（中英附文稿）AI不只帮你写文案：吴恩达谈从重复劳动到主动解
字幕摘录
| 时间 | 英文 | 中文 |
|---|---|---|
| 0:00 | There's been a lot of misinformation about AI. | 对AI的误导很多. |
| 0:03 | This is Andrew. | 这是安德鲁。 |
| 0:04 | 他共同创办了Google Brain和Coursera. | |
| 0:07 | His machine learning course has reached millions of learners, | 他的机器学习课程 已经达到了数百万的学习者, |
| 0:09 | and he is one of the most influential voices in AI today. | 他是今天AI最有影响力的声音之一。 |
| 0:13 | So a handful of leading AI companies | 所以,少数主要的AI公司 |
| 0:14 | have been very loud voices of fear-mongering around AI | 在AI周围发出恐惧的声音 |
| 0:18 | to try to get regulations passed. | 试图让 条例通过。 |
| 0:20 | This drumbeat of fear-based messaging | 恐怖讯息的鼓声 |
| 0:22 | has skewed societal's perception | 扭曲了社会观念 |
展开字幕全文（939 条）
| 序号 | 英文 | 中文 |
|---|---|---|
| 1 | There's been a lot of misinformation about AI. | 对AI的误导很多. |
| 2 | This is Andrew. | 这是安德鲁。 |
| 3 | 他共同创办了Google Brain和Coursera. | |
| 4 | His machine learning course has reached millions of learners, | 他的机器学习课程 已经达到了数百万的学习者, |
| 5 | and he is one of the most influential voices in AI today. | 他是今天AI最有影响力的声音之一。 |
| 6 | So a handful of leading AI companies | 所以,少数主要的AI公司 |
| 7 | have been very loud voices of fear-mongering around AI | 在AI周围发出恐惧的声音 |
| 8 | to try to get regulations passed. | 试图让 条例通过。 |
| 9 | This drumbeat of fear-based messaging | 恐怖讯息的鼓声 |
| 10 | has skewed societal's perception | 扭曲了社会观念 |
| 11 | to be really negative on AI. | 对AI非常负面。 |
| 12 | People talk to me about data centers and job loss. | 人们和我谈论数据中心和失业问题。 |
| 13 | Maybe AI could do 30, 40 percent of many jobs. | 也许AI能做30%,40%的工作。 |
| 14 | And what that means is, well, that 60 percent that a human does | 这意味着 人类60%的工作 |
| 15 | has become even more valuable. | 已经变得更加宝贵。 |
| 16 | What about loss of human control over AI? | 对AI失去人的控制怎么办? |
| 17 | I think about something else that we can't control. | 我想到别的东西,我们不能控制。 |
| 18 | I think you're one of the voices in AI | 我觉得你是AI里的声音之一 |
| 19 | who comes with a huge background | 有着巨大背景的人 |
| 20 | and machine learning and teaching AI, | 以及机器学习和教学 |
| 21 | and also you're a positive voice, | 而且你是一个积极的声音, |
| 22 | because this is something that I've been seeing, | 因为这是我见过的 |
| 23 | especially this summer, how poor lives the society has become, | 特别是今年夏天 社会的贫穷生活 |
| 24 | especially on social media, when I'm posting about AI | 尤其是在社交媒体上 当我发布AI时 |
| 25 | and people talk to me about data centers and job loss. | 人们和我谈论数据中心和失业 |
| 26 | Why do you think this wave started recently? | 你觉得这波为什么是最近才开始的? |
| 27 | What do you think the causes are? | 你觉得原因是什么? |
| 28 | There's been a lot of misinformation about AI, | 关于AI有很多错误信息, |
| 29 | and the root cause of a lot of this is an unfortunate attempt | 原因很多都是不幸的尝试 |
| 30 | that started two, three years ago | 两三年前开始的 |
| 31 | of, I think, PR and registry capture. | 公关和登记记录 |
| 32 | It turns out that one of the most valuable things | 原来最有价值的东西之一 |
| 33 | in AI right now is the giant AI models, | 在AI现在 是巨大的AI模型, |
| 34 | giant large-language models that some companies have trained. | 一些公司训练的大型大语言模型。 |
| 35 | But if you spend billions of dollars training a model, | 但如果你花几十亿美元训练一个模特儿 |
| 36 | it's really inconvenient if someone else trains a model | 如果别人训练一个模特儿 实在太不方便了 |
| 37 | and wants to give it to anyone in the world to use for free. | 并想把它给世界上的任何人免费使用. |
| 38 | So a handful of leading AI companies, I think, as you know, | 所以,一些主要的AI公司,我认为,你知道, |
| 39 | have been very loud voices, fear-mongering around AI | 在AI周围发出非常响亮的声音 恐惧的恐惧 |
| 40 | to try to get regulations passed, | 试图通过条例, |
| 41 | to create an unfair playing field that favors incumbents | 创造一个有利于现任者的不公平竞争环境 |
| 42 | so that we all have to pay a high toll for use of AI, | 因此,我们都必须付出高昂的代价来使用AI, |
| 43 | while stymieing the other teams, | 一方面阻击其他队伍, |
| 44 | be it researchers or other companies, | 无论是研究者还是其他公司, |
| 45 | that want to just give away open way to open source models | 想要给开源模型开辟道路 |
| 46 | that anyone could use much cheaper. | 任何人都可以使用便宜得多。 |
| 47 | Unfortunately, fear-mongering works. | 不幸的是,恐惧的侵扰是有效的。 |
| 48 | When you go and say AI is like nuclear weapons, | 当你去说AI就像核武器, |
| 49 | which is an analogy that has no basis in fact, | 这是没有根据的比喻, |
| 50 | what do they have to even do with each other? | 他们之间有什么关系? |
| 51 | Or when you go around and cherry pick cases of AI, | 或者当你绕圈子去 和樱桃挑AI的案件, |
| 52 | you know, making a misstep | 做错事 |
| 53 | and make it much bigger than it is. | 让它比它大得多。 |
| 54 | Or even spread misinformation about how AI uses data centers, | 甚至散布关于AI如何使用数据中心的错误信息, |
| 55 | uses a lot more water than the actual reality. | 用的水比现实多得多 |
| 56 | This drumbeat of fear-based messaging | 恐怖讯息的鼓声 |
| 57 | has skewed societal's perception | 扭曲了社会观念 |
| 58 | to be really negative on AI, which is unfortunate | 对AI非常负面,这是不幸的 |
| 59 | because this is slowing down American adoption in AI. | 因为这减缓了美国在AI的收养速度. |
| 60 | This is making America less competitive. | 这使得美国的竞争力降低. |
| 61 | And unless we get the truth about AI out there, | 除非我们知道AI的真相 |
| 62 | which is that it's fantastic benefit with some problems, | 这对解决一些问题有好处 |
| 63 | but not nearly the degree to which they're blown up to be, | 但不是被炸得那么严重 |
| 64 | it will hurt individuals. | 它会伤害个人。 |
| 65 | I'm gonna read out some of the problems | 我要宣读一些问题 |
| 66 | that people are highlighting, job loss and inequality. | 人们正在强调,失业和不平等。 |
| 67 | What do you think? | 你觉得呢? |
| 68 | The job apocalypse, the job pocalypse, | 工作启示录, 工作浮夸, |
| 69 | this idea that AI will take over 50% of jobs, | 这个想法,AI将占据50%的工作, |
| 70 | people will be out of work, rioting in the streets, | 人们将失业,在街头暴动, |
| 71 | that's just not gonna happen. | 那只是不会发生。 |
| 72 | With every wave of technology, including AI, | 随着每一波技术,包括AI, |
| 73 | the skills we need to do great work shifts, | 我们所需要的技术 做伟大的工作转变, |
| 74 | and so AI is changing job professions. | 所以AI正在改变职业 |
| 75 | But boy, I wish AI just doesn't work well enough. | 但是,孩子,我希望AI 只是工作不够好。 |
| 76 | I know that handful of businesses want to hype up AI | 我知道有几家企业想哄骗AI |
| 77 | to say, we have super intelligence, | 说,我们有超级智能, |
| 78 | or we'll have artificial intelligence or whatever, | 或者我们有人工智能之类的 |
| 79 | and can do all the stuff that humans do. | 并且可以做人类做的一切 |
| 80 | I wish AI work better, which is not good enough | 我希望人工智能工作做得更好 这还不够好 |
| 81 | to make AI do everything a human does. | 让人工智能做人的一切 |
| 82 | And if you look at the analysis of jobs, | 如果你看看工作的分析, |
| 83 | economists, like my friend Eric Brinovson at Stanford, | 经济学家,像我的朋友埃里克·布里诺夫森 在斯坦福, |
| 84 | Andy McAfee at MIT, | 麻省理工学院的安迪·麦卡菲 |
| 85 | economists have analyzed many people's jobs | 经济学家分析了很多人的工作 |
| 86 | and by breaking it down into individual tasks, | 并把它分成个人任务, |
| 87 | and maybe AI could do 30, 40% of many jobs. | 也许AI能做30%,40%的工作。 |
| 88 | And what that means is, well, that 60% that a human does | 也就是说,人类60%的工作 |
| 89 | has become even more valuable | 越来越珍贵 |
| 90 | because it's called an economic compliment, | 因为这叫做经济赞美, |
| 91 | so the 30, 40% is now cheaper. | 因此,30%,40% 现在更便宜。 |
| 92 | And so what will happen is people that use AI, | 接下来是使用人工智能的人 |
| 93 | maybe people that use AI will replace people | 也许用人工智能的人会代替别人 |
| 94 | that don't use AI, but AI is not in a position | 但是AI没有地位 |
| 95 | for the vast majority of jobs to replace people. | 给绝大多数的工作来代替人。 |
| 96 | Of all the different professions, | 在不同的职业中 |
| 97 | the one that's most affected by AI now is software engineering | 现在受AI影响最大的是软件工程 |
| 98 | because AI is actually fantastic at writing code. | 因为人工智能在写代码方面的确很棒 |
| 99 | And what we see is that the number of job openings | 我们所看到的是 职位空缺的数量 |
| 100 | in software engineering is up, | 在软件工程上, |
| 101 | contrary to what the doom fear mongers would say, right? | 和末日的恐惧相反 对吧? |
| 102 | AI is not actually able to replace software engineers, | AI实际上无法替换软件工程师, |
