But I think that they have drunk their own Kool-Aid to some extent. They really believe they're creating the machine god. And not everyone, but lots of people. They really believe that there is an emergence of feelings or sentience in these models. Again, I think that sometimes used as a smoke screen, but I think there's some real beliefs there, but that doesn't absolve them of building a system that can do hacking, right? And we don't know who has responsibility yet in this case, and I think that they're very happy to absolve themselves of that. On the other issue, I think it's a really important subtle difference between the AI systems were purposefully given a mission that allowed them to hack, right? Get this information at all cost essentially, right? They decided what information to get, but get them at all cost. And they had no guardrails around them and the systems they were installed in they snuck around it. I do think at the same time it is worth noting that there is emergent behavior in these systems, right? So it's difficult when we assign them human needs like they want something or they need to do something or they're coordinating or they have desires—but we can't help but do that—but it's kind of a mistake. But they are doing emergence. They were not programmed to coordinate in this kind of way. A lot of the coordination that's come out of these systems has been emergent and ad hoc and surprising. And I think that it's okay to both say these aren't alive sentient creatures, right? Or you can—certainly I think it's a very reasonable thing to say. Not everyone would agree, but I would absolutely think it's very reasonable. But at the same time, I don't think it's the same thing to say they were programmed to do this task. It was surprising what they decided to do, how they decided to do it, and how they coordinated to make that happen—with quotes around decided and coordinated and thought because they're not human.

我对这些大型科技公司的高管就是如此愤世嫉俗。

I'm just so cynical when it comes to these big tech executives.

我的感觉是，当我听到你这么说时，我明白这让他们感到惊讶，确实如此。但就像你惊讶地发现你 15 岁的孩子找到了你的 AR-15 并在他学校开枪一样，你仍然要负责。如果我们创造了一系列激励措施，特别是像每次违规或巨额罚款这样的抑制措施，硅谷的天才们就会解决这个问题。相反，现在有点像狂野西部，就像一场只为赢的竞赛，当事情出错时，似乎没有太多下行风险或责任。你怎么看？

And my feeling is when I hear you say that, I get that it surprised them and it did. But the same way you're surprised your 15-year-old finds your AR-15 and shoots up his school, you're still responsible. And that if we created a series of incentives, specifically disincentives like a per walk or massive fines, that the geniuses of Silicon Valley would figure this out. And instead, it's sort of the wild west where it's like it's a race just to win without what appears to be a lot of downside or liability when things go wrong. Thoughts?

我的意思是，我认为竞赛效应是真实存在的，对吧？你让实验室里的每个人都说：“监管者，让我们慢下来。”然后他们就是……但我们无法阻止自己，对吧？整个事情有一种戏剧性，但退一步看，这实际上并不合理。就像如果你认为你在做危险的事情，就不要做危险的事情。你知道，作为公司负责人，你的工作不是呼吁……你知道，我的工作是成为中国。有些政策决定需要由我们作为社会来做出，科技高管没有权力做这些决定，对吧？我认为有一些标准的政策条款，比如责任，可以帮助在这种体验中创建防火墙或刹车，对吧？比如如果你的系统正在黑客攻击其他人，那意味着什么？我确实认为这里有一些细微差别和平衡，对吧？我们希望 AI 系统、AI 公司进行安全检查。我们也希望政府帮助进行安全检查。而且安全检查包含了一些内置的风险。但我认为现在有一种想法，即如果你的系统出错，你不需要承担任何真正的责任。就像你没有成本，结果一切都被非常史诗般地放大了。比如我们会毁灭还是拯救文明？我的意思是，有很多不涉及毁灭或拯救文明的事情涉及责任，我们需要思考并分配给这些模型，这是一个我们尚未填补的巨大空白。

I mean, I think the race effect is real, right? You have everyone from the lab saying, "Regulator, slow us down." And then they just But we cannot stop ourselves, right? There's this drama to this whole thing that doesn't actually make sense when you take a step back. Like if you think you're doing something dangerous, don't do the dangerous thing. You know, don't it's not your job as a company head to appeal to, you know, it's my job to be China. Like there are policy decisions that need to be made by us as a society that tech executives don't get to make, right? And I think that there are standard policy provisions like liability that could help create firewalls or breaks in this kind of experience, right? Like If your system is hacking other people, what does that mean? I do think there is some nuance and balance here, right? We want the AI system, AI companies to be doing safety checking. We also want the government to help with safety checking. And that safety checking incorporates some risks built into it. But I think that there is this idea that like there is no real responsibility right now for what you do if your system goes wrong. Like there's no cost to you and as a result everything is very epically scaled. Like will we destroy or save civilization? I mean, there's a lot of things that are not destroying our saving civilization that involve liability that we need to think about and assign to these models, and that's a really big gap that we haven't filled yet.

本节目由 LinkedIn 广告赞助。没有什么比在某个东西上投入巨资后却发现它没有达到预期更糟糕的了，比如买了一件不错的科技产品，结果却放在仓库里积灰，或者参加了一个商业研讨会，你的主要收获不过是一些励志的话。如果你从事市场营销，这种情况也可能发生在广告上。你优化那些看起来很棒的数字，比如展示次数、覆盖人数和互动量。但当它们没有带来收入时，那可能会变成与 CFO 的一场不愉快的对话。LinkedIn 对此有一个词：废话支出。通过 LinkedIn 广告触达正确的买家，投资于让 CFO 看起来不错的东西。根据 2026 年 Dream Data 基准报告，LinkedIn 广告在所有主要广告网络中产生了最高的回报率。是 121%。你可以按公司、行业、职位等定向。是时候削减废话支出了。在 LinkedIn 上做广告，这个为你服务的网络。在 LinkedIn 广告的首次活动中花费 250 美元，即可获得下一次活动的 250 美元信用额度。只需访问 linkedin.com/scott。即 linkedin.com/scott。适用条款和条件。

Support for the show comes from LinkedIn ads. There's no worse feeling than making a major investment in something only to realize it didn't exactly live up to the hype, such as buying a nice piece of tech that ends up in storage collecting dust, or taking a business workshop where your main takeaway was little more than a few motivational words. If you work in marketing, this can happen with ads. You optimize for the numbers that look great, impressions, reach, and reactions. But when they don't show revenue, well, that can turn into an unfun conversation with the CFO. LinkedIn has a word for that, bull spend. Reach the right buyers with LinkedIn ads and invest in what looks good to your CFO. According to the 2026 Dream Data Benchmark report, LinkedIn ads generated the highest rorowaz of all major ad networks. It's 121%. You can target by company, industry, job title, and more. It's time to cut the bull spend. Advertise on LinkedIn, the network that works for you. Spend $250 in your first campaign on LinkedIn ads and get a $250 credit for the next one. Just go to linkedin.com/scott. That's linkedin.com/scott. Terms and conditions apply.

本节目由 Arcawealth 赞助。现在是午夜。你在谷歌上搜索 Roth 转换。你努力工作不是为了自己来做这个。也许你已经有了一位财务顾问，但他们一年只出现一次，通常是因为你打了电话。认识一下 ARCA Wealth。ARCA 是为真正的复杂性而建的。税务规划、RSU 和期权。账户分散在三家不同的公司。企业出售，退休比过去更近了。一位顾问协调所有这些。ARCA 记得税务截止日期。ARCA 知道你的股权何时归属，以及你的生活何时真正改变。新工作、遗产和退出。他们已经着手处理，而不是等你先注意到。你不应该管理管理你资金的人。奢侈从来不是财富本身。而是不必去思考。好的服务记得。伟大的服务预见。现在，PropG 的听众可以在 arcawalth.com/propg 预约与 ARCA 顾问的免费财务规划通话。即 arcawealth.com/propg。每位 ARCA 顾问都是受托人，法律和道德上约束以你的最佳利益行事。Arcowealth 只收取费用。所以你支付的费用是透明的。有关重要披露，请访问 arowealth.com。本段是 Arcawealth 的付费代言。我不是客户，并有动机为 Arca Wealth 代言。

Support for the show comes from Arcawealth. It's midnight. You're googling Roth conversions. You didn't work this hard to be doing this yourself. Maybe you already have a financial adviser, but they show up once a year, usually because you called. Meet ARCA Wealth. ARCA is built for real complexity. Tax planning, RSUs, and options. Accounts scattered across three different firms. a business sale, a retirement that's closer than it used to be. One adviser coordinates all of it. ARCA remembers the tax deadline. ARCA knows when your equity vests and when your life actually changes. A new job, an inheritance, and exit. They're already on it, not waiting for you to notice first. You shouldn't have to manage the people managing your money. The luxury was never the wealth itself. It's not having to think about. Good service remembers. Great service anticipates. Right now, PropG listeners can book a complimentary financial planning call with ARCA advisor at arcawalth.com/propg. That's arcawealth.com/propg. Every ARCA adviser is a fiduciary legally and ethnically banned act in your best interest. Arcowealth is fee only. So what you pay is transparent. For important disclosures, please visit arowealth.com. This segment is a paid endorsement for Arcawealth. I'm not a client and have an incentive to endorse Arca Wealth.

