← 返回任务列表

How to Deslopify Your AI

382 段 · 2 位说话人 · 原片 42:30
M1
M10:00

AI垃圾内容真是到处都是,老兄。

The AI slop is everywhere, man.

听着,当'slop'这个词刚开始被使用时,把它当成那些就是不喜欢AI的人的愤怒反应,也许还算合理。

Look, when the word slop started being used, it was perhaps reasonable to write it off as an angry response from folks who just didn't like AI in general.

然而,现在毫无疑问,随着AI写作、图像生成、视频生成和代码生成变得无处不在,我们正生活在一个内容沼泽里。

However, there is no doubt now that as AI writing and image generation and video generation and code generation have gotten ubiquitous, we are living in a morass of content.

工作中,工作外,就是随时随地都有内容。

Inside work, outside of work, it is just content everywhere all the time.

而这实际上是有代价的。

And it actually comes with a cost.

如果说AI应用于工作的第一阶段是学习如何使用这些工具来产出大量大量的新内容,那么下一阶段就是如何通过善用它们来让自己脱颖而出。

If a first phase of AI for work was learning how to use these tools to produce just volumes and volumes of new content, a next phase is how to distinguish yourself by using them well.

所以朋友们,我们来聊聊如何让你的AI去垃圾化。

And so friends, let's talk about deslopifying your AI.

The AI Daily Brief是一个每日播客和视频,涵盖AI领域最重要的新闻和讨论。

The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.

好了朋友们,在深入之前先快速说几个公告。

All right friends, quick announcements before we dive in.
M1
M11:13

现在,还有两件小事。

Now, 2 more quick things.
M1
M11:35

第二,我今天在外出差,所以我们做一个操作员剪辑,这也是一个对几乎所有人都非常及时的话题,不过明天我会恢复正常集数。

Second, I am on the road today, so we are doing an operator's cut, also one that is very timely for just about everyone, but I'll be back with a normal episode for tomorrow.

好的,Noufar,欢迎回来,又给我们带来一个操作员剪辑。

All right, Noufar, welcome back with another operator's cut.

你好吗?

How are you?
F1
F11:49

我很好。

I'm good.

我对这一期很兴奋。

I'm excited about this one.

我对这个话题非常有热情。

I'm very passionate about this topic.
M1
M11:53

是啊,即使在Operator's Cuts的框架内——这些是《AI Daily Brief》中更偏实用的集数——也仍然存在概念性和应用性之间的区分。

Yeah, this is even, even within the framework of Operator's Cuts, which are sort of the more practical-oriented AI Daily Brief episodes, there's still a band of how conceptual versus applied they are.

而这一集几乎是纯应用性的了。

And this one is about as applied as it gets.

它基本上是一份清单,列出了该做和不该做的事情,而且有一定的结构。

It's pretty much a list of things to do or to not do that have some structure around them.

所以我很期待深入探讨。

So I'm excited to dig in.

希望听众们能学到一些非常具体的技巧、想法和思路等等。其中很多你们可能已经在凭直觉做了,因为我觉得这个听众群体比其他人更关注这些东西。但同时也希望这些内容能为你个人和周围的人提供帮助。

Hopefully, I think folks will come away with some very specific techniques, ideas, thoughts, et cetera, a lot of which they will probably have already been doing intuitively, because my guess is that this is an audience that thinks more about this stuff than many others, but can also hopefully provide resources for both you as the listener and other folks around you.

所以,不管怎样,我很期待开始。

So anyways, I'm excited to dive in.
F1
F12:37

好的。

All right.

在谈论具体细节之前,我们先达成共识。

So before we talk about specifics, let's bring us all to the same page.

而且我敢肯定你们这周甚至今天都已经见过类似的东西了。

And I'm sure that you've all seen it this week and maybe even today.

可能是一封全员邮件,读起来不像任何人写的,或者它太长、太流畅、有点空洞,节奏你们现在已经能在脑子里默念出来了:短句,然后列举三点,再然后一行告诉你某件事真正的意义,你在收件箱里多次看到过这种。

It might've been the all-hands email that sounded like nobody, or it was too long and too smooth, slightly sterile with a rhythm that you can hear in your own head by now, like short sentence, then list of 3, then a line telling you what something is really about, and you've seen that in your inbox in multiple places.

可能是一张特别逼真的图片,或者是一个广播广告,你一听就知道不是人写的。

Maybe it was the especially believable image, or maybe it was a radio ad that you knew after one sentence that nobody really wrote.

也可能是一个食物广告,灯光有点过于金黄,而且还有20个一模一样的广告。

Or maybe it could have been like the food ad with the lighting that is a little too golden and 20 more just like it.

或者是一个活动,随便哪家公司,其全部理念就是'现在用AI了'。

Or it was a campaign for literally any company whose whole idea was now with AI.

而最糟糕的地方之一就是那种每条帖子都以同样的钩子开头、以同样的问题结尾的feed。

And one of the worst places is the feed where every post opens with the same hook and ends with the same question.

而且它到处都是,看起来还都一个样。

And it's everywhere and it looks the same.

时间一长,你就干脆什么都不想读了。

And after a while you stop reading any of it.

说实话,我觉得我们每个人都有份,我们都是这件事的一部分。

And in all honesty, I think we all contribute and we are all part of the thing.

所以 Merriam-Webster 把2025年的年度词汇定成了 slop。

So Merriam-Webster word of the year for 2025 was slop.

那差不多是一年前的事了,他们的定义是:由人工智能批量生产出来的低质量数字内容。

That was almost a year ago, and they defined it as digital content of low quality that is produced usually in quantity by means of artificial intelligence.

而在这之前,它的意思是喂猪的泔水。

And before that, it meant slops you feed the pigs.

所以今年,我觉得情况糟糕多了,因为2026年正是各家公司把 agentic 工具推给所有人的一年,过去一次只来一份草稿的活儿,现在变成了 agent 大量产出文档、演示文稿和代码。

So this year, I think it got lots worse because 2026 is when companies rolled out agentic tools to everyone, and the work that used to arrive one draft at a time comes from agents producing documents, decks, and code in volume.

如果你想看办公室之外、更大规模的情况,Deezer 最近的统计显示,每天上传到平台上的音乐里,有一半以上完全是 AI 生成的。

And if you want even at scale outside of the office, recent stats from Deezer reported that more than half of the music uploaded to the platform every day is fully AI-generated.

但你要是看下载量,AI 生成的内容只占人们真正收听量的1%到3%左右。

But if you look at the downloads, the AI-generated downloads make up about 1% to 3% of what people listen to.

这就是背景,或者说定义。

So that's the background or the definition.

但最近让我对这个话题特别有热情的是,我和一些管理者、还有一位我非常尊敬的产品负责人做了一次培训。他们都非常懂 AI,思维也很敏锐,而且他们在非常用力地推动员工去用 AI。

But the thing that got me to be very passionate about this topic recently is that I had a training session with the managers and one product leader that I really respect a lot, and they're very AI savvy and very sharp thinkers, and they were kind of pushing their employees very hard to use AI.

每个人都用起来了,可这个团队却变成了一个 slop 工厂。

Everybody adopted that, but the team turned into a slop factory.

这是他们自己的说法。

That's their definition.

长长的输出,发出去之前谁都没读过;文案听起来像是同一个模型、同一套结果,哪家公司做出来都一样;最糟的是,几乎没有什么原创的思考。

Long outputs that nobody read before sending, the copy that sounds like the same model and the same results from any company, and worst of all, very little original thinking.

他们跟我说,这让所有人都变浅了——人变浅了,做出来的东西也变浅了。

What they told me is it made everybody shallow, the people and the artifact.

结合这些对话,还有我跟很多领导者进行的许多其他交流,我给它的定义是:本来该做思考的那个人基本上跳过了思考,不管是部分跳过还是完全跳过,而代价则转移到了接手这份工作的人身上。

The way I define that, given these conversations and many others that I'm having with leaders, is the work where the person who should have done the thinking basically skipped it, either partially or altogether, and the cost moved to whoever received the work.

当然,人类也会生产 slop。

And of course, humans produce slop.

早在有 AI 之前就会。

Way before we had AI.

AI 只是让它变得便宜,而且能量产。

AI just made it cheap and in volume.