| 103 | and all the good software engineers I know | 还有我认识的所有优秀的软件工程师 |
| 104 | are busier than ever. | 比以往任何时候都更忙。 |
| 105 | Now, the flip side of it is, | 现在,它的反面是, |
| 106 | if someone's still write code like this 2022 | 如果有人还写这样的代码2022年 |
| 107 | before track GPT, they're in trouble. | 在追踪GPT之前 他们有麻烦了 |
| 108 | They need new skills, like don't do stuff | 他们需要新技能,就像不做东西一样 |
| 109 | that 30, 40% of AI can automate. | 30,40%的人工智能可以自动化. |
| 110 | You've got to start doing that, let AI do that, | 你必须开始这样做, 让AI做到这一点, |
| 111 | but then gain your skills to do the other 60, 70% | 然后获得技能 做其他60,70% |
| 112 | that AI cannot do. | 这是大赦国际不能做的。 |
| 113 | What would your advice be to new graduates? | 你给新毕业生的建议是什么? |
| 114 | Cause I talked to Eric on this podcast | 因为我跟艾瑞克谈过这个播客 |
| 115 | and he was talking about that there's not really | 他当时在说真的没有 |
| 116 | a lot of impact on the job market, | 对就业市场影响很大, |
| 117 | except for, I think he mentioned people from 18 to 25, | 除了,我想他提到 从18到25, |
| 118 | who just graduated. | 谁刚毕业。 |
| 119 | What would be your advice to those people | 你对这些人的建议是什么? |
| 120 | who don't have the expertise | 谁没有专长 |
| 121 | maybe to strategize in their job yet? | 也许他们的工作有策略吗? |
| 122 | They can only do manual work that AI can do as well. | 他们只能做人工工作,而AI也可以做. |
| 123 | So one real challenge for fresh college grads | 所以,一个真正的挑战 对于新鲜的大学毕业 |
| 124 | is that the university system is slow to adapt. | 大学系统适应性很慢 |
| 125 | And so, I love academia. | 所以,我喜欢学术界。 |
| 126 | I think we should all support academia and universities. | 我认为我们都应该支持学术界和大学。 |
| 127 | And when AI comes and transforms | 当AI出现并改变 |
| 128 | the way software is written, | 写软件的方式, |
| 129 | universities often take like a year or two | 大学通常要花一两年时间 |
| 130 | for the faculty to master the skills, | 以便教职员掌握技能, |
| 131 | then create new courses, | 然后创造新的课程, |
| 132 | get curriculum committee approval or whatever, | 得到课程委员会的批准或什么的, |
| 133 | get the faculty senate to vote. | 让参议院投票 |
| 134 | It just takes years. | 只是需要几年时间 |
| 135 | And that speed of change in academia | 学术界的变革速度 |
| 136 | is very poorly matched to the speed of change in AI. | 与AI的改变速度非常差。 |
| 137 | So sadly, many universities are still teaching students | 可惜许多大学还在教学生 |
| 138 | to be ready for the jobs of 2022, | 准备好2022年的工作, |
| 139 | when we shouldn't even be teaching them | 我们不应该教他们 |
| 140 | for the jobs of 2026, | 为2026年的工作, |
| 141 | we should be teaching them for the jobs of 2028 and beyond. | 我们应该教他们2028年及以后的工作 |
| 142 | And what this means is the job openings are there. | 这说明职位空缺就在那里。 |
| 143 | Tons of employers I know just can't find enough skill, | 我知道很多雇主都找不到足够的技能 |
| 144 | you know, people at any level of seniority. | 你知道,任何级别的人。 |
| 145 | But it turns out in my office right now, | 但是现在在我的办公室里 |
| 146 | we have a lot of interns. | 我们有很多实习生 |
| 147 | There are current college students, fresh college grad, | 有现在的大学生, 新鲜的大学毕业, |
| 148 | we also have one high school intern, | 我们还有一个高中实习生, |
| 149 | and they're amazing and productive. | 和它们惊人和生产。 |
| 150 | But the key is they're all very AI native. | 但关键是他们都是AI本地人 |
| 151 | They all use AI tools to do the things that I could do, | 他们都用人工智能工具来做我能做的事, |
| 152 | but then also, you know, | 还有,你知道, |
| 153 | lean in to doing the things that humans can do | 做人类能做的事 |
| 154 | that AI can't for a long time. | AI不能在很长一段时间。 |
| 155 | So there's plenty of work for people to do. | 故众生有余功德. |
| 156 | But so my advice to fresh college grads | 但我建议大学毕业 |
| 157 | or the people currently in college is, | 或者目前大学里的人 |
| 158 | by all means, work hard in classes, you know, | 在课堂上努力工作 |
| 159 | get good grades, learn from the instructors, | 学好成绩 向老师学习 |
| 160 | but to the extent that there are still additional skills | 但只要还有更多的技能 |
| 161 | that the university has not yet adapted to teaching, | 大学还没有适应教学, |
| 162 | then find other ways to learn online, | 然后找其他方法在网上学习, |
| 163 | 无论是来自Coursera还是深入学习AI还是Udemy | |
| 164 | or other places where you can gain the more cutting edge skills, | 或其他可以获得更先进技能的地方, |
| 165 | especially AI skills, | 特别是人工智能, |
| 166 | that universities have not yet worked on the curricula. | 大学尚未制定课程。 |
| 167 | Quick pause, because what Andrew just said about workflows | 暂停一下,因为安德鲁刚才说的工作流程 |
| 168 | connects to something that HubSpot just put out for free. | 连接到HubSpot刚刚免费推出的东西. |
| 169 | The thing is, the real opportunity right now | 问题是,真正的机会,现在 |
| 170 | is not just in the models themselves, | 不只是模型本身, |
| 171 | but in what you build on top of them, | 但你们在他们的上面所创造的, |
| 172 | how you turn AI from a chat window | 从聊天窗口打开AI |
| 173 | into workflows that actually do work for you. | 进入工作流程 真正为你工作。 |
| 174 | And the good news is, | 好消息是,我们 |
| 175 | you don't have to start by building a full agent. | 你不必开始 建立一个完整的代理。 |
| 176 | You can start much simpler with better prompts. | 你可以用更好的提示开始简单得多。 |
| 177 | HubSpot刚刚发布了高级聊天GPT | |
| 178 | Prompt Engineering Playbook, and the idea is very simple. | 即时工程游戏本 这个想法很简单 |
| 179 | In seven days, it helps you move from getting generic AI answers | 7天后,它能帮助你摆脱普通AI的解答 |
| 180 | {\fn黑体\fs22\bord1\shad0\3aHBE\4aH00\fscx67\fscy66\2cHFFFFFF\3cH808080}你更能保持一贯 {\fn黑体\fs22\bord1\shad0\3aHBE\4aH00\fscx67\fscy66\2cHFFFFFF\3cH808080}我很高兴你 | |
| 181 | and useful options, because we've all had this moment. | 和有用的选择,因为我们都有这一刻。 |
| 182 | You type in one vague sentence, | 你输入一个模糊的句子, |
| 183 | get something mediocre back, | 拿些平庸的东西回来 |
| 184 | and then spend the next 20 minutes fixing it yourself. | 然后再花20分钟自己修 |
| 185 | This playbook is basically built for that exact problem, | 这本游戏本基本上是为这个问题而建的 |
| 186 | and it walks you through a single progression. | 它让你经历了一个单一的进化。 |
| 187 | First, you learn how to structure prompts | 首先,你学会如何构建提示 |
| 188 | with things like role, context, format, and parameters. | 包括角色、上下文、格式和参数。 |
| 189 | Then it gets more advanced. | 然后它会更先进。 |
| 190 | Few shot examples, chain of thought reasoning, | 很少有例子 思想推理链 |
| 191 | more precise prompt structures, | 更精确的快速结构, |
| 192 | and ways to make outputs more reliable. | 以及如何使产出更加可靠。 |
| 193 | What I like is that it doesn't stop at individual prompts. | 我喜欢的是它不会停止 个别的提示。 |
| 194 | It also shows you how to build reusable AI personas, | 也告诉你如何建造可重复使用的AI人物, |
| 195 | modular prompt components, | 模块快速组件, |
| 196 | and eventually your own signature prompt engineering system. | 最后你自己的签名即时工程系统。 |
| 197 | So instead of starting from scratch every time, | 所以不是每次从零开始 |
| 198 | you're building a repeatable way to work with AI, | 你正在建立一个可以重复的方式 与AI合作, |
| 199 | whether you're creating content, analyzing data, | 你是否正在创建内容 分析数据 |
| 200 | or making business decisions. | 或做出商业决定。 |
| 201 | The playbook is free, links in the description. | 游戏本是免费的,在描述中的链接. |
| 202 | Thanks to HubSpot for sponsoring this video. | 感谢HubSpot赞助了这段视频. |
| 203 | And now back to our conversation with Andrew. | 现在回到我们和安德鲁的谈话 |
| 204 | I want to say one other thing. | 我想说另外一件事。 |
| 205 | It turns out, if you look at the skill map changes, | 事实证明,如果你看 技能地图的变化, |
| 206 | one of the most important changes is | 最重要的变化之一是 |
| 207 | it's so much easier to build with AI than before. | 和AI一起建起来比以前容易多了 |
| 208 | When something becomes much easier, | 当事情变得容易得多, |
| 209 | a lot more people should do it. | 更多的人应该这样做。 |
| 210 | And so now, not only should professional software engineers | 所以现在,不仅专业的软件工程师 |
| 211 | build software with AI, | 与AI一起构建软件, |
| 212 | it's becoming much easier for everyone to build with AI, | 对AI来说,每个人都更容易建起来 |
| 213 | and people that embrace that and do so | 和接受并接受的人 |
| 214 | will be more productive and will accomplish more, | 将会更加丰硕,并且更加成功, |
| 215 | and I think have more fun than the ones that don't. | 我觉得比那些不开心的人更好玩 |
| 216 | And AI lets you build really fast. | 人工智能让你建得很快 |
| 217 | So for people that are not just software engineers, | 所以对于不仅仅是软件工程师的人来说 |