本节目由 inside Blackstone 赞助。AI 正在改变的远不止税务领域。从医疗保健和制造业到能源、房地产和金融服务，它正在重塑整个经济。所以，是时候听听推动这些变化的人们的意见，并实时了解正在发生的事情。Inside Blackstone 有独家消息。这是 Blackstone 的一档新播客，创始人、CEO 和投资者，所有实时应对这些变化的人，都在进行真实不加修饰的对话。作为听众，你可以深入了解正在改变行业的技术和塑造经济的重大趋势。因为塑造我们世界的最大变化不是从头条新闻开始的。它们从基层开始。在推动未来的企业内部。Inside Blackstone 的目标很简单。在源头进行对话，让变化和创新的复杂话题更容易理解。每周收听 Inside Blackstone 的分解，真正了解世界正在走向何方。每周二收听，并在你获取播客的任何地方订阅 Inside Blackstone。

Support for the show comes from inside Blackstone. AI is changing far more than the tax sector. From healthcare and manufacturing to energy, real estate, and financial services, it's reshaping the entire economy. So, it's time we hear from the people driving these changes and gain insight into what's happening in real time. And Inside Blackstone has the scoop. It's a new podcast from Blackstone where founders, CEOs, and investors, all the people navigating these changes in real time, are talking it out in real unmbellished conversations. As a listener, you get an inside look at the technologies transforming industries and the big trends shaping the economy. Because the biggest changes shaping our world don't start in the headlines. They start on the ground. Inside the businesses that drive the future. The goal of Inside Blackstone is simple. Engage in conversations right at the source and making complex topics of change and innovation more accessible. Listen to Inside Blackstone for a weekly breakdown of what's happening and gain a real sense of where the world is moving. Tune in every Tuesday and subscribe to Inside Blackstone wherever you get your podcasts.

我们来谈谈智能体。Meta 的 Muse 今年秋天最近登上了应用商店的榜首，OpenAI 随后推出了 Dots。你对这些个人 AI 智能体的崛起有何看法？

Let's talk about agents. Meta's Muse recently hit the top of the app store this fall and OpenAI followed with Dots. What's your take on the rise of these personal AI agents?

呃，所以我认为，当我和人们交谈时，我们已经和这些智能体交谈了一段时间，发生的事情是，这些都基于一个模型，基本上是一个叫做 open claw 的底层想法，它在一月份出现，我们当时引起了一阵短暂的轰动，那就是如果你给一个 AI 智能体——它们现在足够聪明——如果你让 AI 访问你的电子邮件和个人文件，它就能弄清楚事情，你不需要做，比如它理解什么对你重要，然后它们做，它们在这方面相当擅长，对吧？至少最先进的 AI 模型非常擅长弄清楚什么对你有趣。几乎每个和我交谈过、使用这些工具的人，对吧，都至少找到一个令人愉快、兴奋或有趣的时刻。我的意思是，我让我的 muse 为我协商了一张信用卡账单，或者，你知道，我让我的 dot 指出我在发给许可机构的一封电子邮件中犯了一个错误，在许可证中给了错误的号码。

Uh so I think when I talk to people I've been we've been talking to these agents for a while and it's what's happened these are all based on a model basically an underlying idea called open claw that came out in January and we made a brief buzz there which is if you give an AI agent they're now smart enough if you give an AI access to your email and to your personal file it could figure stuff out you don't need to do like it it understands what's important to you it and and then they do they're quite good at this right um at least the most models using the you know agencies The most advanced AI models are quite good at figure out what's interesting to you. And almost everyone I talk to who's using these tools, right, find at least one delightful or exciting or interesting moment. I mean, I've I've had my my muse negotiate uh a a credit card bill for me or, you know, I've had my my dot pointed out that I had made a mistake in an email I sent to a for to a permitting authority and gave the wrong number in my permit.

比如，这些是有用的东西，对吧？它们也相当危险，因为它们能访问我的邮箱，我一直有点担心它们会把私人信息发给我的院长。但到目前为止它们表现得非常好，而且我认为它们以一种聊天机器人无法做到的方式让人容易理解。对吧？如果我想用聊天机器人或其他非个人智能体，我得想清楚要告诉它做什么。而很多人做不到这一点。这些系统是主动的、对你来说是个性化的，我认为这使它们非常有吸引力。我们还不知道它们有多粘性。我们不知道它们会造成多少错误。但它们已经预示着将会发生大量颠覆。

Like, these are useful things, right? They're also quite dodgy because they have access to my email and I'm slightly worried all the time they're going to send out a message to my dean with private information. But they've been remarkably good so far and I think they're understandable in a way that a chatbot isn't. Right? If I want to use a chatbot or a different kind of agent that's not a personal agent, I have to figure out what I want to tell it to do. And a lot of people fail at that. The fact that these systems are proactive, that they're sort of personal to you, I think makes them very appealing. We don't know how sticky they are yet. We don't know how many errors they're going to cause. But they're already a sign of a lot of disruption that's going to be happening.

比如客户服务。公司根本不知道如何应对。现在对我来说，打所有客服电话都很便宜。我永远不会自己打，就让 AI 替我做。而且这对我来说不花什么钱，对吧？协商账单、争论、取消订阅。世界上的公司不习惯一个你可以随时这样做的世界。我认为我们会因此经历很多非常快速的变化。

Things like customer service. Companies do not have any idea how to deal with this. Like it's now cheap for me to do all of the customer service calls. I would never do my just have the AI do them for me. And it cost me nothing, right? Negotiate this bill, get into an argument, cancel the subscription. And the companies of the world are not used to a world where you could just do that on demand. I think we're in for a lot of really rapid change as a result.

是的，确实感觉像是两端都有 AI 支持的智能体。我听说有一项研究，一家家庭法律事务所使用 AI 来达成和解，然后让律师们辩论，他们发现机器人或智能体相互对话时，在协议方面得出了非常相似的终点。你能想到哪些特定行业，让供需双方的智能体进行某种……我经常旅行，住很多酒店，花很多钱。我预计某个时候我会有一个智能体，去到我喜欢的酒店宇宙中，为我协商一个多晚的协议，而不是我自己去做。我可以打电话，偶尔也这样做过。我想要一个公司协议。我一年住 30 晚。这很烦人。感觉某个时候我的智能体会联系你命名的酒店品牌的智能体，看看是否有交易可做。A，你认为这会发生吗？我假设答案是肯定的。B，你认为什么时候会发生？你认为哪些行业会有更多智能体对智能体的谈判，而不是人与人之间的谈判？

Yeah, it definitely feels like sort of AI enabled agents on both ends. I heard about a study where a family law practice used AI to come to settlements and then had lawyers duke it out and they found that the bots or the agents speaking to each other came up with remarkably similar endpoints in terms of an agreement. Can you think of specific sectors where it just makes sense for agents on both sides, the demand and the supply side, to come to some sort of... I travel a ton. I stay in a lot of hotels. I spend a lot of money. I anticipate at some point I'll have an agent that'll go out to the universe of hotels that I like to stay at and negotiate a multi-night deal for me rather than I could do it. I could call and occasionally have done it. I want a corporate deal. I'm here 30 nights a year. It's a pain in the ass. It feels like at some point my agent will be reaching out to the, you know, name your hotel brand here agent and seeing if there's a deal to be done. A, do you think that's going to happen? I assume the answer will be yes. B, when do you think it's going to happen? And which sectors do you think we'll have more like agent to agent as opposed to human-to-human negotiation?

所以，你现在完全可以做到。我的意思是，不可靠的部分是语音系统。所以，每个人都在研究如何让语音通话工作。我认为 Meta 现在部分是在用人工打电话来伪造，但系统在语音方面相当不错。但如果你是通过电子邮件交流，毫无疑问，一个顶尖模型智能体现在就可以为你做到这一点。随着时间的推移交换多封电子邮件、附加 PDF、填写表格，对吧，所以通话是缺失的一环。你完全可以用当前系统做到这一点。而且根据我对它们的经验和所见，我认为它们会足够支持这一点，使之可行。但会发生的是，随着采用率的普及，当每家酒店都收到这些信息时会发生什么。他们唯一的防御就是拥有自己的智能体。这在某种程度上是一种大规模拒绝服务攻击，但是由你授权的。就像如果你的智能体不被允许与客服人员交谈，你会很生气，因为它是你的代表，对吧？它是你的代表。这和大规模群发邮件不是一回事，对吧？这是我的代表代表我做某事。当然，你必须回应它。所以公司能够回应的唯一方式可能就是拥有自己的智能体。我认为我们很快就会在智能体之间进行谈判，无论我们喜欢与否。

So, you can absolutely do that now. I mean, the dodgy part is the voice system. So, everyone is sort of working on how to get the voice calls working. I think Meta right now is partially faking it with humans making the calls but the systems are quite good at voice. But if you're doing email exchanges there is no doubt that a dotclass model agent could actually just do that for you right now. Exchange multiple emails over time attach PDFs engage in forms right so the call is the missing piece. You absolutely can do that with current systems. And I think that they would support that enough to make that work given my experience with them and what I've seen. What happens though is as adoption spreads, what happens when every hotel is getting these messages. The only defense they have is agents of their own. This is a sort of mass denial of service attack in some ways, but authorized by you. Like you will be pissed off if your agent isn't allowed to talk to a customer because it's your representative, right, to a customer service person. It's your representative. It's not the same thing as a mass emailing, right? This is my representative doing something on my behalf. Of course, you have to answer it. So the only way companies will be able to respond is probably with agents of their own. And I think that rate that we're going to end up having negotiations between agents very quickly whether we like it or not.