所以这个定义其实也给了你标准:要让对方花最少的力气拿到最有效的产出,那才不算 slop 的工作;而把它压缩成你的受众需要的格式、长度和质量,这个责任在你身上,在负责任的 AI 使用者身上。

So this definition also gives you the standard because the most effective output for the least effort on the other side will be non-slop work and compacting it to the format, the length, and the quality that your audience needs is on you, the responsible AI users.

还有几个数字也印证了这一点。

A few numbers that agree also with that.

Stanford 和 BetterUp 做过一项关于工作 slop 的研究,他们说大约40%的美国办公族在过去一个月里收到过可以算作粗制滥造的东西,而在接收方那边,代价接近两个小时。

Stanford and BetterUp, they did a work slop study and they said that about 40% of US desk workers had received in the past month something to be considered sloppy, and it cost close to 2 hours on the receiving side.

在 Glean 的那份调查里——我记得你在某一集里引用过——今年有6000名员工表示,将近70%的人把工作成果发给别人之前,自己从来没读过。

In the Glean survey, I think you quoted it in one of the episodes, 6,000 workers this year said that almost 70% shipped work without ever reading it to someone else.

我猜你们有些人在说:不对,这些数字还算是低的。

And I suspect some of you are now saying something like, no, these numbers are even on the lower side.

在我们公司,比这多得多。

In our company, it's much more.

所以除了抱怨之类的,我想真正帮到你。

So beyond kind of complaining and so on, I wanna help you.

这一集我想分三部分来帮大家。

And the way I wanna help this episode is in 3 parts.

第一,解释我们是怎么走到这一步的,为什么 slop 会出现在机器里,也会出现在我们身上。

First, explain why we got here, why slop happens in the machine and in us.

第二,你怎么判断某个东西是不是 slop,不管是自己的工作,还是别人的工作?

Also, how do you know that something is slop, either in your own work or someone else's work?

最后,也是最重要的部分:有哪些习惯,能让这份工作始终出自你手,并且按照任务的分量来投入相应的功夫?

And finally, and the most important part, what are the habits that keep the work yours scaled to the stakes of the task?

我想把一件事说得非常清楚。

And I wanna make one thing very clear.

这一期不是在反AI。

This is not an anti-AI episode.

我做任何事情都会用AI来帮我。

I use AI to help me with everything that I do.

当然,也包括这一次的分享。

Including, of course, this session.

我只是想确保你用AI的方式,是真的能产出原创的东西。

I just wanna make sure that you do it in a way that actually produces original work.

所以我想先讲讲,为什么机器会产出垃圾内容。答案是:这些工具其实就是在做它们被造出来要做的事,因为语言模型就是在预测最有可能的答案。

So I wanna start with why the machine produces slop, and the answer is that the tools are doing what they were built to do because a language model is trying to predict the most likely answer.

你给它一个很泛的问题,你得到的大致就是互联网上对这个话题看法的平均值。如果你给工具加上一些记忆和上下文,那确实有帮助,但如果你今天没有把你的想法讲给它听,它们还是没办法让工具知道你到底是怎么看这件事的。

You give it a generic question and you get roughly the average of what the internet thinks about this topic, and if you add some memory and context to the tool, they help, But they still can't tell the tool what you actually think about it if you haven't shared your thinking today.

当然,还有另一个因素也在推波助澜,那就是训练之后的调优。

And then there's, of course, another thing that contributes to that, and that is the tuning after the training.

它是这么运作的:人们会给答案打分,而打分的人会奖励那些让人觉得熟悉的东西。

The way it works is that people rate the answers, and raters reward what feels familiar.

斯坦福的研究人员把这称为 typicality bias,而且在很多情况下,这会被揉进模型里,答案的范围就变窄了,默认输出就变得又长又满是列表,就像我们都知道AI能写出来的那种东西。

Stanford researchers named this the typicality bias, and that is being folded in many cases into the model, and the range of the answers narrows, and the default comes out long and very listy, like we all know that AI can create.

因为篇幅长看起来像是下了功夫,而在很多情况下,人们就是这么奖励它的。

Because length looks like effort, and in many cases, that's how people reward it.

如果你觉得换个厂商就能解决,那去年 NeurIPS 的最佳论文发现,来自不同公司的不同模型,最后会收敛到同样的答案上。

And if you think that swapping vendors can solve it, then the best paper last year from NeurIPS, they found that different models from different companies converge on the same answers.

所以换工具是买不来原创性的。

So switching tools won't buy you originality.

顺便说一句,图像模型那边也没好到哪儿去。

And by the way, with image models, it's not better.

差不多是同一回事,因为每个模型都有自己的招牌风格,如果你不描述你想要的其他东西,它就会退回到那个风格去。

It's kind of the same story because every model has their own house style, and it falls back to that if you don't describe something else that you want to do.

所以食物上那种金光,就是这种招牌风格之一。

So the golden light on the food is one of these things that is like a house style.

为了证明这一点——我也强烈建议你自己做一个类似的练习——我做了个实验。

And in order to prove it, and I strongly recommend to run a similar exercise yourself, I did like an experiment.

我做的就是,在无痕模式下打开了 ChatGPT 4o、Claude Opus 和 Gemini 1.5 Pro 的最新版本。

And what I did, I opened ChatGPT 4o, Claude Opus, and Gemini 1.5 Pro, the latest versions, in an incognito mode.

也就是说,记忆和历史记录都是关掉的,然后我给它们都下了一个非常泛、非常简单的提示词。

So that means that memory and history were off, and I gave them all a very generic and simple prompt.

“写一篇 LinkedIn 帖子,讲讲AI正在如何改变营销。”我得说,连我自己都对结果有多相似感到吃惊。

Write a LinkedIn post about how AI is changing marketing, and I have to say that even I was surprised by how similar the results were.

它们不只是结构非常相似——你们在这儿能看到,我觉得这个结构到现在大家应该都能一眼认出来——更夸张的是,内容也非常相似。

Not only did they have very similar structure, as you can see here, and I think this structure by now will be very identifiable by all of you, but even more so, the content was very similar.

然后我重新跑了一遍同样的任务,这次加上了我的观点、我对AI在营销中的看法、我的经验,还有我喜欢什么。

And then when I reran the same task now with my point of view, what I believe about AI in marketing, what's my experience, and also what I like.

在结构上,还有那些我从来不用的词方面,突然之间我就得到了一个好得多的草稿。

In terms of structure and words that I never use, all of a sudden I got a much better draft.

我能把这个原封不动发到 LinkedIn 上吗?

Could I have posted that on LinkedIn as is?

完全不行,但至少我有了一个好得多的编辑起点。

Not at all, but at least I had a much better starting point for editing.

这就是机器这一边。

So that's the machine side.

问题在于,为什么我们人类能容忍这种垃圾内容。答案是:心理学家大概从八十年代初开始,就把我们的大脑称为“认知吝啬鬼”,因为我们总倾向于在任何能省的地方省下思考和精力。

The question is why we as humans tolerate the slop, and the answer is that psychologists have called our brain a cognitive miser since the, I guess, early '80s because we are prone to save thinking and energy wherever we can.

就算你去跟飞行员聊,他们把这种负面效应叫做“自动化自满”:当机器通常都是对的时候,你就不再检查了。

Even if you talk to pilots, they call the downside automation complacency as something that when the machine is usually right, you just stop checking.

再说知识工作者,研究发现,人们越信任AI,他们对输出用的批判性思维就越少,而这些模型让人很容易就感觉自己特别有把握。

And even if you go into knowledge workers, research found that the more people trusted the AI, the less critical thinking they applied to the output, and the models make it very easy to feel very confident.

所以如果你想想垃圾内容是怎么产生的,基本上有三种方式,而且几乎没有一种是出于恶意。

So if you think about how slop gets made, there are basically 3 ways and almost none of them are bad intent.

因为这很容易。

It's because it's easy.

比如它看起来没问题,而你又压力很大,所以就直接发出去了。

Like it looks fine and you're under pressure, so you just forward it.

而且不懂这门手艺。

And not knowing the craft.