| 218 | but marketers, recruiters, HR professionals, | 但市场营销者、招聘者、人力资源专业人员, |
| 219 | operations specialists, | 业务专家, |
| 220 | I think if they learn to build with AI, | 我想如果他们学会和AI一起建设 |
| 221 | they'll really just do much more, whatever the job role is. | 他们真的会做的更多, 不管工作的角色是什么。 |
| 222 | How do you, by the way, measure the increase in productivity | 顺便说一句,你怎么衡量生产率的提高? |
| 223 | when you deploy AI? | 当你部署AI时? |
| 224 | Do you have a KPI in your company? | 你的公司有KPI吗? |
| 225 | I wish it was a simple answer. | 我希望这是一个简单的答案。 |
| 226 | I find that the business outcome of AI | 我发现AI的商业结果 |
| 227 | is more a function of the business | 更属于企业的功能 |
| 228 | than a function of the AI. | 而不是大赦国际的职能。 |
| 229 | For some, it may be increased customer growth and retention, | 对一些人来说,这可能会增加客户的增长和保留, |
| 230 | or maybe a faster to serve customers, | 或者更快地为顾客服务 |
| 231 | or it may be increased accuracy in some tasks. | 也可能提高某些任务的准确性。 |
| 232 | So the KPI tends to be related to the business | 因此,KPI倾向于与业务有关 |
| 233 | rather than the AI. | 而不是大赦国际。 |
| 234 | So you can't directly measure just AI, | 所以你不能直接测量AI |
| 235 | because it's an interesting thing to do, | 因为它是一个有趣的事情做, |
| 236 | because we've been deploying AI actively in my company, | 因为我们一直在积极部署AI在我的公司, |
| 237 | and I think for me as a media company, | 我认为作为一个媒体公司, |
| 238 | it's probably the amount of views, the output. | 可能是浏览量 输出量 |
| 239 | It's just interesting how different people measure, | 有趣的是不同的人如何衡量, |
| 240 | even revenue, like if you're becoming more effective | 甚至收入,就像如果你越来越有效 |
| 241 | with how you make money. | 你如何赚钱。 |
| 242 | Actually, say more. | 其实,说更多。 |
| 243 | How are you using AI in your business? | 你怎么在你的生意中使用AI? |
| 244 | Oh my God. | 天呐 |
| 245 | First of all, we have Claude for all of us, | 首先,我们有 克劳德为我们所有人, |
| 246 | and we have certain projects | 我们有一些项目 |
| 247 | for every social media that we're on. | 我们所有的社交媒体。 |
| 248 | So for example, for this podcast, | 比如说,对于这个播客, |
| 249 | we have a project that's called Guests, | 我们有个叫"客人"的项目 |
| 250 | and it knows all the analytics from previous guests, | 它知道所有分析 从前宾客, |
| 251 | and it has certain criteria | 而且它有某些标准 |
| 252 | on which we rank every single person | 我们把每个人排在其中 |
| 253 | who comes to the podcast, | 他来见播客, |
| 254 | whether he or she is cited, | 无论他或她被引用, |
| 255 | whether they have a certain opinion on AI, | 他们是否对大赦国际持有某种意见, |
| 256 | whether they've been active with AI in their company, | 他们是否一直活跃 在AI在他们的公司, |
| 257 | or if they're a recent founder in AI. | 或者他们是AI的最近创始人 |
| 258 | So it gives them different weights, | 因此,它给他们不同的重量, |
| 259 | and it comes up with a grade based out of 40. | 并得出一个等级 根据40。 |
| 260 | 40 meaning tier one, 30 meaning tier three, et cetera. | 40表示一级,30表示三级等. |
| 261 | And then we have another one | 礛и临Τ |
| 262 | that analyzes every single podcast, | 分析每个播客, |
| 263 | and gives me tips on how to ask questions. | 并给我如何提问的提示。 |
| 264 | 与Instagram相同,与LinkedIn相同. | |
| 265 | It has my tone of voice, personal dossier, | 它有我的声音,个人档案, |
| 266 | my business strategies. | 我的业务策略。 |
| 267 | So whenever it writes something, | 所以,每当它写的东西, |
| 268 | it knows all the facts about me, how I sound. | 它知道关于我的所有事实 我的声音。 |
| 269 | Every social media is run by a person. | 每个社交媒体都由一个人管理. |
| 270 | So a person makes a strategic call. | 所以一个人做出一个战略号召。 |
| 271 | And by the way, | 顺便说一句 |
| 272 | if you can give me feedback on this, if I can improve. | 如果你能给我反馈, 如果我能改进。 |
| 273 | So what I'm working on right now is closing the loop, | 所以,我现在的工作是 关闭循环, |
| 274 | because sometimes they send me a text, | 因为有时候他们会给我发短信 |
| 275 | and I'm like, oh, we need to change this, this, and that. | 而我喜欢,哦,我们需要改变这个,这个,那个。 |
| 276 | But that happens in a chat, in Telegram. | 但是,这发生在聊天, 在Telegram。 |
| 277 | And we have this feedback. | 我们得到了这个反馈。 |
| 278 | We have a bot that scans all of our chats. | 我们有一个机器人 扫描我们所有的聊天。 |
| 279 | But I really want AI to be able to learn continuously | 但我真的希望AI能够不断学习 |
| 280 | from this feedback to just know my taste better. | 从这些反馈来了解我的品味 |
| 281 | There's one thing I see a lot in AI, | 我在AI里看到很多东西 |
| 282 | which is it turns out for AI's data scientists | 对AI的数据科学家来说 |
| 283 | or AI's brainstorming partner, | 或AI的集思广益伙伴, |
| 284 | it often comes up with one or two good ideas, | 它经常提出一个或两个好主意, |
| 285 | two or three mediocre ones, and four atrocious ones. | 两三个平庸的,四个残暴的。 |
| 286 | And sometimes you wonder, | 有时候你会想 |
| 287 | how could my AI have thought | 我的人工智能怎么会想到 |
| 288 | that could even be a plausible idea? | 这甚至可能是一个可信的想法? |
| 289 | And to me, this relates to the job-pocalypse point of view, | 对我来说,这涉及到 工作启示论的观点, |
| 290 | which is that for a long time, humans, you, me, | 人类,你,我, |
| 291 | everyone watching this, | 大家看这个 |
| 292 | will have a significant context advantage over AI, | 与大赦国际相比,在环境方面具有显著优势, |
| 293 | which is that you know something | 就是你知道一些东西 |
| 294 | that's incredibly obvious to you | 这对你来说是显而易见的 |
| 295 | that that was an awful idea, but the AI did not. | 那是个糟糕的主意 但人工智能没有 |
| 296 | And it turns out that one of the reasons | 结果发现一个原因 |
| 297 | why AI will not replace our jobs or whatever, | 为什么AI不会取代我们的工作什么的, |
| 298 | a lot of times not anytime soon, | 很多时候,不是任何时间很快, |
| 299 | is because humans have a massive context advantage | 因为人类拥有巨大的环境优势 |
| 300 | compared to AI. | 与大赦国际相比。 |
| 301 | We know so much that from our years of experience | 我们从多年的经验中知道这么多 |
| 302 | that we talked to customer, | 我们和顾客谈过 |
| 303 | we saw the funny facial expression that told us, | 我们看到了有趣的面部表情 告诉我们, |
| 304 | ah, they don't like this. | 啊,他们不喜欢这样。 |
| 305 | Or we talked to business or our manager said, | 或者我们跟生意谈了 或者我们的经理说 |
| 306 | hey, blah, blah, blah, I really care about this. | 嘿,嘿,嘿,我真的关心这个。 |
| 307 | And so it turns out that almost all humans, | 结果发现几乎所有的人 |
| 308 | what all humans just know a lot of stuff | 所有人类都知道很多东西 |
| 309 | that the plumbing does not exist, | 使管道不存在, |
| 310 | and I don't think exists for the foreseeable future. | 我认为在可预见的未来不存在 |
| 311 | I know sometimes people talk about the importance | 我知道有时候人们会谈论重要性 |
| 312 | of human judgment or human taste, | 人类的判断或品味, |
| 313 | and sometimes we wonder, all right, what is taste? | 有时候我们想,好吧,什么是味道? |
| 314 | Is this fuzzy thing? | 这是模糊的东西吗? |
| 315 | But to me, the technical thing that underlies | 但对我来说, 技术问题的基础 |
| 316 | why humans have better judgment and better taste than AI, | 为什么人类比AI有更好的判断和品味, |
| 317 | is this context advantage. | 这是上下文的优势。 |
| 318 | And because this is a long-term advantage, | 因为这是一个长期优势, |
| 319 | like no one's gonna solve this, you know, in a few years, | 就像几年后 没人能解决这件事 |
| 320 | this is why we just need a lot more humans | 所以我们需要更多的人 |
| 321 | with that judgment and taste to keep on complimenting AI. | 以判断和品味 继续称赞AI。 |
| 322 | And doesn't this make education even more important? | 这难道不使教育更加重要吗? |
| 323 | Because education gives us context. | 因为教育给我们提供了背景。 |
| 324 | This is another thing I'm hearing about AI, | 这是另一件事 我听到关于AI, |
| 325 | like you won't need education | 就像你不需要教育一样 |
| 326 | because all the information is at your fingertips. | 因为所有信息都在你的指尖上 |
| 327 | You just ask, Chad GPT. | 你只是问,乍得GPT。 |
| 328 | But when you say context and taste, for me, | 但当你说上下文和品味,对我来说, |
| 329 | that's years of acquiring knowledge | 那是多年 获取知识 |
| 330 | and learning from the best and seeing how they perform | 向最好的人学习,看看他们的表现 |
| 331 | versus just asking a chat. | 而不是只是问话 |
| 332 | I'm gonna say something that may be controversial. | 我要说一些有争议的话 |
| 333 | I don't know if I've said this publicly, | 我不知道我是否公开说过 |
| 334 | but I think it's true, which is frankly, | 但我认为这是真的 老实说 |
| 335 | AI models are terrible for learning. | AI模型对于学习来说是可怕的. |
| 336 | I know people think AI is wonderful at getting things done, | 我知道人们觉得AI很擅长做事情 |