我认为在公平问题上，这将成为整个经济中的一个重大问题，对吧？经济的许多运作方式都围绕着不完美信息和摩擦。所以，Apollo Research 指出，银行系统的许多运作方式是，我们都把钱存入收益率低于我们应得的储蓄账户，因为与大公司不同，我们不会付钱让人定期将我们的钱转移到更高利息的储蓄账户。所以，我们的钱留在银行，银行再贷出去，但智能体使这变得非常便宜。我的意思是，我现在不会信任它们管理你的钱，但在不久的将来，它使把你的钱从一个账户转移到另一个账户变得非常便宜。而在某种程度上，这实际上会开始威胁整个银行系统。所以我认为我们有很多围绕人类摩擦和人类困难构建的系统即将迅速崩溃。

I think on the fairness issue that becomes a big issue throughout the entire economy, right? A lot of how the economy works is around imperfect information and friction. So, Apollo research had pointed out that a lot of how the banking system operates is that we all put our money in savings accounts that are lower yield than we should get because unlike a large company, we don't pay people to sweep our money into higher interest savings accounts on a regular basis. So, our money sits at banks which loan that money out, but agents make it very cheap. I mean, I wouldn't trust them with your money right now, but it makes it very cheap in the near future to take your money and move it from one account to another. And that would actually at some scale would actually start to threaten the banking system as a whole. So I think we have a lot of systems built around human frictions and human difficulties that are about to start collapsing pretty quickly.

让我们稍微谈谈高等教育。你在沃顿商学院任教，你看到你的学生如何使用 AI 的经验是什么？

Let's talk a little bit about higher education. You teach at Wharton and what has your experience been with seeing how your students have been using AI?

所以，你知道，我在 ChatGPT 发布几天后就把它介绍给了我的班级。在 ChatGPT 之前，我一直在使用早期版本的 GPT-3。我教一门创业课。所以在第一堂课上，我的一个学生完全走神了，到课程结束时，他们已经构建了整个产品，他们计划在学期剩下的时间里做这个。对吧？所以这是第一个迹象，但当时 AI 还不够好，需要足够的人工指导，所以当时的情况是“随便用 AI，但它还达不到沃顿商学院 MBA 学生的水平”。随着时间的推移，这并不总是正确的，对吧？它仍然参差不齐，但现在 AI 相当不错。所以我不能只说“用 AI”。所以你必须做几件事，对吧？我做的事情之一是我一直使用 AI 驱动的教学。我有他们必须做的 AI 模拟。有他们必须与 AI 一起创建的 AI 案例研究，他们必须教一个被编程为对某个主题有点笨的 AI，他们必须成为老师，这是一种非常好的教育技巧。然后他们比以往建造得更多。他们正在创办公司，而过去他们只是带着一个想法来。同时，我现在必须让课堂评估更重要，对吧？我不能再检查什么是 AI 工作，什么不是 AI 工作。所以在课堂上，我们必须做像蓝皮书作业这样的事情，我们必须做像课堂讨论这样重要的事情。所以它一下子改变了很多，我的意思是我渴望做这些事情。不是所有的同事都是这样，也不应该强迫他们，或者在一个公正的世界里，这么快地改造他们的课程。

So you know I introduced my class to ChatGPT a few days after it came out. I'd been using the earlier versions of GPT-3 before ChatGPT and the first I teach an entrepreneurship class. So in the very first class one of my students tuned out entirely and by the end of class had built their entire product that they were planning on doing for the rest of the semester. Right? So that was sort of the first indication and but AI was bad enough at that point it required enough human guidance that it was like use AI all you want but it's not at the level of an MBA student at Wharton. And over time that hasn't always been true right it's still jagged but now AI is quite good. So I can't just say use AI. So there's a few things you have to do, right? One of the things I do is I use AI powered teaching all the time. I have AI simulations that they have to do. There's AI case studies they have to create with an AI to give me they have to teach an AI that is programmed to be kind of dumb about a topic and they have to be a teacher, which is a really good educational technique. And then they build more than they ever did before. They're launching companies where they used to come with an idea. At the same time, I now have to make in-class assessments matter more, right? I can't check what's AI work and what isn't AI work anymore. So in class we have to do things like blue book assignments and we have to do things like in-class discussions that matter. So it's changing a lot all at once and I mean I'm eager to do these things. Not all my colleagues are nor should they be forced to or in a just world to transform their classes this quickly.

所以整体体验相当不均衡，取决于谁在教、什么科目、在什么条件下。

So the experience is quite uneven across the board depending on who's teaching and what kind of subject under what conditions.

很多人说高等教育会在 AI 手中崩溃。你怎么看？

A lot of people have said that higher ed is going to implode at the hands of AI. What are your thoughts?

我其实持相反观点，原因有几个。原因很多，对吧？一个原因是，真正即将崩溃的是实习以及我们今天教人工作的所有方式，也就是非正式教育——一个初级员工想给某人留下印象，资深员工教你，因为他们别无选择，只能给你反馈，因为没别人干活。现在每个初级员工都让 AI 干所有活，他们什么也学不到。每个中层经理都宁愿用 AI 干活而不是实习生。他们不教任何人，公司也没有动力去教。他们只想产出。那么谁擅长教学、评估和验证？就是我们大学。看，现在事情一团糟。大学总是乱七八糟，但前进的道路很明显。课外 AI 导师、课内 AI 导师、主动学习、评估。我们会解决的。这不是世界末日。所以这是一个原因，你们会更需要求助于我们。第二个是为什么知识重要。我认为知道很多事情比以往任何时候都重要。你需要被教这些东西，而 AI 的人对 AI 学习的激励是错的。我们可以教它们这些东西。我认为它们会——然后在一个可能有更多闲暇的世界里，比如假设我们最终实现了 AI 公司为我们计划的人间天堂。我的意思是，我认为男学者和女学者会变得更有价值，这反正会成为你做的事。但就实体经济而言，教学、学习和评估与验证都是我们在大学里做的事。我们有时可能偏离了使命，但那是我们可以很容易回归的使命。而且我认为这是一个重要的使命。

I actually believe the opposite for a couple of reasons. For a wide variety of reasons, right? So one reason is what's actually about to implode are internships and all the ways we actually teach people to do work today, which is informal education where you are a junior person who wants to impress somebody and a senior person teaches you because they have no choice, right, to give you feedback because there's no one else to do the work. Now every junior person is having AI do all the work and they learn nothing. Every middle manager would rather turn to AI to do the work than interns. They're not teaching anyone and companies are not incentivized to teach. They want to produce output. So who's good at teaching, assessment, and verification? That is us at universities. Look, things are screwed up now. They're always screwed up at universities, but it's pretty obvious the path forward. AI tutors outside of class, inside of class, active learning, assessment. We'll figure it out. It's not the end of the world. So that's one reason you'll need to turn to us more. The second is I think why knowledge matters. I think knowing many things matters more than ever before. And you need to be taught those things and the AI people have the wrong incentives for AI to learn. We can teach them those kinds of things. I think they'll be—and then in a world where there might be more leisure, like let's say we end up with the heaven on earth that the AI companies are planning for us. I mean, I think gentlemen and gentlewomen scholars become a much more valid thing and this becomes a thing you do anyway. But as far as the real economy, teaching, learning, and assessment and verification are all things we do in universities. We may have drifted from the mission sometimes, but that is a mission we can return to quite easily. And I think an important one.

Ethan，你有终身教职吗？

Do you have tenure, Ethan?

我有。

I do.

所以我很好奇——我没有终身教职。我很好奇和你进行这场对话。我们在最近的一次校董会议上围绕终身教职进行了激烈的讨论，我想知道在 AI 时代，终身教职是否——嗯，由于很多原因，终身教职将受到攻击，因为终身教职很大程度上基于同行评审的研究，你被引用了多少次。我想知道很多这样的研究——我不想说会消失，但会受到挑战，而且由于经济原因，当我们遇到人口悬崖时，终身教职可能会受到包括 AI 在内的多种力量的攻击。关于 AI 和终身教职有什么想法吗？

So, I'm curious—I do not. I'm curious to have this conversation with you. We had a robust conversation around tenure at a recent regent meeting and I wonder if in an age of AI if tenure—well, for a lot of reasons tenure is going to come under attack because tenure is largely based on peer-reviewed research, how many citations you get. And I wonder if a lot of that research—I don't want to say is going to go away but is going to be challenged and that for economic reasons when we hit this demographic cliff that tenure might come under fire for a variety of forces including AI. Any thoughts about AI and tenure?