比如说,一个不太了解情况的人在那里 vibe code,他的请求就会很幼稚,很多情况下得到的答案也很幼稚;而一个资深工程师要同样的事情,能得到更好的效果,因为他知道该怎么问。

For example, if you compare someone who vibe codes without knowing much what they're doing, the request is gonna be very naive and they will get in many cases naive results, while a more senior engineer asking for the same thing, they get better work 'cause they know what to ask.

最后,也是让人最不安的部分,是因为我们把思考外包了出去。

And lastly, and that's the most troubling part, that is because we outsource the thinking.

我们在决定自己想什么之前,先问工具怎么想,仅仅因为这很省事,而且这些工具都非常有主见。

We're asking the tool what it thinks before we decide what we think, just because it's easy and the tools are So opinionated.

所以,防止这最后一点,就是我过去几周的使命。

So literally preventing this last one is my crusade over the last few weeks.

我称之为:找回失传的思考艺术。

I call it recovering the lost art of thinking.

在给你们一些关于如何避免 slop 的想法之前,我还想给你们一组问题,用来发现自己或别人作品里的 slop。

And before I wanna give you some ideas on how to prevent slop, I want to also give you like a set of questions to detect slop in yours or in other people's work.

好。

All right.

那么,要区分 slop 和好的 AI 辅助作品,这些是你可以问自己,或者在审查别人作品时能问的一些问题。

So to tell slop from a good AI-assisted work, these are some questions that you might ask yourself or when you review someone else's work.

我想问的第一件事是:发件人能不能不重读,就能为每一个句子辩护?

So the first thing that I want to ask is, could the sender defend each and every sentence without rereading that?

说实话,很多情况下我们做不到。

In many cases, in all honesty, we cannot.

第二件事是:发件人有没有在发出去之前,把这整篇东西都读完?

The second thing is, did the sender read all of it before it went out?

一份 40 页的报告,你读了一两页觉得好像没问题,然后就累了,于是你就把剩下的阅读推给了对方。

The 40-page report seems okay after you read 1 or 2 pages and then you just get tired, so you offload the reading to the other side.

我想让你问的另一个问题是:这里面有没有任何东西是只有这个人才能贡献的,比如一个观点、一个决定,或者真实的数据。

The other question that I want you to ask, is there anything only this person could have contributed, like a point of view, a decision, or real data.

如果随便谁都能在这份东西上署名,那就说明它不是相关的,也不是高质量的作品。

If anybody could put their name to this output, that means that it's not going to be relevant or high-quality work.

有没有一种更有效的方式来传达这些信息,还是说这一份仍然是能够做到的最好的方式?

Is there a more effective way to convey this information, or does this one still do the job the best it can?

也就是说,它总体上是在帮读者节省时间,还是在浪费他们的时间,而不是帮助他们?

Namely, does it overall save the reader time or cost them time instead of helping them?

最后一点,也是最重要、最直观的一点:你会不会愿意把自己的名字,用大写字母,放在这份作品上,而且感到安心?

And lastly, and that's the most important thing and the most intuitive one, Would you feel comfortable putting your name in all capital letters to this output?

如果你觉得署名不舒服,那就说明还有更多工作需要做。

If you don't feel comfortable putting your name, that means that there is more work to be done.

有时候,我们说不清楚为什么不愿意署名,但我们就是知道。

And sometimes we cannot explain why we don't feel comfortable putting our name, but we know.

所以这是我最喜欢这个测试的原因。

So that's why I like this test the most.

现在有一些技术方案试图解决其中一些问题。

There are a couple of technological solutions that try to address some of these problems.

在我看来,它们不怎么样,但我必须要提一下。

In my opinion, they're not that great, but I have to mention them.

其中之一就是用 AI 写作检测器,尤其是你在 LinkedIn 上开启它们的话,所有内容都会亮红,显示为完全或几乎完全是 AI 生成的。

So one is using AI writing detectors, and especially if you turn them on on LinkedIn, everything lights up as being AI-generated fully or almost fully.

这是有趣的部分,但它们要做的其实是衡量结构、冒号密度、要点密度、相同长度的句子,还有其他一些东西。不过你也可以用自己的作品跑一下检测器,作为一种努力跳出平均线的办法。但我不会给它们太多权重,因为我不认为它们能检测出质量或原创性——而那才是我们更应该在意的。

So that's the funny part, but what they're trying to do is measure structure, colon density, bullet density, same-length sentences, a few more things, but you can use a detector on your own work as a means to try and move out of the average, but I wouldn't give them too much weight as I don't think that they detect quality or originality, which is what I think we should care more about.

还有一种是 humanizer,就是像 AI 技能或 AI 流程一样,试图让文本听起来更像是人写的。

And there are also humanizers, which are like AI skills or AI processes trying to make text sound more human.

大部分时候,我觉得它们就是想让自己不像 slop 的 slop,因为 humanize 后的文本仍然有大约四分之三的概率被那些 AI 检测器标记出来。而读者的困扰从来都是缺少人类独特的视角,而不是具体的风格。

Mostly, I think that they are slop trying not to look like slop because humanized text is still flagged about 3/4 of the time by those AI detectors, and the reader's problem was always the lack of a human unique point of view, not the specific style.

所以你可以用它们来改进那些已经有点意思的东西,但我不会用它们作为跳过思考的手段。

So you can try them to improve things that are already interesting, but I wouldn't use them as a means to skip the thinking.

说到跳过思考,我想划清一条界线,一边是把所有事情都外包给AI,另一边是在人类判断仍然重要的地方获得效率提升。

And speaking of skipping the thinking, I want to draw a fine line between outsourcing everything to AI and still getting the efficiency boost where human judgment matters.

所以在某些地方,AI不仅能帮助我们,还能创造更多没有AI就无法实现的机会;但在其他地方,我完全相信那些应该由人类来做,你这里有一个部分清单,但我很好奇,因为我知道你对此有强烈的看法,Nathaniel。

So in some places, not only can AI help us, but it can also create many more opportunities that without AI we cannot do, while in other places I am wholeheartedly under the belief that those should remain human, and you have a partial list here, but I am very curious because I know that you are very opinionated about it, Nathaniel.

那么你如何决定哪些事情自己处理,哪些事情委托给AI?

So how do you decide what to keep in your hands and what to delegate?
M1
M114:12

是的,首先,我认为你说的要理解对于不同产出类型要有不同方法,这一点非常重要,对吧?

Yeah, so I think first of all, what you said about understanding that you're going to have different approaches to this for different types of things that you produce is really important, right?

对于一些你试图创造的原创内容和一个电子邮件,两者之间的要求有很大的不同。

There is a wildly different burden on some original content that you're trying to produce into the world versus an email.

事实上,我不认为有人会反对AI生成的电子邮件,只要它们简洁高效,直击要点。

And in fact, I don't think anyone has any problems anywhere with slop AI emails as long as they're concise and efficient and just getting to the point.

它几乎不算是垃圾内容,就是因为它完成了它的工作,然后就不碍事了。

It's almost like it's not slop by virtue of the fact that it just does its damn job and gets out of the way.

即使用一样的词汇也没关系,因为这就是它的目的。

That it doesn't matter if all the same words are used because that's its purpose.

它的目的不是让你思考或激发想法之类的事情。

Its purpose isn't to make you think or provoke ideas or whatever.

所以我对此有非常多的思考方式。

So there's a huge range of ways that I think about this.

也许让我用这个新播客作为例子。

And I guess maybe let me use the example of this new podcast.

我们是如何走到这一步的,因为我不得不有意识地处理在《AI每日简报》中自然演变出来的所有流程。

How we got here, because I've sort of had to get intentional about all these processes that I've evolved naturally over time with the AI Daily Brief.

而且我不得不以非常具体的方式将这些流程应用到新节目中。

And I had to put them into practice in a very specific way with this new show.

这个新节目,顺便说一句,说到与人们的注意力习惯完全不合拍的东西,对于那些还没听过的人来说,它是一个45分钟到一小时的深度探讨,探讨某个话题背后的整个文化、技术、社会、经济历史,才形成了现在这个样子。

So this new show, by the way, speaking of things that are completely out of sync with people's attention habits, for those who haven't heard it yet, it's a 45-minute to an hour-long deep dive on some topic about the entire cultural, technological, social, economic history that got it that way.