| 337 | use it all the time, love it. | 总是用它,爱它。 |
| 338 | But all the data that's coming out is that | 但所有数据都显示 |
| 339 | when say college students use AI, we know this. | 当说大学生使用AI, 我们知道这一点。 |
| 340 | It's just the studies now back up as well, | 只是现在的学习也有所恢复 |
| 341 | so we also have numbers, but the data is very clear. | 因此,我们也有数字, 但数据非常清楚。 |
| 342 | Students go higher on homeworks when they use AI, | 学生用人工智能做功课会更高 |
| 343 | yay, higher homework schools. | 是的,高等家庭作业学校。 |
| 344 | But retention, their long-term performance is much worse | 但保留,他们的长期表现更差 |
| 345 | because their AI do the work for them. | 因为他们的人工智能为他们工作。 |
| 346 | More and more studies are coming out to back this up now | 越来越多的研究 出来支持现在 |
| 347 | that I think people think, oh, it turns out, | 我认为人们会认为, 哦,结果, |
| 348 | I think Wikipedia is a wonderful tool, it has tons of facts. | 我认为维基百科是一个很棒的工具, 它有大量的事实。 |
| 349 | Web search is a wonderful tool, it has tons of facts. | 网络搜索是一个很棒的工具,它有很多事实。 |
| 350 | But it turns out that when you ask AI to do work for you, | 但事实证明 当你要求AI为你工作时 |
| 351 | you are cognitive offloading to AI, | 你将认知卸载到AI, |
| 352 | which is great because that's how society moves forward | 这是伟大的,因为 社会如何前进 |
| 353 | and it gets work done. | 并且它得到了工作完成。 |
| 354 | But human retention is much worse. | 但人类保留的情况更糟。 |
| 355 | It's just so clear that LLMs, as they are most commonly used, | 很明显,LLMS, 因为最常用的是, |
| 356 | are terrible for learning. | 学习是可怕的。 |
| 357 | I'm not saying there's no way to use it | 我不是说没有办法用它 |
| 358 | in a way that is good for learning. | 以有利于学习的方式。 |
| 359 | I think there are ways to use it that are good for learning. | 我认为有方法使用它,对学习有好处. |
| 360 | But even for myself, there's so many things | 但即使是我自己, 有很多事情 |
| 361 | that are also AI modeled over the last six months | 过去六个月的人工智能模型 |
| 362 | or whatever, like building some project, | 或者什么的,像建设一些项目, |
| 363 | how does this front-end backend component work, whatever. | 这个前端后端组件是如何工作的,什么的。 |
| 364 | Give me the answer, get the job done, it was fantastic. | 给我答案,把工作做好,真是太棒了 |
| 365 | But six months later, I don't remember the answer | 但半年后 我不记得答案了 |
| 366 | when I need to redo that front-end backend component, | 当我需要重做前端后端组件时, |
| 367 | I ask the AI again. | 我又问AI。 |
| 368 | So data's really clear. | 所以数据非常清晰。 |
| 369 | We should stop thinking of AI as helpful for learning. | 我们不应该把人工智能视为有助于学习。 |
| 370 | At least the vast majority of ways | 至少绝大多数的方法 |
| 371 | that the vast majority of people are using AI models today | 绝大多数人今天都在使用AI模型 |
| 372 | is absolutely terrible for learning. | 这对学习是绝对可怕的。 |
| 373 | But you're building a company, helping solve that, right? | 但你正在建立一个公司,帮助解决,对不对? |
| 374 | Because the one-to-one tutoring with AI | 因为跟AI的一对一辅导 |
| 375 | is that where you just announced | 是你刚刚宣布的 |
| 376 | 和Coursera的一亿投资 | |
| 377 | Yes, so I'm excited about leading a new organization | 是的,所以我很高兴领导一个新的组织 |
| 378 | called Learn Vector that is focused | 名为“ 学习矢量” 的焦点 |
| 379 | on building new learning experiences | 积累新的学习经验 |
| 380 | that is much more one-to-one than one-to-many. | 这比一对一多得多 |
| 381 | So 15 years ago, I was privileged to participate | 所以15年前,我有幸参加 |
| 382 | in the online courses movement | 在线课程运动 |
| 383 | that I think changed the way a lot of people learn. | 我认为改变了很多人的学习方式 |
| 384 | But that was, and still remains, | 但这曾经是,现在仍然是, |
| 385 | largely a one-to-many experience | 主要是一对一的经验 |
| 386 | where everyone kind of watches the same video, | 所有人都看同样的视频 |
| 387 | which is actually okay, it actually works well. | 这其实是好的, 它实际上工作很好。 |
| 388 | But the technology now exists | 但技术已经存在 |
| 389 | to create much more personalized, | 创造更个性化, |
| 390 | customized one-to-one experiences. | 定制的一对一体验. |
| 391 | And so our team is working hard on that. | 因此,我们的团队正努力进行这项工作。 |
| 392 | I think we'll have a lot more to show by early next year. | 我想到明年初我们还有很多事情要做 |
| 393 | When think about human skill development, | 想想人类技能的发展 |
| 394 | I feel like because AI has so heavily impacted | 我觉得因为人工智能影响很大 |
| 395 | software engineering, what we see happening | 软件工程,我们看到的 |
| 396 | in the job market for software engineering | 在软件工程的就业市场上 |
| 397 | is a harbinger, or it's a forerunner | 是预兆,还是预兆 |
| 398 | of what we'll see in other disciplines as well. | 我们在其他学科也会看到的东西 |
| 399 | And in software engineering, | 在软件工程方面 |
| 400 | people need to learn new skills, but when they do, | 人们需要学习新的技能,但当他们学习时, |
| 401 | they are thriving and creating more value | 他们正在繁荣,创造更多的价值 |
| 402 | and frankly getting raises | 坦率地说,我得到了高薪 |
| 403 | and doing even more exciting projects. | 和做更令人兴奋的项目。 |
| 404 | And what I'm seeing the early signs of | 我所看到的早期迹象 |
| 405 | in other disciplines as well, | 其他学科, |
| 406 | for example, in software engineering, | 例如,在软件工程方面, |
| 407 | most developers, like front-end, back-end developers, | 大多数开发者,如前端,后端开发者, |
| 408 | have now become full-stack developers. | 现已成为全存储开发者。 |
| 409 | Because of AI help, you could take on broader scope. | 由于AI的帮助,你可以承担更广泛的范围. |
| 410 | I'm seeing early signs of this in other disciplines as well, | 我在其他学科也发现了早期的迹象, |
| 411 | where, for example, someone in marketing | 例如,市场营销中的人 |
| 412 | that did marketing coordination, | 做营销协调, |
| 413 | coordinated marketing campaigns, with AI help, | 在大赦国际的帮助下,协调开展营销活动, |
| 414 | they can now become more of a full-cycle marketer | 他们现在更能成为全周期的营销者 |
| 415 | and take on a broader scope. | 并且扩大范围。 |
| 416 | And I'm seeing, frankly, sources in recruiting | 坦率地说,我看到 招募的来源 |
| 417 | become more full-cycle, do end-to-end recruiting. | 成为更完整的循环, 做端到端的招募。 |
| 418 | So now the good news and bad news is | 所以现在好消息和坏消息是 |
| 419 | for people to step up to these broader roles, | 让人民承担起这些更广泛的角色, |
| 420 | you do need to learn AI skills, | 你需要学习人工智能技能, |
| 421 | but also it's not just learning AI. | 不只是学习人工智能 |
| 422 | You also need to learn these other skills, | 你还需要学习这些其他技能, |
| 423 | like how do you do the other parts of marketing, | 就像你如何做市场营销的其他部分, |
| 424 | of recruiting, or software engineering, or AI engineering. | 招聘,或软件工程,或AI工程。 |
| 425 | So I think this actually creates a heavy need, | 所以我认为这实际上造成了一个沉重的需要, |
| 426 | a big need for people to gain new skills, | 需要人们获得新的技能, |
| 427 | but when they do, which is both AI skills, | 但当他们这样做, 这既是AI的技能, |
| 428 | but also disciplinary skills, | 同时也有纪律技能, |
| 429 | then they can do much more, hopefully have more fun, | 然后,他们可以做更多的, 希望有更多的乐趣, |
| 430 | work on more exciting projects, | 在更刺激的项目上工作, |
| 431 | hopefully get paid more as well. | 希望也能得到更多的报酬。 |
| 432 | And one reason I kind of worry about the fear monitoring | 还有一个原因 我担心恐惧的监控 |
| 433 | is I got an email from someone | 我收到别人的邮件 |
| 434 | that was about to enter college, | 即将进入大学, |
| 435 | and he emailed me saying, | 他给我发邮件说 |
| 436 | hey, Andrew, taking online courses, | 嘿,安德鲁,上在线课程 |
| 437 | but I'm really struggling with what I should major in college | 但我真的在为大学本科生而挣扎 |
| 438 | because in four years, won't AI do all this, | 因为四年之内 AI不会做这一切 |
| 439 | and everything I learn will be obsolete? | 我学到的一切都会过时? |
| 440 | And the answer is no, of course it won't all be obsolete, | 答案是没有,当然不会都是过时的 |
| 441 | but when we keep on pushing these fear messages, | 但是当我们不断推波助澜时 |
| 442 | we make people wonder if they will even be relevant, | 我们让人们怀疑他们是否相关 |
| 443 | and it makes people not lean in to gain these skills, | 它使得人们不 倚靠获得这些技能, |
| 444 | they'll put them in a much better position. | 他们会让他们处于更好的地位。 |
| 445 | So I see very clearly that these fear monitoring messages | 所以我非常清楚地看到 这些恐惧监测信息 |
| 446 | are distorting how many people, | 扭曲了多少人, |
| 447 | including high school students, college students, | 包括高中生、大学生、 |
| 448 | fresh grads, think about the economy, | 刚毕业,想想经济, |
| 449 | and frankly, making people give up | 坦白地说,让人们放弃 |
| 450 | is one of the worst things we've been doing | 是我们做过的最糟糕的事之一 |