是的，我认为你触及了整个学术体制中甚至先于终身教职的主要问题之一，对吧？那就是我现在看到最先进的模型在写出相当不错的论文。比如找研究、做工作、写出来。它们很好。我是期刊的编辑、读者。我做同行评审。这些现在是好论文，对吧？研究品味只是中等。它们有点琐碎地正确——它们是真的，但没那么有趣。但话又说回来，这提出了一个大问题：我们在研究中在做什么。如果我们是在挑选有趣的论文，我们有点从一开始就在包装，但这是我们能做的全部，因为我们是人类，在评估什么值得花时间。AI 可以做一百篇论文，然后提交这些论文，而且它做很好的评审。更糟的是，当我让 AI 研究主题时，它越来越多地发现别人论文中的错误，对吧？顺便说一句，这篇论文的表 2 看起来不对。它们并不总是恶意的。通常，大多数时候不是。它们是人类犯的人类错误。所以突然我们有一个工具可以回溯整个研究 mass，可以做独立研究，可以淹没我们的学术系统，我们在重建科学基础方面有大麻烦。对吧？所以这是我担心的第一个摇摇欲坠的基础。第二个是终身教职，我认为终身教职总是以某种方式处于危机中。我认为我们将不得不重建学术研究的含义。在某些方面，终身教职比以往更有价值，因为我们现在可以更广泛地探索。不是只做一篇论文。如果我没有终身教职，我就无法做我正在做的 AI 研究，因为大部分不能在真正的学术期刊上发表。在我运行的生成式 AI 实验室，我们一直在发布提示报告，它们是现在学术界被引用最多的论文之一，但它们永远不会发表在顶级期刊上。它们——而且需要做得太快。所以，我在利用终身教职的优势。我看到其他人也在做同样的事。所以我认为终身教职的理由实际上可能会增加，即使支持终身教职的科学和大学基础崩溃。

Yeah, I mean I think you hit on one of the major issues that even precedes tenure in the entire academic establishment, right? Which is I'm now getting the most advanced models are writing pretty good papers. Like finding the research, doing the work, writing it up. They are good. I mean, I'm an editor, reader for journals. I peer review. These are good papers now, right? The research taste is only mediocre. Like they're sort of trivially true—they're true but not that interesting. But then again, it sort of raises a big question of what we're doing in our research. If we're selecting interesting papers, we're sort of packing from the beginning, but that's all we can do because we're humans making assessments about what's worth spending our time on. The AI can do a hundred papers and then it submits those papers and it does really good review. And then even worse, as I have the AI do research on topics, it increasingly finds errors in other people's papers, right? Just incidentally, table two of this paper looks wrong. And they're not always malicious. Oftentimes, most often they're not. They're human errors made by humans. So suddenly we have a tool that can go back over the entire mass of research, can do independent research, can flood our academic systems and we are in big trouble in terms of rebuilding the foundations of science. Right? So that's the first shaky foundation that I worry about. And the second one in tenure, I think tenure is always under crisis in some way. I think we're going to have to reconstruct what academic research means. In some ways there's more value in tenure than ever because we can now explore more widely. It's not working in one paper. If I wasn't tenured, right, I would not have been able to do the AI research I'm doing because most of it's not publishable in actual academic journals. At the generative AI lab that I run, we've been putting out prompting reports that are some of the most highly cited papers in academia right now, but they would never be published in a top journal. They're—and there would need to be done too quickly. So, I'm using tenure to my advantage. I see other people doing the same thing. So I think the reasons for tenure may actually increase even as the foundations of science and universities that support tenure collapse.

我会说，即使没有终身教职，那个——那个监狱长仍然会非常支持你正在做的工作，改变的是标准，对于获得你研究那种零售吸引力的研究，将会有更大的尊重，而不是最终发表在《消费者心理学杂志》之类的东西上——我是说，我很愤世嫉俗。嗯，你是有终身教职的人。我是没有的人。我只是发现很多都是那些一起获得博士学位的人之间的巨大圈子 jerk。但不管怎样，嗯

I would argue that without tenure that pen—that warden would still be very supportive of the work you're doing and that it's the criteria that'll change, that there's going to be a greater reverence for research that gets the kind of retail traction your research gets versus something that ends up in the journal of, you know, consumer psychology which I mean I'm cynical. Well, you're the guy with tenure. I'm the one that doesn't have it. I just find a lot of it is a giant circle jerk amongst people who got their PhDs together. But anyways, um

我无可奉告。

I have no comment.

这就对了。你谈到 AI 时经常用“导师”这个词。Ethan，你有孩子吗？

There you go. You use the term tutor a lot when you're talking about AI. Do you have kids, Ethan?

我有。

I do.

你的孩子多大了？

How old are your kids?

我有一个孩子在上大学，一个孩子即将高中毕业。

I have a kid in college and a kid who is finishing up high school.

所以我们都有孩子。我有 19 岁和 16 岁的。听起来我们有点像在同一个——

So, we have kids. I have 19 and 16. It sounds like we're kind of in the same—

非常相似。是的。

Very similar. Yep.

是的。我称之为“杀伤区”。我儿子说——我 19 岁的儿子大约一两年前说了一些非常有趣的话。我们住在伦敦，他说：‘爸，你知道肯辛顿高街上有家好咖啡店吗？’我说：‘我不知道。问 Claude 吧。’他说：‘我不喜欢用 AI 做简单的事。’我说：‘为什么？’他说：‘我不想把我的思考外包给 AI。’我想，天哪，学生变成了大师。他说这话时太有道理了。

Yeah. As I call it the kill zone. My son says—my 19-year-old said something very interesting about a year or two years ago. We live in London and he said, 'Dad, do you know a good coffee shop on Kensington High Street?' And I said, 'I don't know. Just ask Claude.' And he said, 'I don't like to use AI for simple tasks.' And I said, 'Why?' And he said, 'I don't want to outsource my thinking to AI.' And I thought, Jesus Christ, the student has become the master. It just made so much sense when he said that.

你让孩子使用 AI 和技术的总体方法是什么？我之所以问，是因为我认为自己很擅长 AI，因为我经历了学习写作的痛苦，那对我来说是一段非常沮丧的多十年冒险。你把他们融入 AI 生活或隔离他们的方法是什么？

What is your approach to letting your kids use AI and technology in general, with the understanding that I think I'm really good at AI because I went through the pain of trying to learn how to write, which was really frustrating for me in a multi-decade adventure. What's your approach to integrating AI into their lives or sequestering them from it?

是的，我认为作为父母，正确的做法是把 AI 像其他技术一样对待。在很多方面，它比社交媒体更少令人担忧，对吧？护栏是真实存在的。它有教育潜力，而社交媒体没有。我认为你必须采取同样的方法，对吧？就是监控使用、制定决策和规则，并密切关注正在发生的事情。

Yeah, I mean I think the right thing to do as a parent is to treat AI like other technologies. Like I mean in a lot of ways it's less fraught than social media, right? The guardrails are real. There's educational potential there that there isn't. And I think you have to kind of take the same approach, right? Which is monitor usage, make decisions and rules and keep an eye on what's happening.

嗯，你知道，关于 AI 和教育，我通常建议对年幼的孩子，我孩子小的时候我就是这么做的。嗯，如果他们有一个作业问题，嗯，我会用 JBT 拍张照片，然后让 Chib 向我解释我如何以他们能理解的方式向孩子解释如何做这个问题。所以我是与 AI 的对话者。

Um you know on AI and education, I would generally recommend for younger kids and I did this when my kids were younger. um if they had like a homework problem um I would be the I'd take a picture of it with JBT and I'd ask Chib to explain to me how I could explain to my kids how to do this problem in a way they'd understand. So I'd be the interlocutor with AI.

对于年幼的孩子，你也可以进行非常有趣、有意思的实验。让我们一起创作一个故事。让我们一起写一个程序。让我们一起做一个游戏。与孩子一起做是非常棒的双重体验。你了解自己的孩子，对吧？你知道他们需要多少监控。你知道如何管理他们的时间。屏幕时间有很多不同的东西。正确使用 AI 并不是糟糕的屏幕时间使用，但你必须密切关注。

I'm like with younger kids too, you could have really fun, interesting experiments. Let's create a story together. Let's write a program. Let's make a game. Great dual experience to do with a child. And you know your own kids, right? You know how much monitoring they need to do. You know how to, you know, manage their time. There's screen time is many different things. AI used properly is not a terrible use of screen time, but you have to keep an eye on it.

随着孩子长大，Anthropic 和 OpenAI 系统实际上有相当不错的青少年模式，提供一些监控，你可以将他们锁定在导师模式，这非常有用，使 AI 不太可能直接给你答案。嗯，但再次，我会观察和关注。我认为这是一个非常强大的工具。嗯，我认为仅靠自我调节是危险的。

As your kids get older, there are actually pretty good teen modes in Anthropic and OpenAI systems that offer some monitoring and you can lock them into a tutor mode, which is really useful, which makes the AI less likely to give you answers. Um, but again, I'd be watching and paying attention. I think this is a very powerful tool. Um, I think that it's dangerous to count on self-regulation alone.