第一集是关于导致美国对科技失去兴趣的15年历史,这一历史也影响了数据中心的争论。

First episode is about the 15-year history that led to America falling out of love with tech that shapes the data center debate.

所以这是一个信息密度极高、篇幅很长、存在各种潜在问题的内容。

And so this is an extremely information-dense, long-form content piece that has all sorts of potential problems.

有信息来源的问题。

There's the problems of where the information comes from.

有保持观众兴趣的问题。

There's the problem of keeping people interested in it.

有推动内容的观点问题。

There's the problem of the point of view which drives it along.

有叙事结构的问题。

There's the problem of the narrative architecture.

这个节目若无AI,我不可能以现在的方式制作。

And this is a show that would be impossible for me to do as I do it without AI.

但如果处理不好,它也完全可能陷入AI的所有陷阱。

But also has all the potential to fall into every single trap of AI if not done well.

这个节目不是那种——我不是每个话题的专家。

And this is not a show where I can— I'm not an expert on each of these topics.

我最初做的几期,我更接近专家,因为这就是我选择从它们开始的原因。

The first ones that I'm doing, I'm much more, much closer to an expert in because that's why I'm starting with them.

但随着我越往远处走,比如做GLP-1的历史之类,即使我并非深入了解那个领域,我也需要把自己变成一种翻译者。

But as I get farther out and I'm doing the history of GLP-1s or something like that, like it's not— I'm going to have to sort of turn myself into the translator even without being deep in that space.

所以我思考如何‘去垃圾化’或确保过程中不产生垃圾内容的方法是,首先要清楚哪些部分是我的,哪些部分是AI的,以及它们如何结合,对吧?

So the way that I think about deslopifying that or making sure that there isn't slop that enters into that process is one, having a point of view first on which parts are mine and which parts are AI's and how those go together, right?

并且要知道这将是一个有意识的计划。

Knowing that it is going to be an intentional plan.

这是它建立的基础。

That's the foundation upon which it stands.

当我思考这个计划时,我知道的首要两件事是,第一,我无法完成所有的研究。

Two first things that I know when I'm thinking about that plan are that one, there is no way for me to do all of my own research.

所以这部分必须交给AI。

So that has to be AI.

但第二点,我不愿意让 AI 替我形成对一个话题的观点。

But two, I am unwilling to allow AI to form my point of view on a topic.

所以这就是——这些就是我的出发点。

So that's—those are my starting points.

所以我做任何一期节目,第一件事就是充分利用 AI 的能力——它能非常快地读完一个话题现存的、曾经存在过的一切材料,然后把它转化成任何一种能帮我快速上手的、有用的形式。

And so the first thing that I do with any episode is I am taking full advantage of AI's ability to very quickly read everything that exists and has ever existed about a topic and turn that into any medium of getting myself up to speed that's useful.

我可以要一份 5 分钟的概览报告。

I can have the 5-minute overview report.

也可以让它是一份 50 页的研究报告,什么都行。

I could have it be the 50-page research report, whatever it is.

但我会用各种各样的模型做出各种不同类型的研究报告,因为我发现它们在强调什么这件事上是有差别的。

But I'm producing research reports of all different varieties using all sorts of different models because there are differences I find in what things they choose to emphasize.

倒不是谁更好谁更差,我就是想把它们全都要。

Not so much better or worse, I just want all of it.

我用这些来确保:我要么拥有一个真正属于我自己的观点,要么就是在这个观点上不断质疑它、把它完善出来,不管我的起点是什么。

And I'm using that to make sure that I either have a point of view that is my own or that I'm sort of challenging it and building it out from whatever point I'm starting with.

为什么你真的没法把内容里的观点外包给 AI,我觉得有个特别好的例子。这个过程里我经常做的一步,就是让一个 AI 模型为两个截然不同的视角分别做最强论证,让它针对某个议题或现象的每一方都激烈地辩起来。

I think a great example of why you literally cannot outsource your point of view in content to AI is one part of that process that I often do is I will have an AI model steelman two wildly different perspectives and have it argue vociferously for each side of a particular issue or phenomenon.

这意味着,即便某一次运行里它碰巧选了这一边而不是那一边,它都完全能辩得像那是它最深的信念一样,你想让它站哪边,它就能站哪边。

And that means that it's even if in one run it happens to decide one versus another, it's completely able to argue as though it is its deepest conviction, whichever thing you want it to.

所以归根结底,它不是一个拥有观点的人,它只是它能搜集到的所有论点的总和,这就意味着你确实必须得有自己的观点。

And so the fact that it ultimately, it is not a person that has a perspective, it just has the sum total of all the arguments that it can go out and source, means that you really have to have that sort of point of view.

一切都从这里往上搭。

Everything builds from there.

所以我有研究这一块,基本交在 AI 手里,不过我确实也有一个研究团队帮着跟它互动,然后就是观点。

So I've got the research, which is largely in the hands of AI, although I do actually have a research team that helps interact with it, and then a point of view.

另外,就我自己而言,我在搭建叙事和论证流程方面也非常强。

I also tend to be very strong, for myself, in architecting the narrative and the flow of the argument.

这对我来说极其重要,而且我发现我几乎从来都不同意 AI 想怎么构建一个论证。

That's something that is extremely important to me and that I find I almost never agree with how AI wants to construct an argument.

我倒不觉得每个人都必须这么做。

I don't necessarily think that's something that every person has to follow.

你知道,我做这种长内容已经做了这么久,对论证应该怎么搭起来,我有非常具体的直觉和感觉。

You know, I've been doing this type of long-form content for so long now that I have very specific instincts and feelings about how arguments should be built.

但这又是那种我不愿意外包出去的事。

But that's one where, again, I'm not willing to outsource it.

所以人的这一侧,我有观点和论证的架构。

So the human side, I've got the point of view and the architecture of the argument.

然后是 AI 那一侧,我有研究基础。

Then I've got the AI side, I've got the research base.

从这儿开始,我讲的东西一律不写稿。

From there, I don't script anything that I do.

我都是即兴讲。

I do it extemporaneously.

但我之所以能做到这一点,是因为我先把那个叙事流程搭好,然后把我要用的所有素材、还有我边讲边放的视觉素材,全都按它们出现的顺序排好。

But the way that I'm able to do that is I architect that narrative flow, and then I put all the sort of materials and talkover visuals that I'm going to use in the sequence that they're going to come.

而这一步,又把 AI 请回来了。

And that once again, I've now brought AI back in.

所以我拿到的是非常精确、具体的叙事结构大纲。

So I've got the narrative architecture in very sort of precise, specific outlines.

然后我让 AI 来写要点,基本上就是关键事实或者关键数字,就是那些我特别想讲到的东西——跟我们为 AI Daily Brief 实际由人来做的研究档案是一个类型。

And then I have AI do bullets basically that have either key facts or key numbers, like basically the things that I wanna hit, the same type of research dossier that we actually produce with a human for the AI Daily Brief.

那是新节目 How We Got There 的人机合力成果。

That's a combined human-AI effort for the new show, How We Got There.

然后我就在这上面做我平常那种即兴讲述。

And then I do my normal extemporaneous talk over that thing.

所以这基本上就是一个没完没了来回往复的过程,其中最重要的部分——观点和叙事架构——始终握在我手里。

So that's sort of, it's basically this never-ending back and forth where the most important parts of perspective and narrative architecture remain with mine.

我用 AI 来让这一切真正可行,能按它需要的产出速度做出来。

And I use AI to make that all viable to do it at the speed at which it needs to get produced.

还有一点我觉得值得提一下,因为我已经大体上说过,基本上从一开始就没有使用任何粗糙生成的内容。

The one other piece that I thought was worth noting though, because I've sort of given this, basically there's no slop from the ground up.

我不会用AI来生成内容的脚本。

I don't use AI to produce the script of the content.

但对于每期节目,我都会写一篇配套文章。

But for each of these shows, I'm producing a companion essay.

而且我根本没办法在每周做这些节目和AI每日简报的同时,再写一篇6000字的配套文章。

And there is simply no way for me to write a 6,000-word companion essay in the same week that I'm doing these shows on top of the AI Daily Brief.

也许有人能做到,但我不行。

Maybe someone could, but I can't.

所以这不可避免地意味着,如果让AI写6000字的东西,那必然会是粗糙拼凑的内容。

And so that inevitably, there is no way to have AI write something that's 6,000 words without it being slop.