| 451 | in this era when people that lean in will thrive. | 在这个时代, 人倾斜将繁荣。 |
| 452 | Andrew has taught over 8 million people AI. | 安德鲁教过超过800万人AI. |
| 453 | He started teaching machine learning online back in 2011, | 他从2011年开始在网上学习机器, |
| 454 | years before the current AI boom. | 距当前AI繁荣还有好几年 |
| 455 | Now one of the companies he's building | 现在他正在建设的公司之一 |
| 456 | is focused on AI agents. | 重点是人工智能特工 |
| 457 | From the way that it sounds, | 从听起来, |
| 458 | it can still feel way too technical. | 感觉还是太技术了 |
| 459 | So I put together a step-by-step guide | 所以我编了一部"一步一步"的指南 |
| 460 | to building your first AI agent with no coding required. | 来建立第一个不需要编码的AI代理. |
| 461 | It walks you through what to automate, how to set it up, | 它让你通过什么 自动化,如何设置它, |
| 462 | and how to make it actually useful. | 如何让它真正有用。 |
| 463 | It is in my newsletter this week. | 我这周的通讯里有这个 |
| 464 | The newsletter is called Future Proof. | 该通讯名为"未来证明". |
| 465 | It's free. | 自由无碍. |
| 466 | Link is in the description. | 链接在描述中. |
| 467 | What would you reply back to that email | 你对那封邮件有什么反应? |
| 468 | that somebody sent you? | 有人派你来的? |
| 469 | What would you say is the best major to study now, | 你会说什么是最好的主 现在学习, |
| 470 | to thrive in AI era? | 在AI时代蓬勃发展? |
| 471 | Do you think it's like going deep into a niche | 你觉得这就像深入到一个小地方 |
| 472 | or just broader computer science | 或者只是更广泛的计算机科学 |
| 473 | so that you can acquire AI skills really fast? | 这样你才能很快获得人工智能技能? |
| 474 | You know, I don't know what's the best major. | 你知道,我不知道什么是最好的主修课。 |
| 475 | There are awful lot of great majors. | 有很多伟大的专业。 |
| 476 | It's like, I kind of feel like | 我觉得... |
| 477 | what's the best job in the world? | 什么是世界上最好的工作吗? |
| 478 | What's the best major in the world? | 何等为最上大修. |
| 479 | Something that you love, right? | 你爱的东西,对不对? |
| 480 | Yeah, my daughter wants to be an astronaut. | 对,我女儿想当宇航员 |
| 481 | I don't know if she can major in becoming an astronaut. | 我不知道她能否成为宇航员 |
| 482 | I have to think about that. | 我得考虑一下 |
| 483 | When she gets older, she may change her mind. | 她长大后可能会改变主意 |
| 484 | I see so many opportunities across so many job roles. | 我看到很多机会 贯穿这么多的工作角色。 |
| 485 | It all seems very exciting to me. | 这一切似乎都让我兴奋。 |
| 486 | But do learn AI, do learn to build with AI. | 但要学AI,要学与AI一起建设. |
| 487 | The other thing that my team's been working | 还有一件事,我的团队一直在工作 |
| 488 | on an AI engineering skills map to try to map out | 在AI的工程技能地图上尝试映射 |
| 489 | the most important skills for AI engineering. | AI工程最重要的技能. |
| 490 | One thing that I felt intuitively, | 有件事我直觉地觉得 |
| 491 | but I was surprised to see it show up in the data, | 但我惊讶地看到它出现在数据中, |
| 492 | was that a lot more job descriptions seem to be saying | 更多的职务说明似乎在说 |
| 493 | that they want people that demonstrate | 他们想要展示的人 |
| 494 | a very high sense of agency. | 一种很高的代理意识。 |
| 495 | Because it turns out with AI, | 因为有AI的发现 |
| 496 | there are a lot more opportunities for individuals | 个人有更多机会 |
| 497 | to spot problems and go build something | 发现问题,去制造一些东西 |
| 498 | or do something to go solve it. | 或做一些事情来解决它。 |
| 499 | So I think we're really evolving. | 所以,我认为我们真的在进化。 |
| 500 | Well, we've long been evolving, | 我们早就在进化了 |
| 501 | but we're accelerating past the era | 但我们正在加速穿越时代 |
| 502 | where people sit around and wait for their boss | 在那里人们坐着等待他们的老板 |
| 503 | to tell them what to do. | 告诉他们该怎么做。 |
| 504 | This is what I've been feeling a lot, | 这就是我一直以来的感受, |
| 505 | especially when we started doing remote work. | 特别是我们开始做远程工作的时候 |
| 506 | I want people to be entrepreneurs within their niche. | 我希望人们在自己的专长内成为企业家. |
| 507 | Like if you're helping me with LinkedIn, | 就像你帮我处理LinkedIn一样 |
| 508 | you're an entrepreneur there. | 你是一个企业家在那里。 |
| 509 | You can hire more contractors. | 你可以聘请更多的承包商。 |
| 510 | You can deploy different tools. | 你可以部署不同的工具。 |
| 511 | You make the strategic decision | 你来决定战略 |
| 512 | whether this topic is good or not. | 不论这个议题是否良好。 |
| 513 | Shall we proceed with it? | 我们继续吧? |
| 514 | I really think, and tell me if you agree with me, | 我真的觉得 如果你同意的话 告诉我 |
| 515 | we're moving into that job market | 我们要进入这个职业市场 |
| 516 | where everyone is kind of independent in their workplace. | 在那里,每个人都在他们的工作场所是独立的。 |
| 517 | I think people will have much more autonomy and creativity. | 我认为人们会有更多的自主权和创造力. |
| 518 | So I agree with that. | 所以我同意这一点。 |
| 519 | And I'd even go one step further, | 我甚至更进一步, |
| 520 | which is I talked to a lot of people, | 我和很多人谈过 |
| 521 | engineers and others in large companies | 大公司的工程师和其他人 |
| 522 | that told me that the manager tells them | 这告诉我,经理告诉他们 |
| 523 | to stay in their swim lane. | 留在他们的游泳道。 |
| 524 | They'll say, oh, I have this creative idea, | 他们会说,哦,我有一个创意的想法, |
| 525 | but the manager says, no, | 但经理说,不, |
| 526 | and they should have focused on this one thing, frankly, | 他们应该把注意力集中在这件事上 坦白地说 |
| 527 | often because the manager's career depends on it. | 经常是因为经理的职业生涯取决于此. |
| 528 | But I feel like the number of opportunities | 但我觉得机会之多 |
| 529 | for people to spot things outside their swim lane | 在游泳道外发现东西 |
| 530 | and then in a responsible way explore how to get it done, | 然后以负责任的方式探索如何完成它, |
| 531 | that feels very exciting to me. | 感觉很刺激 |
| 532 | And I think that in the future, | 我认为在未来, |
| 533 | the businesses that set up a culture | 建立文化的企业 |
| 534 | that encourage people to learn AI, build fast, responsibly, | 鼓励人们学习AI, 建设快速,负责, |
| 535 | talk to customers would drive a lot more value | 和顾客聊天会吸引更多的价值 |
| 536 | than the more hierarchical silo organizations. | 而不是更分级的组织。 |
| 537 | Yeah, it starts with hiring the right people | 是啊,它从雇用合适的人开始 |
| 538 | and then nurturing this in your organization. | 然后在你的组织里培育这个 |
| 539 | When you say learn how to use AI | 当你说学习如何使用AI |
| 540 | and become proficient with AI, | 并且精通AI, |
| 541 | can you give me some benchmarks? | 你能给我一些基准吗? |
| 542 | Like of a person who's like say marketer, | 就像说市场营销员一样 |
| 543 | knowledge worker, advanced with AI, | 知识工作者,高级AI, |
| 544 | what are you looking for | 你们在找什么? |
| 545 | when you're interviewing this person? | 当你采访这个人时? |
| 546 | I'm pretty sure my team's ahead of the curve. | 我敢肯定,我的团队 领先曲线。 |
| 547 | All of my marketers know how to code. | 我所有的推销员都知道如何编码 |
| 548 | So as part of how I interview marketers, | 因此,作为我如何采访营销员的一部分, |
| 549 | we ask them what they've built | 我们问他们建造了什么 |
| 550 | and if they have not built any software. | 如果他们没有建造任何软件, |
| 551 | If it's a dashboard, is it good or bad? | 如果是仪表板,是好是坏? |
| 552 | Like is it too basic or? | 这太基本了还是? |
| 553 | A dashboard, again, my team's pretty, | 仪表板,再次, 我的团队很漂亮, |
| 554 | you know, somewhat ahead of the curve, | 你知道,有些在曲线前, |
| 555 | but all of my marketers have built | 但我所有的推销员都造了 |
| 556 | much more sophisticated things than dashboards. | 比仪表板更精密的东西 |
| 557 | Oh, I feel like, I don't know, | 哦,我觉得,我不知道, |
| 558 | the other day, one of our, | 那天 我们中的一个 |
| 559 | someone on the marketing team | 有人在营销团队 |
| 560 | was talking about the tools that he had built. | 他正在谈论他建造的工具。 |
| 561 | When he's considering writing an article on something, | 当他考虑写一篇关于某事的文章时, |
| 562 | it will crawl the web, find related work, | 它会爬网 找到相关的工作 |
| 563 | has a custom desktop app, | 有自定义桌面应用程序, |
| 564 | actually build a desktop app that runs on his Mac | 实际构建一个运行在其 Mac 上的桌面应用程序 |
| 565 | to highlight related articles for him. | 为他突出相关文章. |
| 566 | Then he can chat to the whole system, | 然后他可以和整个系统聊天 |
| 567 | navigate, you know, the thing he's writing, | 导航,你知道, 东西他写, |
| 568 | as well as the related work | 及有关的工作 |
| 569 | and he had a large dashboard for crawling the internet | 他有一个巨大的仪表板 爬上网 |
| 570 | to highlight to him exciting things that are popping up. | 告诉他一些激动人心的事 |
| 571 | Now, even on my team, | 现在,即使在我的团队, |
| 572 | I think that marketplace is ahead of the curve. | 我认为市场领先于曲线。 |