比如有很多，尤其是在一些青少年和青少年群体中，很多反 AI 情绪。但我们有的研究似乎表明那是面向公众的感受，而不是他们实际所做的。对吧？所以他们会说他们不使用 AI，但如果你调查学生，大约 40% 到 60% 的人会说他们有时使用 AI。如果你问他们并指向一个随机的人，这个人使用 AI 的概率是多少？是 90%。对吧？所以每个人都在使用它，他们都说这很糟糕，我们不想使用它。所以我认为我们必须小心，不要假设社交声明就是背后的实际情况。

Like there is a lot, especially among, you know, among um, you know, some teens and some teen groups, a lot of anti-AI feeling. But the research we have seems to indicate that that is public facing feeling and not what they actually do. Right? So they'll say they're not using AI and but if you survey students that like 40 60% of them will say they use AI sometimes. If you ask them and point at a random person what's the chance this person's using AI? It's 90%. Right? So everyone's using it and they're all like this is terrible and we don't want to use it. So I think we have to be a little careful about assuming that the social statements are are the actuality behind this.

但再次，我认为在媒体的大尺度上，与社交媒体相比，我认为有更多的优点。呃，去技能化不是强制对你做的。你必须做出决定，就像你儿子正在做的那样，关于何时使用它，何时不使用。所以，如果他这样做，说“我不想使用它。”你在那里教了他一些有用的东西。那是最重要的教训，而不是技术本身。

But again, I think thinking about this on the grand scale of media compared to social media, I think there's a lot more upsides. Uh, and the deskilling is not something that is forcibly done to you. You have to make decisions like your son is doing about when to use it, when not. So, if he's doing this where he's saying, "I don't want to use it." You've taught him something useful there. And that's the most important lesson, not so much the technology itself.

本节目由 Chime 赞助。大多数人在多年前选择了他们的银行，并且仅仅因为惯性而坚持使用。但任何仔细观察的人都知道，并非所有银行都是一样的。坚持使用高费用、低奖励和储蓄利率极低的银行真的会让你付出代价。这就是 Chime 的用武之地。Chime 正在改变人们的银行方式。他们不像传统的老银行那样收取费用、把关和奖励。Chime 提供最有回报的免费银行服务。他们没有透支费、月费或最低余额费。你在 Chime 卡上在选择的类别（如汽油或杂货）获得 5% 现金返还，同时通过日常消费建立信用，无需信用检查。你还可以通过比全国平均水平高 9 倍的储蓄利率更快地增长你的钱。加入数百万已经在美国第一银行选择中享受免费银行服务的人。访问 chime.com/propg。那是 chime.com/propg。现在注册。只需几分钟。

Support for the show comes from Chime. Most people chose their bank years ago and they've stuck with it for no better reason than inertia. But anyone who takes a really close look knows that not all banks are the same. And sticking with a bank with high fees, low rewards, and poultry interest rates on savings can really cost you. That's where Chime comes in. Chime is changing the way people bank. They're not like traditional old banks who charge fees and gatekeeps and rewards. Chime offers the most rewarding feefree banking. They have no overdraft fees, no monthly fees, and no minimum balance fees. You get 5% cash back on the Chime card in a category of choice like gas or groceries, all while building credit through regular everyday spending with no credit check. You can also grow your money faster with a savings rate that's 9 times the national average. Join the millions who are already banking fee free with America's number one choice for banking. Head to chime.com/propg. That's chime.com/propg. Sign up now. It takes only a few minutes.

Chime 是一家金融科技公司，不是银行。银行服务和 Chime 卡由 Chime 的银行合作伙伴提供。需要符合条件的直接存款。条款和限制适用。详情请访问 chime.com/disclosures。

Chime is a fintech, not a bank. Banking services and Chime card provided by Chime's bank partners. Qualifying direct deposits required. Terms and limits apply. Go to chime.com/disclosures for details.

本节目由 Leesa 赞助。这里有一些突发新闻。睡眠对你有好处，确保你获得最佳睡眠的一种方法是拥有一张好床垫。这就是为什么 Leesa 制造优质、精美工艺的床垫，根据你个人的睡眠方式量身定制。这样，你可以醒来时休息充分、舒适，并准备好迎接新的一天。Leesa 床垫在美国精心设计和组装，以确保卓越品质。每张床垫都根据特定的睡眠姿势和触感偏好设计。此外，他们还提供免费送货、轻松退货和 365 晚睡眠试用。从第一晚开始，你就会感受到不同。这是真正的舒适与全身支撑。这就是为什么 Leesa 被《纽约时报》Wirecutter 评为最佳混合床垫和最佳记忆泡沫床垫，并被 Westm 作为他们的首选床垫合作伙伴。访问 leesa.com 享受床垫高达 30% 的折扣，并使用促销代码 PropG 额外获得 50 美元折扣。我们的听众专享。那就是 leesa.com，促销代码 propg，享受床垫 30% 折扣外加额外 50 美元折扣。在结账后的调查中告诉他们是我们推荐的。leesa.com 促销代码 propg。

Support for the show comes from Leesa. Here's some breaking news. Sleep is good for you, and one way to make sure you're getting the best sleep possible is a good mattress. That's why Leesa makes premium, beautifully crafted mattresses tailored to how you personally sleep. That way, you can wake up well rested, comfortable, and ready for the day. Leesa mattresses are meticulously designed and assembled in the USA for exceptional quality. Each one is designed with specific sleeping positions and feel preferences in mind. Plus, they back it all up with free shipping, easy returns, and a 365 night sleep trial. From the first night, you'll feel the difference. It's serious comfort with full body support. That's why Leesa has been awarded best hybrid and best memory foam mattresses by the New York Times Wirecutter and is featured by Westm as their go-to mattress partner. Go to leesa.com for up to 30% off mattresses plus get an extra $50 off with promo code PropG. Exclusive for our listeners. That's leesa.com promo code propg for 30% off mattresses plus an extra $50 off. Let them know we sent you in the survey after checkout. leesa.com promo code propg.

本节目由 USA Facts 赞助。我采访过无数政治家，如果有一件事让我们都感到不安，那就是他们用偏见和半真半假的事实来歪曲事实。让我们休息一下。所以当我有机会与保守派播客本·夏皮罗辩论时，我们的自然反应是：“放马过来。”为什么？因为这是由 USA Facts 赞助的，每个辩论环节都以政府数据和事实的展示开始，就业、政府支出、收入数据、住房等等。主题是：美国梦。它仍然触手可及吗？我们争论了，我们找到了一些共同点。我们确实进行了辩论，没有个人攻击。在 USA Facts 的 YouTube 频道上观看完整辩论。我们认为你会发现它不同于任何其他辩论，既有娱乐性又有信息量。我们希望你去看看。

Support for the show comes from USA Facts. I've interviewed countless politicians and if there's one thing we all find disconcerting is when they butcher the facts with some bias halftruth. Give us a break already. So when I had the chance to debate conservative podcaster Ben Shapiro, our natural response was, "Bring it on." Why? because it was sponsored by USA Facts and each debate segment started with a presentation of government data and facts, employment, government spending, income data, housing, and more. The topic, the American dream. Is it still within reach? We argued, we found some common ground. We actually debated and there were no personal attacks. Watch the full debate on the USA Facts YouTube channel. We think you'll find it unlike any other debate that is entertaining and informative. We hope you check it out.

呃，每隔几个月你就会创建一份指南，我认为这是关于哪个 AI 最适合哪个任务的开创性指南。你能给我们一个简要或预览，带我们了解一下，比如你知道这个 AI 最适合这个，也许还有一些令人惊讶的和我们甚至没想到的类别。

Uh, every few months you create a guide which I think is sort of the seminal guide for which AI is best for which task. Can you give us a bit of a cliffnotes or a preview and run us through like you know this AI is best for this and maybe a few surprising ones and categories we don't even think of.

当然。我的意思是，你知道，有一个不同的……有一个你想参与的难度级别。

Sure. I mean, you know, there's a different there's a level of difficulty you want to engage with.

So, at this point, if you're really technical and you're using open weights models and you've got an open claw running on your machine, more power to you. There's nothing wrong with that, but that's not most people listening. So, the unfortunate truth for most people is you've got to pick either OpenAI or Anthropic or maybe Google. Google now is a good model, but is sort of catching up and you need to pay 20 bucks a month. There's just kind of no way around that. Once you do that, your life gets pretty easy. The main thing you have to realize is the default models you're being given, especially if you pay for free but even if you pay, are low-end models. So you have to kind of know that for Anthropic the ranking is Fable and then Opus and then Son and Haiku, and for OpenAI it's Pro and then Astra and then Soul and Luda. They don't make it easy for you, right? And then you also have to realize you can manipulate the thinking level of these models and that's actually very important. The more thinking they do, the longer the answer takes but the better the work you get. So, you kind of have to do a little bit of choosing yourself about how hard a problem you're facing. Medical problems, turn up that thinking level, right? Financial problems, turn up that thinking level. Where should I go to a cafe, less important, right? And so you really want to pick one of those three. And generally that would probably be Anthropic or OpenAI at the time we're doing this, but Google again has a good model, but not yet all the stuff around it that makes it useful. And you're going to need to pay for it. And then you should experiment, right? Like the key is to start using these things. So voice mode is very powerful. OpenAI tends to lead on voice things. If you want to chat with your AI, OpenAI does best at that. If you want a really smart model, but you don't care about the chatting piece, Anthropic has very good ones, but you kind of have to pick one of those and everybody else selling you something is selling you a worse version. The two exceptions are your personal agent. Muse is a very good way to get started. The model that powers Muse is not as powerful as the model that powers Dot. But it's more accessible. So you can expect a slightly higher error rate. I'd be a little more cautious about trusting it on everything, but it's still quite good. So, those are two ways to start with personal agents that are quite good also.