所以我基本上必须想办法应对这个情况。

And so I basically had to go figure out how to handle this.

因此,我用两种方法来‘去糙化’那个特定内容,因为我确实觉得它非常重要,因为很多人想阅读长篇内容而不只是听,或者他们两者都需要。

And so the two pieces to deslopify that particular piece of content, 'cause I did think it was incredibly important because a lot of people want to read something long-form like this rather than just listen to it, or they want both.

而且这种书面内容可以加入图表和视觉效果,这是纯音频无法实现的。

There's a chance to have charts and visuals that are totally different with that type of written content.

所以我做了两件事:第一,我没有让它仅仅根据我们构思的大纲来写,而是给了它我的旁白文本的原始转录。

And so the two things that I did were, one, instead of having it just write from the outline that we had concocted, I actually gave it the raw transcript of my talkover text.

这样它拥有的就不只是大纲了。

So it has more than just the outline.

它还包含了我实际的措辞和语气。

It has the actual way that I phrase things and intone things.

而且AI可以相当贴近这些,即使我允许它进行扩展,加入更多细节、语境、背景等等。

And it can hew fairly closely to that, even if I'm giving it permission to sort of expand and bring in more details, context, background, or whatever.

这是第一步。

So that was part one.

但即使如此,第二步是我发现这些文章还是不可避免地需要5到10小时的改写过程。

But even with that, part two was I found that there is just a can't-escape-it 5-to-10-hour rewrite process that I'm going to go through with those essays.

我实际上在How We Got There网站里建了一个内容管理系统,这样一旦我输出了这篇高度视觉化的文章,我就可以逐段地检查修改。

I actually ended up building a content management system into the How We Got There website so that once I've printed that highly visual essay, I can just go through paragraph by paragraph.

我可以重新排列段落。

I can reorder paragraphs.

而且我做了大量编辑,因为我的写作和讲话方式不像普通AI写出来的东西。

And I edited just a huge amount because my writing, my talking doesn't sound like the average AI writing.

有些地方甚至严格来说是错误的,就像英语老师会说写得不好,但那些错误正是我的风格。

There's things that are literally wrong, like an English teacher would say I do poorly, but they sound like me.

所以这是一种妥协,我认为要接受效率的局限性,本质上要承认:如果没有AI,这篇5000字的文章根本不可能存在。

And so this is a concession where I think accepting that there is a limit to the efficiency here, to basically embrace the fact that without AI, there's no way that this 5,000-word essay even exists.

但它仍然有成本。

But there is still a cost to it.

这个成本就是我要花5到10小时来编辑,让它不再有那些难以忍受的AI腔调,即使它遵循了我自己的叙述结构。

And that cost is the 5 to 10 hours that I'm going to spend editing it to make it not have these sort of unbearable AI-isms, even if it follows my own narrative architecture.

这些就是我最近经常思考的各种版本。

So those are the versions of this that I've been thinking about a lot recently.

我认为,主要的要点包括:第一,要有观点和视角,清楚哪些部分应该属于你,哪些属于AI。

I think, you know, in terms of big sort of takeaway points, the key elements are, one, having a perspective and point of view on which parts are supposed to be yours versus AI.

第二,我认为最终不能外包自己的观点,那个关键的东西是你试图实现或回归的核心。

2, I think not outsourcing one's point of view ultimately, that, that key thing that you are trying to achieve or come back to.

第三,接受没有完美效率这个事实,即使AI可以做到一切。

And then 3, accepting that there is no perfect efficiency, even if AI can do it all.

而且你必须在剩余的时间预算内想清楚,你愿意进行多少‘去糙化’工作,并视其为业务成本的一部分。

And you have to figure out with whatever time budget you have left, how much de-sloppifying you're willing to do and consider that part of the cost of doing business.
F1
F123:17

我觉得你可能用你那5到10小时的编辑时间吓跑了一些人,但让我们尝试提供一个更通用的视角,我认为这与你说的一切一致,但还给那些立场不同的人提供一些其他思考点。

And I think you might have scared away some people with your 5 to 10 hours editing, but let's try and give like a more general perspective that I think agrees with everything that you're saying, but gives maybe a few other points of thinking for people with other positions and so on.

听到你如何用AI优先和AI原生工作产出高质量的内容,真是太棒了。

So amazing to hear how you run high quality with like AI-first and AI-native work.

还有一点我想说,尤其是对那些觉得你的话听起来很吓人的人:需要根据重要程度来决定投入多少。

And one thing that I want to say, especially to the people who feel that what you're saying sounds intimidating, that it needs to be proportional to the stakes.

所以就像你说的,内部邮件里那行没人关心的文字,绝对可以完全由AI生成。

So like you said, the internal email, the one line that nobody cares about can definitely be fully AI-made.

而且我认为没人想要那种手工精心制作的会议记录或摘要。

And I think nobody's looking for like artisanal meeting transcript notes or summaries.

但我确实认为,任何代表你品牌的东西,应该是你的原创作品,面向客户且对公司影响很大的,这类工作需要这种刻意为之的态度,就像Nathaniel刚刚指出的,而且每个人都应该塑造自己的信念过程与优先级。

But I do think that anything that represents your brand, that's supposed to be your original work, that is customer-facing and has a high impact to your company, this is the type of work that requires this intentionality, like Nathaniel just pointed out, and that everybody should kind of carve their own belief processes and priorities.

至少,要思考一下这个过程。

At the very minimum, just think about the process.

这就是我关于每个人都应该如何deslopify的总体策略。

And that's my overall playbook on how I believe that everybody should deslopify.
F1
F127:10

我想把它分成几个阶段,第一阶段发生在你打开工具之前。

I wanna divide it into stages, and stage 1 happens before you open the tool.

第一件事,就像你刚才听到Nathaniel说的,是我认为你必须在打开工具之前知道你在想什么。

And the very first thing, like you just heard Nathaniel saying, is that I think that you have to know what you're thinking before you open the tool.

去散步,阅读研究资料,直到你形成自己的观点,不管它多杂乱。

Go take a walk, read the research until you have a view of your own, no matter how messy it is.

我最喜欢的方式是,一旦你有了一些原创想法,就把你想到的所有东西都进行brain dump。

My favorite version is to brain dump everything that you're thinking once you have some original thinking.

口述,对AI工具随意说,不要关心结构,就把所有东西都告诉它,你听到了什么,你担心什么,你在想什么,你确定不想做什么。

Dictate, ramble to the AI tool, don't care about the structure, just tell it everything, what you heard, what worries you, what you're thinking, what you're sure you don't wanna do.

然后让模型和工具来找出结构。

And then let the model and the tool figure out the structure.

比如这一集,对我而言,一开始是我对着AI工具讲了15分钟,非常混乱,但我只是把关于slope的所有想法都告诉它,来帮助写下这些内容。

This episode, for example, for me, started with 15 minutes of me talking to the AI tool, very messy, but I just told it everything I'm thinking about slope to help write this down.

然后AI工具会帮你梳理很多想法,或者把你说的内容回馈给你。

And then the AI tool will help you shape many of the thinking or speak back to you what it heard.

而且我认为,在很多情况下,这恰恰是人们在跳过的一件事,却有着最大的下游影响。

And I think that in many cases, this is funnily enough, the one thing that people are skipping that has the highest downstream impact.

所以这是第一点。

So that's number one.

你必须在打开工具之前思考。

You have to think before you open the tool.

第二个习惯是,我希望你也告诉工具你对你想的东西有多自信,因为模型天生就是赞同你的。

Habit 2 is I want you to also tell the tool how confident you are in whatever it is that you're thinking, because the model is built to be agreeable.

所以如果你带着非常强烈的观点,它很可能会帮你证明每一个强烈观点。

So if you arrive with a very strong opinion, it will probably help you justify every one of your strong opinions.

就像你说的,它会对你的观点进行steelman。

As you said, it will steelman you.

所以说出你有多确定,以及你希望它多努力地挑战你。

So say how sure you are and how hard you want it to challenge you.

比如,我大概有60%的把握。

For example, I'm about 60% sure.

问题在于onboarding。

The problem is onboarding.

在你同意我之前,请强烈反驳。

Push back hard before you agree with me.