| 573 | But that's great to hear. | 但这是伟大的听到。 |
| 574 | Any other interesting use cases | 任何其他有趣的使用案例 |
| 575 | that will inspire people to build something similar? | 那会激励人们 建立类似的东西? |
| 576 | Let's see, maybe my finance team uses AI extensively. | 让我们看看,也许我的金融团队 广泛使用AI。 |
| 577 | So I think one of my CFOs realized that, | 所以我认为我的首席财务干事意识到, |
| 578 | you know, her team was spending hours every week | 你知道,她的团队 每周花几个小时 |
| 579 | clicking through documents, open this copy, | 点击文档,打开本副本, |
| 580 | paste this number here. | 在这里粘贴这个号码。 |
| 581 | And so she started building automation scripts | 于是她开始制作自动化脚本 |
| 582 | that runs on a routine that automatically opens files, | 运行在自动打开文件的常规上, |
| 583 | checks what's in there, | 检查里面有什么 |
| 584 | 校对:Soup | |
| 585 | if there's something they need to be paying attention to | 如果有什么需要他们注意的 |
| 586 | of a new document that showed up. | 显示的新文件。 |
| 587 | So I find that rather than waiting around | 所以,我发现这 宁可等待周围 |
| 588 | for an engineer to do the work for them, | 一个工程师为他们工作, |
| 589 | the team's ability to kind of not just build dashboards, | 团队的能力 那种不只是建造仪表板, |
| 590 | but build kind of a data management infrastructure | 但建立某种数据管理基础设施 |
| 591 | that's taking ingest data, | 这是摄入数据, |
| 592 | alert them if something's happening. | 发生什么时通知他们 |
| 593 | I think my finance and marketing teams are doing that. | 我认为我的金融和营销团队正在这样做。 |
| 594 | Oh, my recruiting team, | 哦,我的招募团队, |
| 595 | well, we actually have recruiting engineers, | 其实我们招聘了工程师 |
| 596 | which are really professional engineers | 他们是真正的专业工程师 |
| 597 | that sit in a recruiting team. | 在一个招募团队。 |
| 598 | They're building very sophisticated tools for recruiting. | 他们正在建造非常复杂的招募工具. |
| 599 | And this is actually the other trend. | 这是另一个趋势。 |
| 600 | I think marketers, recruiters, HR professionals, | 我认为市场营销者,招聘者,人力资源专业人士, |
| 601 | ops people should all learn AI. | 行动人员应该学会人工智能 |
| 602 | But the other thing is when you take an engineer | 但另外一件事是当你接受工程师 |
| 603 | and embed them in these teams, | 把他们安置在这些队伍中, |
| 604 | then that further accelerates what you can do. | 然后,这进一步加快了你能做的。 |
| 605 | We do the same. | 我们也是 |
| 606 | We start with something basic, build it ourselves. | 我们从基本的东西开始, 建造它自己。 |
| 607 | Then we hit the wall and engineer comes in. | 然后我们撞上了墙 工程师进来了 |
| 608 | We build it further. | 我们进一步建造它。 |
| 609 | Frankly, when you look at not just software engineers, | 老实说,当你看到 不只是软件工程师, |
| 610 | but recruiting engineers, marketing engineers, | 但招聘工程师,营销工程师, |
| 611 | HR engineers, I think there's so much valuable | HR工程师,我觉得有很多有价值的东西 |
| 612 | engineering work that can now be done. | 现在可以做的工程工作。 |
| 613 | I'm just not worried about running out of, | 我只是不担心 跑出来, |
| 614 | frankly, all my friends were so busy. | 坦白说,我的朋友都很忙 |
| 615 | We think, boy, how could we run out of engineering jobs? | 我们想,孩子,我们怎么能用完工程工作? |
| 616 | Yeah, there's so many cool ideas you can experiment on, | 是啊,有很多很酷的想法 你可以实验, |
| 617 | but you touched upon something | 但你碰了个东西 |
| 618 | that is actually one of the fears. | 这其实就是其中之一的恐惧。 |
| 619 | When we talk about like financial information, | 当我们谈论像财务信息, |
| 620 | how much you're giving to AI. | 你要给AI多少钱? |
| 621 | So I gave my perplexity permission | 所以我才允许我感到困惑 |
| 622 | to scan my fidelity account so it can track my portfolio, | 扫描我的忠诚账户 这样它就可以追踪我的投资组合 |
| 623 | tell me when to rebalance. | 告诉我什么时候再平衡 |
| 624 | It doesn't do anything on my behalf, but it has access. | 它不代表我做任何事,但它有权限。 |
| 625 | Do you think there is any problem with that? | 你觉得有什么问题吗? |
| 626 | This is complicated. | 这很复杂 |
| 627 | I think AI and privacy is a complex area. | 我认为AI和隐私是一个复杂的领域. |
| 628 | And it depends a lot on the company | 这要靠公司 |
| 629 | that you are sharing your data with. | 您正在分享您的数据。 |
| 630 | So for example, I trust all the hyperscalers | 比如说,我信任所有的超标器 |
| 631 | to really 100% follow their terms of service | 百分之百遵守他们的服务条款 |
| 632 | and to do what they say. | 并按他们说的做。 |
| 633 | My personal opinion, not giving legal business advice, | 我个人的意见,不提供 法律业务建议, |
| 634 | but I'd be shocked if the largest hyperscalers | 但是,如果最大的超标,我会震惊 |
| 635 | publish the terms of service with some privacy notice | 发布一些隐私通知的服务条款 |
| 636 | and if they reach that, | 如果他们到了那里, |
| 637 | because that'd be not the culture, | 因为那不是文化 |
| 638 | it'd be so damaging to the long-term business model. | 这会对长期商业模式造成如此大的伤害。 |
| 639 | Now that's on the largest hyperscaler side. | 现在,这是最大的超规模。 |
| 640 | If you look at the AI company side, | 如果你看看AI公司那边 |
| 641 | there's been at least one company that I won't name | 至少有一家公司我不点名 |
| 642 | that seems to occasionally change the terms of service | 这似乎偶尔改变服务条件 |
| 643 | and if you're using it, you go to the website, | 如果你在使用它, 你到网站, |
| 644 | so you pop up, hey, we changed the terms of service | 所以,你出现了,嘿, 我们改变了服务条件 |
| 645 | to retain your data or train your data. | 保留数据或训练数据。 |
| 646 | And if you aren't paying attention | 如果你不注意的话 |
| 647 | and click the wrong button, | 点击错误的按钮, |
| 648 | then they suddenly gave themselves permission | 然后,他们忽然答应了。 |
| 649 | to access your data | 访问您的数据 |
| 650 | in a way that I'm not that comfortable with. | 我不太喜欢这样 |
| 651 | I feel like I handle some sensitive information. | 我觉得我处理一些敏感的信息。 |
| 652 | So then I tend to be very careful with the businesses | 所以,我总是非常小心 与企业 |
| 653 | that I just don't feel their culture and their DNA. | 我只是感觉不到他们的文化和DNA |
| 654 | And frankly, the long-term business model | 坦白说,长期商业模式 |
| 655 | is as tied to protecting individual user privacy | 与保护个人用户隐私一样 |
| 656 | than the hyperscalers. | 比超标者还要大 |
| 657 | And I see businesses get this as well. | 我认为企业也得到了这个。 |
| 658 | 例如,我的一个团队,AI Aspire, | |
| 659 | we work with very large corporations, | 我们和大公司合作 |
| 660 | including banks with incredibly sensitive financial data. | 包括拥有极其敏感金融数据的银行。 |
| 661 | And as you can imagine, AI Aspire and our clients | 你可以想象,AI Aspire和我们的客户 |
| 662 | do not willingly share really sensitive, | 不愿分享真正的敏感, |
| 663 | often material non-public information, right, NMPI, | 经常提供非公开资料,对,NMPI, |
| 664 | with frontier labs without really careful thinking | 与前沿实验室 没有真正仔细的思考 |
| 665 | about the guardrails in privacy. | 私底下的护栏 |
| 666 | So I think it's complicated. | 所以我觉得这很复杂。 |
| 667 | So trusting hyperscalers, | 如此信任超标者, |
| 668 | but also another thing that you can do, | 但还有一件事你可以做, |
| 669 | you can download an open-source model | 您可以下载一个开源模型 |
| 670 | and just run it on your computer | 在你的电脑上运行 |
| 671 | and then it just stays on your computer, right? | 然后它只是留在你的电脑上,对不对? |
| 672 | Yes, I think, yes, it turns out a lot of banks | 是的,我想,是的,结果很多银行 |
| 673 | will actually run the things | 将实际运行的东西 |
| 674 | in the virtual private cloud or on-prem, | 在虚拟的私密云中,或在梦中, |
| 675 | so they never even leave their control. | 所以他们从不离开他们的控制。 |
| 676 | But I think for individuals, | 但我认为对于个人来说, |
| 677 | for the really sensitive things, | 对于真正敏感的东西, |
| 678 | I sometimes run a local model. | 我有时会经营本地模特儿 |
| 679 | And it's been interesting with the open-way models. | 并且对开通车型颇感兴趣. |
| 680 | So the latest open-way models | 因此,最新的开放模式 |
| 681 | are approaching frontier capability | 正在接近边境能力 |
| 682 | and that are actually small enough. | 这其实足够小。 |
| 683 | There are actually really good models now | 现在真的有很好的模型了 |
| 684 | they can run on. | 他们可以继续前进。 |
| 685 | Yeah, the one from Meta, right? | 是啊,梅塔的那个,对吧? |
| 686 | The recent one. | 最近那个 |
| 687 | Oh yes, Metamuse is a good model. | 哦,是的,Metamuse是一个很好的模型。 |
| 688 | And I'm thinking also the latest version of Quen | 我还在想最新版的Quen |
| 689 | is also very good. | 也很不错。 |
| 690 | But I think frankly, these models change every other week. | 但坦率地说,这些模型每隔一周就改变一次。 |
| 691 | So I think the best practice is to not get stuck on one, | 所以我认为最好的做法 就是不要被困在其中 |
| 692 | but keep on trying new models. | 但继续尝试新的模型。 |