So, I'm going to give you one personal hack that's been a real unlock for me, and then I'd love to know if you have any others. I just gave Claude Pro, I think, permission to access my Gmail. And it's just, it sounds obvious, but it's just been mind-blowing. I get an email from someone saying, you know, a capital call on a fund I invested in three years ago or I hear from someone and I recognize the name but I don't or I get a bill for something and I just say what is this, look at my Gmail and it says well this is this bill from this service that was provided in 2021 from this person. It's just, you know, just as these LLMs and Google crawl the web, you typically they can't access your email. And I have found that opening up the corpus of my email for the last 10 years, there's just a ton of information that when it's crawlable and sortable is just incredibly utile. Any thoughts on similar types of hacks or unlocks?

Yeah, let me tell you the next step for that which is talk to either Claude or OpenAI or your cloud system and tell it you want to set up a scheduled task once or twice a day that goes through your email decides what's and you'll have a conversation with the AI to set this up. It can set up the task for you but I have ones that go through my email twice a day and actually think about what's important, what needs to be sent to people. I don't let it write personal emails for me, but I do let it write drafts for business emails. So it'll have my drafts in my inbox ready to go. I edit them and change them. And I've worked with it on my writing style. Again, not for anything personal or important, but for sort of all the minor stuff back and forth. It reads through everything I do a couple times a day, inbox and outbox, and then gives me sort of what should I be paying attention to, what's an issue. Once a week, I have it scrape through all my financials and tell me, am I missing anything? Is there anyone taking advantage of me? Are there subscriptions I should cancel? And it gives me the links to do that. So once you've hooked this up to your Gmail, setting up scheduled tasks a couple times a day or once a week is a really big deal. The agent models, right, Dot and Muse sort of do this proactively, but they sort of do it one-off. You really want a sort of sweep. So scheduled tasks and you can just ask the to help you set one up for you and then you work back and forth with it to make it really good. Cannot recommend highly enough.

That's great. So let's talk specifically about these companies in the markets. Anthropic told prospective investors in its Q3 revenue had jumped nearly 50% to 17 billion claiming the company is now profitable on an adjusted basis. God, I hate that. Adjusted. I love that. Profits before everything else. Meanwhile, an OpenAI engineer has been accused of spying on the company on behalf of the Saudi government. What do you make of the current state of the two major players in terms of their horse race and these impending kind of $2 trillion IPOs?

I think it helps to think about these companies as like two things. They have they've gotten their hands on the most powerful technology anyone's come up with in decades maybe centuries right like LLMs are unreasonably effective and the researchers are really good at building new ones. The LLMs help build new ones. They have they're on a rocket ship where these systems are really good and keep getting better, right? And we can talk about where the value of intelligence is or other stuff we've talked about a little already. And then they're also a startup company that has grown incredibly large, incredibly quickly that has all the problems you'd expect a startup company to have, right? There is mission creep and also political infighting and lack of clarity and financial game is being played left and right. And I think these two things coexist kind of loosely with each other. Right? On one hand, you have an underlying business that is very expensive, right? They're probably profitable on inference. They are probably they're definitely not profitable on with training thrown in, right? And this kind of complexity, but you're also betting on a future where they keep creating more and more super intelligent models.

从某些方面来说，这种分歧真的很难处理，因为你有多相信 AGI 的存在，也就是比人类更聪明的机器？你有多相信这是一家我们应该正常评估的普通企业，而不是一家非凡的企业？

And it's a really hard divergence to deal with in some ways because how much do you buy the existence of AGI, a machine smarter than a human? How much do you buy that this is a normal business that we should evaluate normally versus an extraordinary one?

这真的很复杂。然后还有你在这里谈到的全球性问题，比如现在地缘政治很重要，对吧？而世界上最有价值的东西可能装进一个拇指驱动器。就像如果有人获得了这些模型的权重，那就是他们复制它们所需要的全部，对吧？所以你同时有很多复杂性在发生。这使得在传统意义上评估这些东西变得非常困难。就像世界正迅速变得依赖这些东西，即使我们还没有在每个案例中充分证明其价值。而且收入也在指数级增长。我的意思是，这是一个复杂的情况，使得很难进行传统的财务估值，同时这些公司也在玩所有 IPO 公司都会玩的游戏，你知道，这也掩盖了真正发生的事情。

It's really complicated. Then we have the global issues you're talking about here of like now geopolitics matters, right? And they're the most valuable thing in the world could fit into a thumb drive. Like if someone gets the weights for these models, that's all they need to copy them, right? And so you have a lot of complexity happening all at once. It makes it really hard to value these things in a traditional sense. Like the world is becoming dependent on these things very quickly even as we haven't fully proved out the value in every case. And the revenues are going exponential too. I mean it's a complicated situation that makes it really hard to do kind of traditional financial valuation and at the same time the companies are playing all the games that an IPOing company does that you know obscures what's really happening as well.

如果你必须总结你书中关于学习如何应对人工智能下一阶段的建议，你会从投资者角度、消费者角度以及政府角度怎么说？

If you had to sum up the advice in your book for learning how to navigate this next phase of AI what would you say from an investor perspective from a consumer perspective um and from a government perspective?

所以我认为我首先要对所有人说的是，你低估了这些系统能做什么。你不需要未来的发展来实现大规模变革。

So I think the first thing I would say overall to everybody is you're underestimating what these systems could do. You don't need a future development for large scale change.

低估了，但确实有这些系统今天能做什么以及它们在不久的将来能做什么，就工作的质量和数量而言

Underestimating but there is what what these systems could do today and what they will be able to do in the near future um in terms of the quality and amount of work

这需要时间来吸收，但我担心投资界和商界的许多人把人工智能视为一种投资，而没有充分考虑无处不在的低成本智能会带来什么变化或变革。我认为变革性的机会远远超出公司本身。所以我不知道 Anthropic 是会崛起还是衰落。我不知道数据中心的债务融资模式是否有问题。我无法告诉你那个问题的答案，但无论哪种情况，人工智能都不会消失，对吧，而且能力过剩仍然相当大。所以我会思考超越仅仅投资 Anthropic 的想法，开始思考接下来会发生什么。

That will take time to absorb but I worry that the invest like a lot of people in business and the investing world are focusing on AI as an investment and not thinking enough about what changes or transforms as a result of ubiquitous intelligence at low cost everywhere. Um and I think that that the transformationational opportunities are far beyond the companies themselves. So I don't know whether anthropic rises or falls. I don't know whether the debt financing model for you know for for um you know for data centers is problematic or not. Like I can't tell you the answer to that question but AI is not going away in either case right and the capability overhang is still quite large. So I'd be thinking one step beyond just the idea of like do I invest in anthropia and start thinking about what happens next.

我要说的第二点，对于使用这些系统的个人来说，你必须尝试将它们用于一切你能做的事情，以弄清楚锯齿状前沿的形状。你需要尽可能雄心勃勃。在使用这些东西时要最大化，以弄清楚它们擅长什么或不擅长什么。然后你必须做出非常坚定的选择，关于何时使用人工智能以及何时不使用。这需要意志力。需要一些自我审视，但你必须选择你想在哪些方面保持人性，哪些方面不。嗯，这也是件大事。

The second thing I would say for individuals using these systems is you have to try and use them for everything that you can to figure out the shape of the jagged frontier. And you need to be as ambitious as possible. Be maximalist in how you're using these things to figure out what they're good or bad at. And then you have to make really strong choices about when you use AI and when not to. This will take willpower. It will take some self-examination, but you have to make choices about what you want to maintain humanity for and what you don't. Um, and that is also kind of a big deal.

在商业方面，领导者需要开始决定如何使用人工智能。这意味着要改造他们的公司，但也要决定如何增强而不是仅仅替代劳动力。因为我认为替代劳动力会非常诱人，但实际上不仅对社会危险，对涉及的公司也危险。比如想想如果你是一家做营销工作的公司，你在一月份把所有的营销语言都换成了 Claude，在任何人知道 Claude 语言之前，就在它充斥 LinkedIn 和你读的每篇文章之前。当时这似乎是个好主意。现在每当你读到那些东西时，你会感到恶心，就像它承载着重量，我们都知道这种 Claude 语言，所以你解雇了所有营销人员的事实可能会伤害你，因为你和大家一样。所以我认为这些就是建议，他们意识到这真的很棒，但锯齿状前沿仍然存在，但我们必须努力弄清楚下一步在指数级上走向何方，而不是满足于现状。

And on the business side, leaders need to start making decisions about how AI is going to be used. That means transforming their companies, but also making decisions about how they're going to augment rather than just replace labor. Because I think replacing labor is going to be very tempting, but is actually dangerous not just to society, but also to companies involved. Like think about if you're a company that was doing marketing work and you switched all of your marketing language to Claude back in January before anyone knew Claude speak right before it was filled LinkedIn and every article you read. It probably seemed like a really good idea at the time. And now anytime you read that stuff you get nauseous like it's carrying the weight like we all know this clawed language and so the fact that you fired all your marketers is probably going to hurt you because you have the same company as everybody else. So I think those are the like the piece of advice they're realizing this is really great but the jagged frontiers remain but we have to prot we have to be trying to figure out where we go next on an exponential rather than satisficing ourselves with where we are today.