那就会开始一场完全不同的对话。

That starts a completely different conversation.

即使你很确定,也要询问反方观点。

Even when you are sure, ask for the counter view.

这通常有助于磨砺你自己的观点。

Often it will help sharpen your own point of view.

在开始使用工具之前,我想要你做的另一件事是思考受众是谁,你对他们了解什么,以及你希望他们获得什么样的体验,无论你在创作什么。

The other thing that I want you to do before you start with the tool is to think who the audience is and what you know about them and what the experience is that you want them to come out with, whatever it is that you're creating.

可以是感觉,他们想要采取的行动,或者其他任何重要的东西,包括长度标准、格式、语气,任何能让它成为你的作品并且适合当前任务的东西。

It can be the feeling, the action that they want to take, or anything else that matters, including the standards of length, of format, tone, anything that will make it yours and appropriate to the task at hand.

它给模型一个目标,也帮助你在一起完成工作后对照某些东西进行验证。

It gives the model a target and it also helps you validate against something once you do the work together.

当然,所有这些都是在实际开始使用工具之前完成的。

And of course, all of that is done prior to the work that you actually kick off with the tool.

然后就是真正去引导 AI 做执行的部分,我们不妨叫它“带着执行”。

And then comes the actual, let's call it shepherding of the execution with AI.

那会是第二阶段。

And that's gonna be the second phase.

而在这里,我觉得你可以做的一件事就是多铺开几个方向,意思是第一个答案往往是最可能的那个,也常常是最平庸的那个。

And here, one thing that I think that you can do is to go wide, meaning that the first answer is probably the most probable one, and that's often the most average one.

但如果你让它给出 6 个方向,这些方向在方法、角度、结构、视觉,或者你正在做的任何东西上都不一样,它就会被推得更远。

However, if you ask for 6 directions that differ in approach, angle, structure, visuals, whatever it is you're producing, it pushes it further.

你可以要求一些从其他领域借鉴来的方向。

You can ask for directions that borrow from other domains.

比如,一个游戏设计师或者一个新闻编辑室会怎么处理这件事,然后让工具分别替每个方向做论证。

For example, how a game designer or a newsroom would handle this, and then make the tool argue for each direction on its own.

然后你眯起眼睛,把这些不同方向都看一遍。

And then you squint at the different directions.

看看每个方向里你喜欢什么、所有方向又都缺了什么,然后选择最吸引你的那些方向,也就是你觉得能服务当前目标的那些。

Look at what you like about each, what's missing from all of them, and then follow the ones that you are most drawn to, that represent something that you feel will serve the purpose at hand.

做完这些之后,我希望你非常有意识地去决定你的介入点,就像你听到 Nathaniel 做的那样。

After you do that, I want you to be very intentional, like you heard Nathaniel doing, about your intervention point.

看起来最高效的工作流——这是我要的,你从头到尾做完——往往造成的伤害最大,主要是因为当你面前摆着一个打磨好、完成度很高的东西时,沉没成本就进来了,然后很多时候你会说,好吧,我不想再回到真正的叙事上,就这么发出去吧。

The workflow that might look the most efficient, here's what I need, go and do it end to end, it does the most damage mostly because when you get like a polished and finished piece sitting in front of you, the sunk cost gets in and then you, in many cases, will say, okay, I don't wanna go back to the actual narrative, let's just ship it like that.

而且用工具的时候,通常也很难把它从这种状态再改回去。

Also with the tools, it's often harder to move it from this.

所以提前决定好你要在哪些点介入,我希望你把这些点定在非常战略性的位置,并且提前定好,比如大纲,然后第一部分,然后也许是第一版叙事,然后草稿,然后成品。

So decide in advance what the points are where you want to step in, and I want you to do that at very strategic points and decide them in advance, like the outline and then the first section and then maybe the first narrative and then the draft and then the finished good.

也就是说,要有意识地设一些战略性的检查点,你在那里做审阅,让修复或者彻底转向的成本尽可能低。

So there are intentional strategic points where you're reviewing that when the cost of fix or a complete pivot is as low as possible.

还有几个额外的审阅习惯,我觉得会非常有帮助,我个人非常非常喜欢,其中一个就是我所说的召集一个评审团,意思是任何重要工作发出去之前,让 AI 从不同的人物视角来审阅它。

And a few additional review habits that I think will go a long way, which I really, really like, include what I call convening a board, meaning that before any important work goes out, you have AI review it from distinct personas.

所以可以是第一次读到的读者、一个持怀疑态度的目标受众、一个领域专家、一个会收到它的最忙的人。

So it can be the reader seeing it cold, a skeptical target audience, a domain expert, the busiest person who will receive it.

很多情况下也要包括专门唱反调的人,因为这个人也可以是那个主张它根本不该发出去的人。

And in many cases also include the devil's advocate because that can also be the person who is arguing it shouldn't go out at all.

根据这个评审团能给你的不同视角去修改。

Fix based on the different perspectives that this board of reviewers can provide you.

当然,它也可以是你的目标客户里的不同人群。

And of course it can also be different segments of your target customers.

当然,只采纳你觉得有价值的反馈。

And of course, only use the feedback that you find valuable.

我觉得一旦你用上这个,你就能借助 AI 真正获得更多视角,并在发布前提高信心。

And I think that once you use that, you leverage AI to really give you more perspectives and improve the confidence before it ships.

所以当然,这既可以在早期关键节点做,也可以在成品上做。

So of course, this can be done both early on in critical junctions as well as on the finished good.

还有几个额外的审阅习惯,能让你的质量再上一个台阶。

And there are a few additional review habits that will take you even further with the quality.

关于审阅,我想让你做的第一件事就是尊重阅读预算。

The first thing that I want you to do with regards to review is to respect the read budget.

我觉得其中最重要的一点是,永远不要发出你自己都不会读的东西。

And I think the most important part of it, that you never send something that you wouldn't read yourself.

所以如果它对你来说太长或者不够清楚,那你的受众肯定也不会读。

So if it's too long or not clear for you, then your audience for sure will not read it either.

他们只会把它丢给自己的 AI 去总结。

They'll just hand it over to their own AI to summarize.

然后现在我们就变成两个模型在说话,而两个人在假装沟通。

And then now we have 2 models that are talking while 2 people pretend to communicate.

所以把一页内容浓缩成 5 句话,会逼着你去决定自己到底怎么想;或者把它生成成一套幻灯片、一个播客、一个 HTML、一张图或者一个视频。

So distilling a page into 5 sentences forces you to decide what you think, or generating that as a deck or as a podcast or as an HTML or as an image or a video.

用 AI,你有太多选择。

You have so many options with AI.

所以要有目的地选择格式,甚至在开始制作之前,先问问自己受众是谁,他们会以什么方式最有效地消费它,然后在这个方向上尽情发挥。

So choose the format on purpose, and before you even start the production, ask yourself about the audience, what they will consume in the most effective way, and then go wild with that.

你能做的最好的事情是,如果你自己喜欢用那种形式来回顾它,那么很可能你的目标受众也会喜欢消费那样的内容。

And the best thing that you can do is if you will enjoy reviewing that in that format, odds are that your target audience will also enjoy consuming that.

所以这就是我们如何让一个单一的作品出炉,而且质量更高,带有人类的品味等等。

So that's how we all get a single output out of the door in a higher quality with like human taste and so on.

但我也想提几个想法,关于如何随着时间的推移让它变得更简单、更个性化。

But I want to also propose a few ideas on how to make it easier and more personalized over time.

有一件事会很有帮助,那就是教AI你自己的品味。

And one thing that will help a lot is to teach AI your own taste.

当然,你不需要能用语言描述自己的品味,因为我们大多数人没有那样的词汇量,能像语言学家或艺术总监那样解释我们喜欢什么。

And of course, you don't need to be able to describe your own taste in words because most of us just don't have the vocabulary to be like a linguist or an art director that explains what it is that we like.

但如果我们给一个非常强大的模型看我们最好的作品、我们欣赏的别人的作品,以及我们不喜欢的作品和我们不喜欢的东西,然后问它那些好的作品有什么共同点,或者让它刻画你自己的品味和声音等等。

But if we show a very capable model either our best work, the work by others that we admire, and the work that we dislike and the things that we dislike, and then ask what the good ones have in common or how will it profile your own taste and voice and so on.