| 693 | So basically when there is a situation | 所以,基本上当有情况 |
| 694 | that you don't trust anyone, | 你不信任任何人, |
| 695 | you run a local model | 您运行本地模式 |
| 696 | and this is how you keep your data safe. | 这就是你保护数据安全的方法 |
| 697 | I do trust the hyperscalers, | 我确实相信那些超规模的人 |
| 698 | but sometimes for literally NMPI, | 但有时对NMPI来说 |
| 699 | material non-public information, | * 非公开资料, |
| 700 | that I won't even send to that, | 我甚至不会送去, |
| 701 | I just can't even send that to the cloud. | 我甚至不能把那个寄到云里 |
| 702 | So that, I'll either do it manually without AI help | 所以,我要么手动做,没有AI帮助 |
| 703 | or if I really need to use AI, | 或者,如果我真的需要使用AI, |
| 704 | then really carefully only use a local model. | 然后非常小心地使用一个本地模型。 |
| 705 | Interesting. | 有意思,这很有趣 |
| 706 | Okay, this is an interesting one. | 好吧,这是一个有趣的。 |
| 707 | Okay, what about loss of human control over AI? | 好吧,那人类失去对AI的控制呢? |
| 708 | 因为我和Yoshua Benja谈过了 | |
| 709 | who is very negative when it comes to free AI | 在自由AI方面,谁是非常消极的 |
| 710 | without any regulation. | 没有规定。 |
| 711 | And he painted me some very scary pictures | 他给我画了一些非常恐怖的照片 |
| 712 | of AI taking over control, | 由大赦国际接管, |
| 713 | because basically the whole scenario is | 因为基本上整个情景 |
| 714 | we can't control something that's smarter than us. | 我们不能控制比我们聪明的东西 |
| 715 | And if AI gets smarter and smarter, | 如果AI变聪明了 |
| 716 | where do we end up? | 我们到哪里去? |
| 717 | What do you think about that? | 你觉得怎么样? |
| 718 | I think about something else that we can't control, | 我想到别的东西,我们不能控制, |
| 719 | which is airplanes. | 这是飞机。 |
| 720 | No one can build an airplane | 没人能造飞机 |
| 721 | that you can fly perfectly. | 你可以完美飞行。 |
| 722 | Winds will buff it around. | 风会吹动它周围。 |
| 723 | And then candidly, | 然后坦率地说, |
| 724 | in the early days of developing airplanes, | 在发展飞机的初期, |
| 725 | some airplanes crashed and people died | 有些飞机坠毁,有人丧生 |
| 726 | and it was tragic and awful. | 悲剧和可怕。 |
| 727 | But through the early lessons learned, | 但通过早期的教训, |
| 728 | we then learned to control airplanes better and better | 然后我们学会了更好控制飞机 |
| 729 | so that today we can mostly get in an airplane | 这样今天我们就可以乘飞机了 |
| 730 | and not fear too much for our lives. | 也不要太害怕我们的生活。 |
| 731 | And it's really like that too with AI. | 对AI来说也是这样 |
| 732 | No one can perfectly control AI | 没有人能完全控制AI |
| 733 | because it generates tokens or outputs | 因为它会产生符号或输出 |
| 734 | that are a little bit random, | 这有点随机, |
| 735 | so we don't really know what exactly it'll do. | 所以我们不知道它到底能做什么 |
| 736 | But as we run them, | 但当我们运行他们, |
| 737 | and there's been a small number of mishaps, | 发生了少量的意外 |
| 738 | which is unfortunate, | 这是不幸的, |
| 739 | and some number of mishaps have done some real damage, | 和一些灾祸 已经做了一些真正的损害, |
| 740 | but the way we engineer almost any system, | 但是我们设计系统的方式, |
| 741 | from an airplane to electric circuits to now AI, | 从飞机到电路到现在的AI |
| 742 | is carefully grow their capabilities | 仔细地提高他们的能力 |
| 743 | so that we can have a controlled environment | 以便我们有一个控制的环境 |
| 744 | in which to measure what's wrong | 用来衡量什么是错的 |
| 745 | and then to shape it, | 然后造出来, |
| 746 | to make sure we can control it well enough | 确保我们能很好地控制它 |
| 747 | that it behaves responsibly and safely. | 它的行为是负责和安全的。 |
| 748 | And to this day, | 时至今日, |
| 749 | we can't perfectly control any airplane | 我们不能完全控制任何飞机 |
| 750 | and we will never perfectly control AI either. | 我们也不会完全控制AI |
| 751 | But I think we are certainly controlling them well enough | 但我认为我们控制得够好了 |
| 752 | that this loss of control doesn't feel like science, | 这种失控的感觉不像是科学 |
| 753 | if it's a science fiction. | 如果这是个科幻小说 |
| 754 | Yeah, yeah. | 对 对 |
| 755 | What about deep fakes? | 那深层假货呢? |
| 756 | Deep fakes are a problem. | 深假是烦恼. |
| 757 | Well, one of the most disgusting things | 最恶心的事情之一 |
| 758 | I've ever seen or heard of | 我见过或听说过 |
| 759 | is non-consensual, intimate deep fake imagery. | 是未经同意的 亲密的深层假象 |
| 760 | I'm really glad that U.S. Congress has been moving | 我很高兴美国国会在搬家 |
| 761 | to let's pass laws, get rid of that, penalties for that. | 让我们通过法律, 摆脱它,惩罚它。 |
| 762 | I think there's some really problematic uses of AI | 我觉得人工智能的用法有点问题 |
| 763 | that we should outlaw, heavily penalize, | 我们应当严惩他们, |
| 764 | let's just get rid of that. | 我们把那个扔掉吧 |
| 765 | What do you think about children and social connection | 你对孩子和社会联系有什么看法? |
| 766 | when it comes to AI with kids | 当涉及到AI与孩子 |
| 767 | using more of AI? | 用更多的人工智能吗? |
| 768 | Because we've seen social media, | 因为我们看过社交媒体 |
| 769 | how there are people who are dumb-scrolling all day. | 怎么会有人整天都傻傻地滚 |
| 770 | And my daughter, who is five years old now, | 还有我女儿 她现在五岁了 |
| 771 | whenever I don't have an answer, | 当我没有答案时 |
| 772 | she's like, ask Chad GPT. | 她喜欢,问查德GPT。 |
| 773 | Who's that person? | 那个人是谁? |
| 774 | I'm like, I don't know. | 我很喜欢,我不知道。 |
| 775 | 问查德GPT. | |
| 776 | She thinks Chad GPT knows everything. | 她觉得Chad GPT什么都知道 |
| 777 | What would you say about kids' future with AI? | 你会怎么说 孩子们的未来与AI? |
| 778 | First, I think kids have a bright future. | 首先,我觉得孩子有一个光明的未来。 |
| 779 | It's such an exciting time | 这么刺激的时候 |
| 780 | to be a child to grow up in this environment | 在这样的环境下长大的孩子 |
| 781 | with tools that none of us ever had before. | 拥有我们从未有过的工具 |
| 782 | At the same time, we've seen that social media, | 同时,我们看到社交媒体, |
| 783 | I think social media has probably been blamed | 我觉得社交媒体可能已经被指责了 |
| 784 | a bit more than it deserves, | 比它应得的要多一点 |
| 785 | but it does deserve blame, | 但它确实值得责备, |
| 786 | has kind of not been great for kids. | 有点对孩子不好 |
| 787 | I actually worry a lot about, it's a wonderful tool, | 其实我担心的很多, 这是一个很好的工具, |
| 788 | but AI damaging learning is something I worry a lot about. | 但是,AI破坏的学习 是我担心的很多。 |
| 789 | So it turns out, I have a five-year-old and a seven-year-old, | 原来我有一个5岁和7岁的孩子 |
| 790 | when I teach them math, they're so young enough | 当我教他们数学时,他们那么年轻 |
| 791 | that I can basically not let them use the calculator, | 我不能让他们使用计算器, |
| 792 | can say, how do you multiply these numbers? | 可以说,你如何乘以这些数字? |
| 793 | And I don't give them a calculator | 我不给他们一个计算器 |
| 794 | and practice that with them. | 和他们一起练习。 |
| 795 | But as they're a little bit older, | 但是,由于他们有点老, |
| 796 | I worry a lot about students using cognitive offloading | 我非常担心学生使用认知卸载 |
| 797 | to AI in a way that damages the long-term learning retention. | 以损害长期学习保留的方式向大赦国际提出。 |
| 798 | But then at the same time, I actually built an app. | 但与此同时,我实际上还制作了一个应用程序。 |
| 799 | I did not like any of the free online learning | 我不喜欢任何免费的网上学习 |
| 800 | to type types of things, so I actually built my own | 来打几类东西, 所以我实际上建造我自己 |
| 801 | to have my daughter learn to type, | 让我女儿学会打字 |
| 802 | and I'm hoping that she's actually getting pretty decent now | 我希望她现在变得相当体面 |
| 803 | for a seven-year-old. | 给一个七岁的孩子。 |
| 804 | Oh, so she's typing already? | 哦,所以她已经打字了? |
| 805 | Oh yeah, she actually typed all the lowercase letters. | 哦,是的,她实际上打 所有小写字母。 |
| 806 | She's a little bit, her shift uppercase letters | 她有点小, 她的换班大写字母 |
| 807 | are a little bit not quite there, | 有点不完全在那里, |
| 808 | but I think that this unlocks responsible adult-supervised use | 但我认为这解锁 负责任的成年人监督使用 |
| 809 | of online tools, and I think it's really tricky. | 我认为这是非常棘手的。 |
| 810 | I think adult-supervised use of digital tools | 我认为是成年人监督使用的数字工具 |
| 811 | seems a great thing for kids, | 似乎对孩子们来说是件好事 |
| 812 | but too many adults don't have time | 但太多的成年人没有时间 |
| 813 | to supervise the use of the tools, | 监督工具的使用, |
| 814 | and then the incentives of, say, social media, right, | 然后是社交媒体的激励 |
| 815 | to do funny things. | 来做一些有趣的事情。 |
| 816 | Yeah, has to be the right incentive when it comes to AI. | 对,对AI来说,应该有正确的激励 |
| 817 | Okay, you mentioned, we talked about the fears, | 好吧,你提到, 我们谈论了恐惧, |