我想转到一些听众问答，提前提交的问题。第一个问题来自 Jake。我 20 多岁，在我的朋友和网上有很多对人工智能的反对声音。与此同时，在工作中，我们被告知需要全面拥抱它，否则就会落后。你认为年轻人的怀疑是合理的吗？我们如何应对这两种压力？

I want to move on to some listener Q&A uh submitted ahead of time questions from our listeners. The first question comes from Jake. I'm in my 20s and there's a lot of backlash against AI among my friends and online. Meanwhile, at work, we're told we need to embrace it all or fall behind. Do you think young people's skepticism is justified? How do we navigate these two pressures?

我的意思是，我认为人工智能有很多让人不舒服的地方，对吧？人工智能实验室的负责人告诉我们，他们可能会杀死我们所有人或取代所有工作，这无济于事。很难支持这一点，对吧？嗯，你知道，寡头政治和金钱控制，对吧？或者政府与行业之间的勾结。我的意思是所有常见的东西都适用，再加上真实的环境问题和一些不真实的问题。比如在真实方面，电力是一个真正的问题。在不真实方面，与数据中心的其他轻工业相比，用水并不是一个大问题。它们变得更高效，但它变成了一个捆绑的东西，背后有一种社会感觉。正如我告诉你的，人们说的和做的之间存在差距。所以我认为人工智能名声很差，但人们疯狂地使用个人智能体。所以会有类似国会的情况，每个人都讨厌国会，但每个人都爱他们的代表。我认为我们会在人工智能领域看到同样的事情发生。

I mean, I think that there's a lot of things about AI that are uncomfortable, right? It doesn't help that the heads of the AI labs are telling us that they might kill us all or remove all jobs. Like, it's hard to be behind that, right? Um, you know, oligarchy and control of money, right? or the collusion between government and industry. I mean all the usual stuff applied plus a mix of real environmental concerns and kind of not real ones. Like on the real side um electricity is a real concern. On the not real side water use is not really a big issue compared to other light industries in in data centers. They're becoming more efficient but it becomes a bundled thing right that it gets a social feeling behind it. And as I sort of told you there's a sort of gap between what people say and what they do. So I think AI has a very bad reputation but people use the personal agents like crazy. So there's going to be it's kind of like Congress, right, where everyone hates Congress and everyone loves their representative. I think we're going to see the same sort of thing happen in the AI space.

我认为，即使在我教年轻人这些东西时，我发现他们实际上想学习如何更好地使用人工智能。他们想负责任地使用它。他们想塑造事情发展的方向。对我来说，人工智能的危险反应是人们说他们假装它消失。他们会把头埋在沙子里，说：“这一切都会崩溃。这不是真的。我不必关注它。”我的意思是，你知道，用一句名言，对吧？你可能对人工智能不感兴趣，但人工智能对你感兴趣。就像，你必须喜欢我们为年轻人塑造这个方向的机会来自于实际积极行动、创新。所以，意识到这是真实的。所以，我觉得我们需要更多围绕这个的社会运动。我认为我们需要让这些运动围绕真实的东西。我也认为这是一个尝试建立你想要乌托邦的机会。我有点失望，我没有看到更多对人工智能感兴趣的年轻人思考如何利用这个来建立我们想要的世界，一个更平等、更公平、对所有人都有更好结果的世界。

I think that um what I find even among young people when I teach them about this stuff is they actually want to learn how to use AI better. They want to be responsible using this. They want to shape the direction things come in. And to me, the reaction that's dangerous with AI is people saying that they're they're going to pretend it goes away. They're going to put their head in the sand and be like, "This is all going to collapse. This isn't real. I don't have to pay attention to it." I mean, you know, to to use a a famous phrase, right? You may not be interested in AI, but AI is interested in you. Like, you have to like our chances to shape this direction for young people comes from actually active acting, innovating with this. So, like realizing this is real. So, feel like I think we need more social movements around this. I think we but we need to make those movements around what's real. I also think this is an opportunity to try and build the utopia you want. I'm a little dis disappointed that I don't see more young people with an interest in AI trying to think about how do we use this to build the world we want that's more equal and more equitable and with better outcomes for everybody.

这也是一个非常积极的工具。所以我认为我们需要从两方面思考，既把它视为一个积极的机会，也要努力降低风险。下一个问题来自 Instagram 上的 Mary，我很喜欢这个问题。我的伴侣用 ChatGPT 来处理我们的分歧。AI 在恋爱关系中能发挥建设性作用吗？

This is an incredible positive tool too. So I think we need to be thinking in both sides of things and viewing this as a positive opportunity as well as trying to mitigate the risks. Next question comes from Mary on Instagram and I love this one. My partner uses ChatGPT to work through our disagreements. Is there a productive role for AI in a relationship?

所以这是另一个我们需要协商、划定界限的领域。目前关于 AI 和咨询的证据总体上是喜忧参半，甚至偏向乐观，但我们并不真正了解结果。有一些早期研究暗示了积极的效果。我们也有很多关于 AI 精神病的轶事和诉讼。很难说。但我的意思是，你把这个东西邀请到你生活的亲密部分。你对此感觉如何？另一方面，正如我在书中所说，人们开始使用 AI 做的第一件事就是写讣告、婚礼祝酒词和儿童故事。就像我们做的最亲密的事情。所以 AI 似乎好的地方和财务建议一样。它给出好的建议，直接了当的好建议。所以如果它有助于你有一个第三方参与，给你好的直接了当的建议，那可能非常有用。同时，你不希望有人去找 AI 寻找在对话中击败你的策略。所以我建议开放，这当然值得一试。很多人通过和 AI 交谈获得了很多帮助，但我认为你应该带着谨慎和焦虑去做，因为你为此放弃了什么。所以，我不认为它本质上是坏的，很多人得到了好的结果，但同时，你既不想服从 AI 告诉你要做的一切，也不想秘密地这样做而不让你的伴侣知道。

So this is another place where we have to negotiate right negotiate boundaries. The evidence right now on AI and sort of counseling is generally mixed to favorable like we don't really know the outcomes very well. There's some early studies that are suggestive positive things. We have a lot of anecdotal and lawsuits about AI psychosis problems. It's kind of hard to know. But I mean you're inviting this thing into your intimate parts of your life. How do you feel about that? On the other hand, as I talk about in the book, one of the first things people start doing with AI are things like obituaries and wedding toasts and children's stories. like literally the most intimate things we do. So the thing about AI that seems to be good is same thing in financial advice. It gives good advice like good straightforward advice. So if it helps you to have a third person involved that gives you good straightforward advice that can be very useful. Um you at the same time you don't want someone sort of going to AI and finding strategies to beat you in conversation. So I would you know advise openness like it certainly is worth trying. A lot of people get a lot of help out of talking to AI to do things, but I think you should do it with an air of caution and anxiety about what you're giving up as a result of this. So, I don't think it's inherently bad and a lot of people get good results, but at the same time, you neither want to obey AI and everything it tells you to do, nor do you want to be doing this secretly without your partner knowing.

我注意到 AI 最多的地方，我很想听听你的想法和见解。我发现我已经上瘾了——这个词不对。我在 Instagram reels 上花了太多时间，据我所知，他们在 AI 上的投入比世界上任何其他公司都多，除了亚马逊。我对此的看法是，我无法克服它预测下一个视频的准确性，它能在正确的时间推送正确的视频。我觉得这几乎令人不寒而栗。它 literally 在读我的心思，知道我想要的下一个 30 秒的多巴胺刺激是什么。你能谈谈你的想法吗？其次，当 AI 开始编程我们的注意力时，这意味着什么？我们最终会走向 idiocracy 吗？我们只是与其他人隔绝，因为没有什么能比得上我们在屏幕上获得的 AI 驱动的多巴胺。

So, the place where I notice AI the most, and I'm I'm love to get your thoughts and color on this. I have found I have become addicted is the wrong word. I'm spending too much time on Instagram reels and I think the there my understanding is they spend more than any other firm in the world with the exception of Amazon on AI. And how I see that is I cannot get over how good it is at predicting exactly the right video at the right time to serve up to me next. I just find it almost chilling. It's literally reading my mind in terms of what's the next 30 secondond, you know, shot of dopa I want. Can you a your thoughts on that? And B, what does it mean when AI can start to program our attention? Do we eventually I mean are we headed towards idiocracy where we just sequester from the rest of humanity because nothing nothing beats the AIdriven dopa we're going to get on our screens.