它能比你自己更精准地说出你的偏好。

It'll put words to your preferences much more precisely than what you can.

对视觉来说,我真的很喜欢做一个情绪板,放一堆图像、颜色,还有你想要从中获得的感觉,这比你能用的任何形容词都管用。

And for visuals, I really like to create like a mood board of a bunch of images and colors and the feelings that you want to get out of that, and that beats any adjectives that you'll be able to use.

一旦你对视觉、叙事、语气等有了很好的偏好画像,你就可以把它变成一个技能,或者变成你的AI工具里可以用的东西,然后那就可以成为任何想要听起来像你的工作的起点。

Once you have a good profile of your preferences with regards to visuals or narrative or tone of voice and so on, you can turn that into a skill or something usable with your AI tool, and then that can be the beginning of any work that is intended to sound like you.

不过,你仍然需要自己做思考,但至少它能帮助你听起来更像你自己。

Still, you need to do the thinking, but at least it will help you sound much more like you.

当然,拥有这样一个技能的缺点是,你可能会开始只产出同一样东西,因为这个技能会帮你一遍又一遍地做出来,而观众确实会记得你有重复性。

And of course, the downside of having such a skill is that potentially you will start to produce just the one thing because the skill will help you produce it again and again, and audience does remember that you have a sameness.

所以比如说,也许在我的例子中,你已经非常受够我的Falcons了。

So for example, maybe in my case, you're very much fed with my Falcons.

这可能是我把它固化下来的一个弊端,但你能做的是有意识地去打破它。

That might be a downside of me codifying, but what you can do is to break it down on purpose.

所以每当受众、范围或目标改变,或者时间过去了足够久,就从新的灵感中带来新鲜的想法,试着走出你自己的模式,但仍然教它你的偏好等等。

So whenever the audience, the scope, or the goal changes, or just enough time has gone by, bring fresh idea from new inspirations and try to kind of go outside your mold, but still teach it what you prefer and so on.

这就是我们要创造文化差异的地方。

This is where we want to create a cultural difference.

我想让你做的第一件事就是,每当你从别人那里得到slop,就明确指出来。

The first thing that I want you to do is every time you get slog from someone else, name it.

而且顺便说一句,如果你处于领导职位,尤其要这样,因为当有又长又泛泛而谈、明显没人会读的东西出现时,就直接说出来。

And especially, by the way, if you're in a leadership position, because when something long, generic, and clearly unread arrives, just say so.

你可以说得客气一点,也可以不那么客气。

You can do it kindly or not so kindly.

这不是在好好利用我的时间。

This isn't a good use of my time.

我希望你用五句话把你的观点发给我,或者我希望你发给我你更认真思考过的作品。

I want you to send me your take in 5 sentences, or I want you to send me work that you think about more carefully.

就把好关,别让slop到你这里来。

Just hold the gate and don't allow for slop to come your way.

当然,当是面向客户的工作时,这是你绝对不允许做的。

And of course, when it's customer-facing work, this is where you're never ever allowed to do.

因为当我们做对外作品时,我们绝不允许自己的名字被放在那种东西上。

Because we don't ever put our team name when we're producing an external work.

最后,我不一定要求你隐藏AI。

And lastly, I don't want you to necessarily hide AI.

如果你还记得之前,Nathaniel,你做了一期关于写作的节目,你们在讨论Stanley Druckenmiller在《华尔街日报》发表的那篇AI写的专栏文章,讨论的内容是我们是否应该去掉所有的破折号,去掉所有AI的痕迹,而你当时的观点主要是,那只是一种让观众觉得你没有投入时间的替代手段。

If you remember a while back, and Nathaniel, you did an episode on writing and you were debating on the Stanley Druckenmiller AI-written op-ed in Wall Street Journal, and the discussion was whether or not we should remove all the em dashes and remove all the AI telltales, and your point of view there was that primarily, that was a proxy to let the audience feel as if you didn't invest the time.

所以在我看来,你不应该完全试图隐藏AI。

So in my opinion, you shouldn't completely try to hide the AI.

你只需要让它感觉像是你本来就会创作出来的东西。

You just need to make it feel like something that you originally produced.

关于这个你有什么要补充的吗?

Anything to add on that?

因为那时候你非常固执己见。

Because you were very opinionated back then.
M1
M136:24

我今天本来要在今天的节目里用“load-bearing”这个词,但我真的是不能再用了,因为AI用得实在太多了。

I was going to use the phrase load-bearing in today's episode, and I literally just couldn't anymore because of how much it's been used by AI.

所以,从公司的角度再举个实际的例子:你开始看到一些走在最前沿的公司,把他们对这些问题的处理方式正式化、制度化。

So maybe another example of this that's practical from a company perspective: you're starting to see companies that are on the cutting edge formalize and codify their approaches to these issues.

所以 Figma 已经分享了他们关于 AI 写作的规则。

So Figma has shared their rules for AI writing.

其中一条是,如果是用 AI 写的,这本身不是问题,但要在开头标明,这样人们进入内容时就有这个背景信息。

And one of them is if it's written with AI, that's not an issue, but label it as such on top so that people have that context to go into it.

比如,很多时候你拿那份 40 页的报告,想做的并不是说“这是完成的作品”,而是说“咱们快速过一遍,找出有意思的部分”,对吧?

So for example, there's going to be lots of times where what you're trying to do with that 40-page report is not say, "Here's the finished work," but to say, "Let's rip through this, find the parts that are interesting," you know?

所以就明确标注出来。

And so you just identify it as such.

我觉得你一开始提到的那个观点——slop 的其中一个定义,就是把工作成本从产出它的人转嫁给消费它的人——我觉得这说得太准了。

I think the point that you made at the beginning of one of the definitions of slop being transferring the cost of work from the person who produces it to the person who consumes it, I think is acutely accurate.

我觉得这基本上几乎完全解释了为什么人们会对它有这么强烈的本能反应:你现在是在把成本强加给他们。所以这甚至不只是你不负责任。

I think that basically almost wholly explains why people have such a visceral reaction to it: you are now imposing a cost upon them. So it's not even just a dereliction of your responsibility.

某种意义上,你是把干活的包袱留给他们背。

You're leaving them holding the work bag in some ways.

所以,当人们足够聪明,能看出来在一个你本该向他们呈现东西的场景里,你把自己放在他们之上,我觉得这就是大家有意见的地方。

And so, when people are smart enough to know if you are prioritizing yourself over them in a context where you're supposed to be presenting something to them, I think that's what folks have an issue with.

所以我觉得,透明和披露很重要。

So I think transparency, disclosure.

明白不同场景、不同类型的工作有不同需求,这些背后都有一条共同的主线:像尊重自己的时间一样尊重别人的时间。

Understanding that there are different contexts and different needs for different types of work all have a common through line of respecting other people's time as much as you respect your own.
F1
F138:09

是啊,说到这儿,我们差不多可以总结一下了。我觉得往前看,模型肯定会越来越好,而它们越好,那种通用答案就会越诱人。所以我想劝你看看我们讲过的这五个阶段,并且承认不管接下来出什么新东西,它们都照样成立。因为我有个直觉:等眼下 agentic 工具的一大波热潮过去,再加上用 AI 生产东西这么容易,手工做的作品反而会变成高级货,人们会愿意为带有人类印记的作品多付钱。

Yeah, and I think with that, we can kind of summarize, and I think looking forward, for sure, the models will keep getting better, and the better they get, the more tempting the generic answer will become, so I want to urge you to look at the five stages that we went through and acknowledge that they hold whatever ships next, because my hunch is that once a lot of the current hype of the agentic tools and how easy it is to produce with AI, the handmade work will become the premium thing and people will pay more for work with a human fingerprint.

就像你说的,哪怕它有点乱,语法不一定完全正确,但它听起来像你,很有趣,能让人产生一种很人类、很直觉的感受。

Like you said, even if it's messy, if it's not necessarily grammatically correct, but it sounds like you and it's intriguing and it creates like a human gut feeling.

而且我真的相信,自律、高质量的工作,对每个担心自己工作的人来说就是职业保障,不管那工作是什么。

And I do believe that disciplined, high-quality work is the job security for everybody who is afraid about their jobs, regardless of what it is.