| 818 | we talked about how you can improve your work with AI. | 我们谈过如何改善你和AI的工作 |
| 819 | Can you name some of the biggest opportunities in AI in 2026 | 你能说说2026年AI里最大的机会吗? |
| 820 | for people who want to build? | 给那些想建楼的人? |
| 821 | For an individual that wants to build, | 对于一个想要建设的人来说, |
| 822 | I don't think it's one size fits all, | 我不认为它是一个尺寸适合所有, |
| 823 | but because the cost of building has plummeted, | 但由于建筑成本暴跌, |
| 824 | I encourage people to learn AI, build fast, | 我鼓励人们学习人工智能 快速建设 |
| 825 | and talk to customers. | 和客户交谈。 |
| 826 | I find myself building things, I don't know, | 我发现自己建造的东西, 我不知道, |
| 827 | every week, every weekend, | 每个礼拜,每个周末, |
| 828 | because I or someone on our team will have some problem | 因为我或我们团队中有人会有问题 |
| 829 | and I have some idea for building some AI thing | 我有点想法 建造一些AI的东西 |
| 830 | to automate it. | 把它自动化。 |
| 831 | Last weekend, I had, really, I was using a Frontier model | 上个周末,我真的在用边疆模型 |
| 832 | to analyze a lot of our key business metrics | 来分析我们许多关键的商业标准 |
| 833 | because I didn't have time to do it myself, | 因为我没有时间自己做, |
| 834 | but it was kind of measuring, you know, | 但它是一种测量, 你知道, |
| 835 | deep standardized key business metrics, | 深入标准化的关键业务衡量标准, |
| 836 | and I didn't have time to go find a data scientist | 我没有时间去找数据科学家 |
| 837 | to go work me on it, | 去工作我做到这一点, |
| 838 | so I just did a variety of Frontier models, | 所以我刚刚做了各种边疆模型, |
| 839 | being really careful on their data retention policies. | 对数据保存政策非常小心 |
| 840 | I did not use models with data retention policies | 我没有使用有数据保存政策的模型 |
| 841 | that I don't like in order to analyze data. | 我不喜欢分析数据 |
| 842 | But, and then I find that what's happened with AI | 但是,然后我发现 发生了什么事AI |
| 843 | is the cost of building has plummeted, | 建筑成本暴跌, |
| 844 | and so the challenge is shifting to deciding what to build, | 因此,挑战正在转向决定建设什么, |
| 845 | which I've been calling | 我一直在打电话 |
| 846 | the product management bottleneck. | 产品管理瓶颈。 |
| 847 | And so people, founders, engineers, product managers, | 于是人们,创始人,工程师,产品经理, |
| 848 | they can talk to customers, | 他们可以和顾客交谈, |
| 849 | get a sense for the taste of judgment on what to build, | 体会到判断如何构建的滋味 |
| 850 | and then build with AI and iterate quickly. | 然后用人工智能快速建造 |
| 851 | I think there's just a ton of exciting things to do. | 我觉得只有一吨刺激的事情要做。 |
| 852 | And you've been starting so many companies. | 你开了这么多公司 |
| 853 | You're like, when I looked at your portfolio, | 你就像,当我看着你的组合, |
| 854 | do you think, for beginners, | 你认为, 对于初学者, |
| 855 | when you said you built something during the weekend, | 你说你在周末造东西的时候 |
| 856 | how do you decide what to focus on? | 你如何决定什么是重点? |
| 857 | Or you can pursue multiple ideas because of AI now, | 或者你可以追求多种想法 因为AI现在, |
| 858 | and you can just be playing in different companies | 你可以在不同的公司里玩 |
| 859 | at the same time. | 同时。 |
| 860 | It turns out building a company is still really, really hard. | 事实证明,建立一家公司仍然非常困难。 |
| 861 | And so there's a lot to be said | 所以有很多话要说 |
| 862 | for a single-threaded leadership | 单线领导 |
| 863 | or someone that's fully focused on just one thing. | 或者某个完全专注于一件事的人 |
| 864 | I find that over a weekend, | 我周末发现, |
| 865 | I can often build an own wrapper, | 我常常可以自己做包装 |
| 866 | build a simple application, | 建立一个简单的应用程序, |
| 867 | but I wish it was that easy to build a large company. | 但我希望有那么容易 建立一个大公司。 |
| 868 | I find that building something meaningful | 我发现建造一些有意义的东西 |
| 869 | often takes either real technical depth | 往往需要真正的技术深度 |
| 870 | and or deep customer insight and integration of customers. | 或客户的深刻见解和融入。 |
| 871 | And yes, we can now use AI to code something | 是的,我们现在可以用AI 来编码一些东西 |
| 872 | in a few hours, | 几小时后 |
| 873 | but that's a small piece of the puzzle. | 但这是谜题的一小块 |
| 874 | So spending time understanding the technical complexity | 所以花时间了解技术的复杂性 |
| 875 | and building the really complex software, | 和建造非常复杂的软件, |
| 876 | that takes us like months, maybe years, | 我们花了几个月,也许几年, |
| 877 | or having that deep customer insight | 或者是客户的深刻见解 |
| 878 | to decide what to build. | 决定建设什么。 |
| 879 | That also just takes a lot. | 这也需要很多钱。 |
| 880 | Talking to people, reading facial expressions, surveys, | 与人们交谈,阅读面部表情,调查, |
| 881 | it's doing that over and over | 它的做 一遍又一遍 |
| 882 | until we figure out what to build. | 直到我们想出 建造什么。 |
| 883 | And so I think sometimes there's a lot of value | 所以我觉得有时有很多价值 |
| 884 | to sampling widely, | 广泛采样, |
| 885 | but then having that focus for an individual | 但后来又以个人为重点 |
| 886 | to go really deep in a couple of sectors, | 深入到几个部门, |
| 887 | that still seems important for building a business. | 这对建立企业来说仍然很重要。 |
| 888 | My last question, I know we don't have much time, | 我最后一个问题,我知道我们时间不多了 |
| 889 | but I wanted to ask you about AGI, | 但我想问你关于AGI的事 |
| 890 | just because people use this word so much. | 只是因为人们使用这个词这么多。 |
| 891 | And some people say, I think Jensen Huang said, | 有人说 黄詹森说 |
| 892 | we already reached AGI. | 我们已经到达AGI。 |
| 893 | You said it's decades away. | 你说过还有几十年 |
| 894 | What's the one criteria when you're gonna say | 当你要说,一个标准是什么? |
| 895 | we reach AGI? | 我们找到AGI了吗? |
| 896 | So different people say we reach AGI at different times | 不同的人说我们在不同的时间到达AGI |
| 897 | because of different definitions of AGI. | 因为AGI的定义不同. |
| 898 | The definition I'm most familiar with | 我最熟悉的定义 |
| 899 | is AI that could do any intellectual tasks | 是AI,可以做任何智力任务 |
| 900 | that a human can. | 人类可以。 |
| 901 | But so the human brain can take say five years | 但是人类的大脑需要五年的时间 |
| 902 | to study and do a PhD thesis. | 学习和做博士论文。 |
| 903 | And so can AI write a PhD thesis? | AI也能写博士论文吗? |
| 904 | Or a human can learn to drive a truck | 或者人类可以学会开车 |
| 905 | through a dense rainforest | 穿过茂密的雨林 |
| 906 | with tens of minutes of practice. | 和数十分钟的练习。 |
| 907 | So when can AI do that to drive a new environment | 因此,什么时候AI可以这样做 驱动新的环境 |
| 908 | with tens of minutes of practice? | 练习数十分钟? |
| 909 | It feels like there's a long list of these things | 感觉有一大堆这些东西 |
| 910 | that AI cannot do for what feels to me decades. | 对我而言,大赦国际不能做几十年的感受。 |
| 911 | I hope it's only decades. | 惟愿数十载. |
| 912 | Maybe it will turn out to be longer. | 也许它会变成更长。 |
| 913 | So that's what I think for that definition of AI or AGI, | 所以,这就是我的想法 对于AI或AGI的定义, |
| 914 | AGI还很远 | |
| 915 | But it turns out because of economic incentives, | 但结果却因为经济刺激 |
| 916 | I think open and Microsoft had an agreement. | 我认为开放和微软有协议。 |
| 917 | There's actually been renegotiated now, | 已经重新谈判过了 |
| 918 | so that's gone away. | 所以这已经消失了。 |
| 919 | 但OpenAI有经济刺激 | |
| 920 | to try to declare reaching AGI earlier. | 试图更早地宣布到达AGI。 |
| 921 | And so it turns out that if you come up | 结果发现如果你上来 |
| 922 | with other definitions of AI, | 与大赦国际的其他定义, |
| 923 | depending on how far you lower the bar, | 看你把酒吧放低多远 |
| 924 | then you could totally have reached AGI already | 那你就完全可以联系到AGI了 |
| 925 | or even 30 years ago, | 甚至在30年前, |
| 926 | depending on how you want to define it. | 取决于您要如何定义它。 |
| 927 | Yeah, true. | 是的,没错。 |
| 928 | Andrew, thank you so much for this positive conversation. | 安德鲁 非常感谢你这么积极的谈话 |
| 929 | Very applicable. | 非常适用。 |
| 930 | I like when you watch something | 我喜欢你看东西 |
| 931 | and then you go and you measure yourself | 然后你去测量自己 |
| 932 | against what people are doing with AI, | 与人们对AI的所做所为相对应, |
| 933 | look at your process and maybe expand it. | 看看你的过程,也许可以扩展它。 |
| 934 | So thank you so much for showing what your team is doing | 所以,非常感谢你展示 你的团队在做什么 |
| 935 | and thank you for your insights. | 谢谢你的洞察力 |
| 936 | Yeah, I think given the huge benefits of AI to come, | 是啊,我认为考虑到AI的巨大好处 来, |
| 937 | I hope whoever's watching this is motivated | 希望看这个的人有动机 |
| 938 | to really go learn AI, apply it, | 真正去学习AI,应用它, |
| 939 | and even to go build some things. | 甚至去制造一些东西。 |
该视频共有字幕 939 条。解锁更多字幕为会员功能，请移动到 价格