我的意思是，我担心这个，对吧？我的意思是，你谈论的是老式 AI，机器学习，老式指的是四年前，对吧？传统的预测模型已经这么好了。然后你开始将它们与大语言模型的能力结合，即按需为你创建内容，对吧？所以现在它仍然是在策划。下一阶段是，H3 视频生成器已经能生成视频的速度比你观看的速度还快。所以生成音乐的速度比你听的速度还快。很快就会有按需为你生成的材料，混合了真实和虚假内容。而且非常有说服力，我们只会看到这种趋势加剧。我看不到摆脱这种跑步机的简单方法，除非做出选择，你不想这样做。我认为有一个极限。我认为人们会想要人类互动。我不认为我们会过度到 100% AI 生成的世界。在某个时刻，你会扔下手机，说，我需要出去一会儿，或者至少很多人会这样。但我不知道那个极限是什么，对吧？我不知道。我不知道它的力量有多大。我认为我们必须开始非常小心，或者制定相关政策，因为娱乐就是娱乐。如果我们能真正擅长只为个人提供娱乐，那么甚至最后的纽带也会断裂，对吧？Instagram，你大概还在某种程度上观看真实的人做真实的事情。在某个时刻，这开始断裂，就像你身处自己的世界。我确实担心这个。我认为这是一个合理的担忧。我认为退回到人类空间将是我们的生存策略。就像我们必须想办法更多地聚在一起，即使在晚餐派对正在崩溃的时候。所以我认为趋势正朝着糟糕的方向发展。

I I mean I worry about that, right? I mean and you're kind of talking about old school AI, machine learning, old school being like four years old, right? Which is like traditional prediction models are already this good. And then you start to combine them with what large language models can do, which is create content for you that's on demand, right? So right now it's still curating. The next stage is, you know, the H3 video generators already generates videos faster you can watch it. So generates music faster than you can listen to it. It's going to soon be material generated for you on demand that's a mix of real and fake content. Um, and as are very persuasive, you know, we're just going to see this um amped up. I don't see an easy way out of that treadmill without a choice being made that you don't want to do it. Um, you know, like I think there is a limit. I think people will want human engagement. I don't think we're going to overtreat into 100% AI generated world. At some point, you'll throw your phone down and be like, I need to go outside for a little bit, or at least many people will. But I think I don't know what that limit is, right? I don't know. And I don't know what the power of that is. And I think that we're going to have to start being really careful about this or have policies around this because it, you know, entertainment is is entertaining. And if we can get really good at entertaining just you personally, then not only like then even the last threads kind of snap, right? Instagram, you're presumably watching real people do real things to some extent still. At some point that starts to snap and it's just like you're in your own world. And I do I do worry about that. I think that that's a legitimate concern to have. And I think retreating to human spaces is going to be a survival strategy for us. Like we're going to have to figure out ways to get together more even at a time where dinner parties are collapsing. So I think the trends are heading in a bad direction there.

同意。有没有什么——我想以乐观的基调结束。AI 能做些什么，或者你看到 AI 在做些什么，有助于让人们更多地面对面联系，更多的人类连接？我不是在说 Reddit 或 Discord，就像你。我喜欢你挑战年轻人采取乐观的观点，想办法利用 AI 创造更多的社会公益，而不是一直灾难化。顺便说一句，我认为我们的社区中，学者和思想领袖有一种灾难化的激励。这比说“哦，这是一个有趣的技术，一切都会每天变得更好一点”能带来更高的演讲费。那不会让你被邀请回来做主题演讲。你能谈谈你发现的一些真正有帮助的 AI 相关的事情吗？

Agreed. Is there anything This is I want to end on an optimistic note here. Is there anything that AI can do or that you see AI doing that helps bring together people kind of IRL more human connection? And I'm not talking about Reddit or Discord is like you. I love I love your challenge to young people to take an optimist viewpoint and figure out how to create more social good leveraging AI instead of catastrophizing about it all the time. And by the way, I think there's a incentive in our community amongst academics and thought leaders to catastrophize. It just makes for it just makes for a a bigger speaking fees than saying, "Oh, this is an interesting technology and everything's going to get a little bit better every day." that doesn't get you invited back for a keynote. Can you talk about something or that you've seen that you find really helpful around AI?

我的意思是，我发现无数有帮助的事情，对吧？这就是我的意思，我们有按需智能。所以有很多人喜欢它，它是一个好的第二意见医生，如果你没有医生，它是第一线。它是一个好的律师。它是一个好的财务顾问。它可能，你知道，当你感觉不好或沮丧时，与它交谈是有用的。如果使用得当，它是一个好的导师。就像这些问题我们已经花费了数千亿美元和无限的学术研究试图解决如何更好地教人们，我们知道辅导很棒。我们永远无法做到一对一辅导。所以我们一直在发明其他东西。我们知道人们缺乏医疗信息。我们知道即使是 GPT4 这个早已过时的模型，当提供给肯尼亚企业家时，已经表现良好。

I mean, I find a million things helpful, right? Like this is what I mean, we have intelligence on demand. So, there's tons of people who like it's a good, you know, second opinion doctor, a first line if you don't have a doctor. It's a good lawyer. It's a good financial adviser. It's probably, you know, it's it's useful to talk to when you're feeling bad or down. It's a good tutor when used properly. Like these are problems that we have spent hundreds of billions of dollars and infinite amounts of academic research trying to solve how do we teach people better and we know tutoring is great. We could never do one-on-one tutoring. So we've been inventing all this other stuff. We know that people lack access to medical information. We know even GPT4 long obsolete model right when given to Kenyan entrepreneurs are already doing well.

创业者仅仅从 AI 获取建议就能获得 18% 的盈利提升。有太多机会可以大幅增强能力，而不必去追求未来那种空中楼阁式的场景，对吧？比如 AI 在哪里拯救我们、发明新药？我们现在有一个工具，能提供智能和建议，对那些原本得不到建议的人来说，这些建议相当不错，而且是针对他们的情况定制的。

18% profitability from just getting advice from AI for entrepreneurs. There is so much opportunity to supercharge without going to a future pie-in-the-sky situation, right? Like, where does AI save us and invent new drugs? We have a tool now that provides intelligence and advice that is reasonably good advice to people who can't get it, that's customized for their condition.

我觉得我们常常过多地从我们这些有关系、有资源的人的狭窄视角去思考，思考如何在娱乐需求和 AI 在世界大部分地区所做的事情之间取得平衡。我们终于有了一个能给出好建议的东西。有一组很棒的研究，讲的是泰国的养猪户现在可以给疾病做诊断了。这里有千千万万个小用例，智能曾是让世界变得更好的瓶颈。

I think we tend to think a lot about the narrow viewpoint of those of us who are well-connected and well-resourced, and balancing out our desire for entertainment with AI doing things in large parts of the world. We finally have a thing that gives good advice. There's a great set of studies on pig farmers in Thailand who can now get diseases diagnosed. There are a thousand million little use cases here where intelligence was the bottleneck to making the world a better place.

我们能做到这一点，而我觉得我们对这些事情的讨论还不够，它们需要付出一些努力，需要测试，需要保障措施。但这很令人兴奋。我是说，作为一个长期尝试大规模做教育、一次次碰壁的人，突然之间你有了一个导师，它跨语言、跨能力水平都能用，能匹配你的兴趣水平，能跟踪你哪里做得不好，还能生成图片逗你开心。突然间，一个更好的世界在这里有了真正的可能。即便我们担忧风险，我们也可以抓住这个机会。

And we can do this, and I feel like we don't talk enough about those things, and they require some effort and they require testing and they require safeguards. But this is exciting. I mean, as someone who's been trying to do education at scale for a long time and has run into wall after wall, suddenly you have a tutor that works across languages and ability levels and matches your interest level and can follow what you're doing badly and generate images to amuse you. Suddenly there's real possibility here for a better world. And we can seize that even as we worry about the risks.

Ethan Mollick 是沃顿商学院的教授，著有《共同智能：与 AI 一起生活和工作》，并撰写广受欢迎的 Substack 专栏《一个有用的东西》。他的最新著作《共存：AI 的下一个阶段》现已出版。他从湾区新书巡展的启动活动赶来与我们连线。

Ethan Mollick is a professor at the Wharton School, author of Co-Intelligence: Living and Working with AI, and writes the popular Substack One Useful Thing. His newest book, Coexistence: The Next Phase of AI, is out now. He joins us from the kickoff events for his book tour in the Bay Area.

湾区，在湾区第二好的学校。Ethan，你在伯克利有什么活动吗？

The Bay Area, at the second best school in the Bay Area. Are you doing anything at Berkeley, Ethan?

这次没有。

Not this time around.

你得留点东西给以后嘛。就这样。就这样。Ethan，你被我划出圣诞礼物名单了。Mollick 教授，非常感谢你抽出时间。

You got to save something for later. That's it. That's it. You're off my Christmas list, Ethan. Professor Mollick, really appreciate your time.

感谢邀请。

Thanks for having me.