因为我相信,有原创思维的人会越来越吃香;而那些不让 slop 进入工作成果流转、尤其是对外流转的公司,会脱颖而出,并且以几乎无人能及的方式吸引客户——只要它们把通用产出和千篇一律的东西排除在外。

Because original thinkers, I'm sure, will be rewarded more and more, and companies that keep the slop out of how work travels, especially outside of them, they will stand out and they will attract customers in a way that almost nobody else will be able to manage should they leave the generic output and the sameness out there.

所以一定要用 AI 来产出更高质量、更多的工作,突破你能做到的边界;但我也希望你把它当成一个工具,用来更敏锐地思考、拓展视野和知识、看到更多方向,然后扛住一个更严格的审稿人可能会替你做的那些检验。

So definitely use AI to produce more quality, more work, push the envelope of the things that you can do, but I want you to also use it as a tool to think more sharply, to expand the horizon and the knowledge, and to see more directions and then survive the thing that perhaps a tougher reviewer might do for you.

所有东西都在配套页面上,你大概会把它放进 AIDB 网站上的节目笔记里。

We have everything on the companion page that you will probably put in the show notes in the AIDB website.
M1
M140:05

我想最后再补一句我比较愤世嫉俗的看法——说愤世嫉俗可能有点过了,应该说是从自身利益出发的看法。

I guess maybe I'll add my cynical take on this just to close it out, or cynical is overstating it, the self-interested take.

我相信,很多听这个节目的人都很有决心,铁了心要成为用这项技术领先的那群人,不让这项技术“发生在自己身上”,而是用它来实现他们想达成的任何目标。

I believe that a lot of the folks who listen to this show are determined and hell-bent on being in the group that uses this technology to get ahead, to not allow this technology to happen to them, but to use it for whatever goals they're trying to achieve.

我觉得这基本就是大家来这儿的原因,哪怕他们是为了新闻而不是实用内容来的。

I think that's mostly why people are here, even if they're here for the news versus the practical stuff.

总的来说,他们想成为这个 AI 参与其中的新时代里最顶尖的 1%。

Overall, they wanna be in the top 1% of this new era where AI is a part.

简单说,这意味着什么已经发生了变化。

And simply put, what that means has undergone a shift.

我觉得我们都意识到了这一点。

And I think we've all recognized it.

但一开始,用好它、进入那最顶尖的 1%,甚至前 10%,意味着你意识到你可以用 Gamma 更快地做出更好的演示文稿,视觉上更统一,对事情也有更全面的答案。

But at the beginning, using it well, being in that top 1% or even the top 10% meant realizing you could use Gamma to produce better presentations that were more visually consistent more quickly and having more comprehensive answers to things.

但现在,这些作为一种技能已经完全普及化了,对吧?

But all of that has been completely democratized as a skill now, right?

每个人都能接触到这些东西。

Everyone has access to those things.

不再只是那 1% 的事了。

It's no longer just the 1%.

而是所有在用这些工具的人,除非他们到现在还在理念上反对它们。

It's everyone who's touching these tools unless they're philosophically against them at this point.

所以,如果你想成为最顶尖的那部分用户,价值、差异化价值,就从只是产出一堆东西,转向把它们用好。

And so the value, the differentiated value, if you're trying to be in that top percent of users, shifts from just producing a bunch of stuff to using them well.

而“用好”,结果证明,意味着有品位、有意图。

And well, it turns out, means with taste and with intention.

不幸的是,这个阶段在很多方面都比第一个阶段要困难得多。

And unfortunately, it's, it's a much harder phase in a lot of ways than the first one was.

第一阶段,我们只是早早到场就获得了分数。

The first one, we all got points for just showing up to the party early.

现在产品的差异化实际上会变得重要,但在内容无限丰富的世界里,存在着巨大的机遇。

Now the differentiation in the product is actually gonna matter, but there are huge opportunities in a world of endless seas of content.

好的内容会变得更加突出,人们甚至不需要刻意过滤噪音就能获得极好的体验。

The stuff that's good will stand out even more and people will get enormously good even without trying at filtering out all of this noise.

所以这只是一个新的机遇,它需要不同的技能,但这就是接下来要做的事。

So it's just the next opportunity and it involves a different set of skills, but it's the thing that's next.

所以我认为这就是我们都要做的事情。

So I think it's what we're all gonna do.

太棒了。

Awesome.

好的,Noufar,很高兴再次邀请你来到节目。

All right, Noufar, great to have you on the show again.

就像Noufar所说,我们在网站上提供了一个配套体验,里面不仅有这些内容,而且比Noufar刚才提到的背景幻灯片要实用和具体得多。

Like Noufar said, we have a companion experience up on the website that has this, but even way, way more sort of practical and tangible than the background slides that Noufar was talking over.

所以请到aidailybrief.ai去看看吧。

So go check it out at aidailybrief.ai.

还有,呃,再次感谢你来到这里。

And, uh, thanks again for being here.
已剔除 4 处广告(点击展开查看)
M1
M10:57广告 · 已剔除

First of all, thank you to today's sponsors: KPMG, Robots and Pencils, CData, and Blitzy.

To get an ad-free version of the show, go to patreon.com/aidailybrief, or you can subscribe on Apple Podcasts.

And if you want to learn more about sponsoring the show, send us a note at [email protected].

M1
M11:14广告 · 已剔除

First, it is Monday, which means the new episode, just the second so far, of How We Got Here is out.

This one is called How We Made the Wrong Things Cheap or the Wrong Abundance.

And it's about how stuff got cheap while life got expensive and all of the attendant consequences of that change.

That is now available on all major podcast apps.

Again, the show is How We Got Here.

M1
M124:36广告 · 已剔除

A new study from KPMG and the University of Texas at Austin found that when people work with AI, similar skills don't guarantee similar outcomes.

Researchers studied more than 500 early-career professionals and found that the best performers consistently amplified the value of AI by guiding, evaluating, and refining its outputs.

These top performers, called AI Amplifiers, weren't defined by what they knew alone, but by how they worked with AI.

Learn more about what separates AI Amplifiers from everyone else at kpmg.com/us/aiamplifiers.

The best teams don't have a single star carrying everyone else.

They know their own strengths and each other's weaknesses and play to both.

That's the team Robots and Pencils has built on purpose.

Nobody there is grinding through busywork to pad a headcount number.

People come for the hard problems and they stay because everyone around them is leveling up at the same time.

In a market full of companies that are just trying to hire fast, that's worth a look.

Check out robotsandpencils.com/careers.

This episode is brought to you by CData.

Companies rolling out AI keep hitting the same wall.

Token spend is climbing, security is holding agents back, and no one knows how AI is truly being used.

The root cause: AI is cut off from the systems and context your business runs on.

That's why CData built Connect AI Gateway, one control point between your AI and your systems.

Built on the connectivity layer powering Palantir, Google, Salesforce, and hundreds more, each agent gets exactly the data and tools it needs.

It governs down to the record and lets you finally see your org's complete AI usage in one place.

Accurate answers, a fraction of the token cost, and a gateway that learns your business with every interaction.

See how at cdata.ai.

Every AI coding tool on the market does the same thing first: it starts writing code.

Blitzy does the opposite.

Before writing a single line, Blitzy spends days reverse engineering your entire codebase.

Thousands of agents ingest millions of lines, mapping every dependency, every undocumented constraint, every architectural decision made over the last decade.

The result is a dynamic knowledge graph that understands your software the way a principal engineer would after 30 years in the building.

Other tools guess at context with grep searches and markdown files.

Blitzy never guesses.

It builds true understanding first, then delivers over 80% of entire software epics autonomously.

Validated, end-to-end tested, production-grade pull requests.

That's why Fortune 500 engineering teams trust Blitzy with the codebases that matter most.

See for yourself at blitzy.com.

That's B-L-I-T-Z-Y.com.

F1
F139:40广告 · 已剔除

And as always, I have to say in one sentence that part of what we teach in the various courses that we do with the Superintelligent training is exactly to teach you how to be an intentional AI user, like from best effort to best-in-class, that does not outsource the thinking and is able to produce original work while still getting the immense value that the technology can give us, but doing so in a more strategic way.