从Tableau到人工智能：数据可视化的发展方向与安迪·柯克

欢迎回到PolicyViz播客。

Welcome back to the PolicyViz podcast.

我是你们的主持人约翰，再次陪伴大家探索数据可视化的世界。

I'm your host, John once again for your journey around data visualization.

这一次，在本期节目中，我很高兴邀请到我的老朋友Kirk，我们已经很久没见面了。

And this time on this episode and this week, I'm joined by my good friend, Kirk, who I haven't seen in many ages.

很高兴看到Andy回归节目。

It is good to see Andy, have him back on the show.

我们聊了很多内容。

We talk about well, talk about a lot of stuff.

Andy出了一本新书，是他著作的新版。

Andy has a new book, a new edition of his books.

我们谈到了这本书。

We talk about that.

我们还讨论了数据可视化的未来。

We talk about the future of data visualization.

我们聊了客户工作。

We talk about client work.

我们当然也聊了人工智能。

We talk about AI of course.

我们会讨论人工智能以及它将如何影响我们未来的发展。

We're gonna talk about AI and how it's going to impact our field going forward.

如果你从未听过我和安迪的对话，我想你会喜欢这一期的。

If you've never heard, got a chance to listen to me and Andy talk, I think you're gonna enjoy this one.

我们只是轻松愉快地聊天。

We basically just have a good time.

见到安迪总是很棒。

It's It's always good to see Andy.

所以我不再拖延了。

So I'm not even gonna delay any further.

接下来是我在Visualising Data与安迪·柯克的对话。

Here's my conversation with Andy Kirk from visualisingdata.

我相信你会喜欢的。

I'm sure you're gonna enjoy it.

给你。

Here you go.

他来了。

There he is.

看看这个帅小伙。

Look at that good looking guy.

哦，你是在说我吗？

Oh, you mean me?

诺亚在你后面，诺亚在那儿。

There's Noah there's Noah behind you.

抱歉。

Sorry.

你有你的信息，我根本不知道。

You've got you've got your inform I didn't even know.

你头顶上挂着的是你那些精美的奖杯。

You've got your information's beautiful trophies above you.

你还有那个大型数据可视化霓虹灯招牌。

You've got your big DataViz neon sign.

那么，这些奖杯你打算怎么处理呢？

Well, know what with those, right?

是的，它们就是。

So they're Yeah.

它们是2015年和2016年的胜利奖杯。

They're the victorious trophies from 2015 and 2016.

问题是，奖杯在什么时候不再是一种荣誉的象征？

And the question is, at what point is a trophy no longer a badge of honor?

它有保质期吗？

Does it have an expiration date on it?

对。

Yeah.

没错。

Exactly.

你知道，这现在真是，天哪。

You know, this this is now oh, god.

他还挂着那些吗，你知道的？

Has he still got those up, you know?

嗯，现在情况是，数据可视化奖由《信息之美》主办，或者也可能反过来。

Well, there is there is so like, you know, so now it's like the DataVisuals Awards brought to you by Information is Beautiful or maybe it's other way around.

我不知道。

I don't know.

所以

So

那些现在是有价值的。

those are like those have value now.

就像你收藏的那款火柴盒小汽车，全世界只有六辆。

That's like, you know, the matchbox car that own there's only like six of them.

那就是你拥有的东西。

That's that's what you've got.

好吧，就是这样。

Well, there you go.

我们走着瞧吧。

We'll see.

目前来说。

For now

事情怎么样？

How are things?

事情怎么样？

How are things?

很好，伙计。

Very good, mate.

谢谢。

Thank you.

是的。

Yeah.

一切都好。

All good.

很忙，这从来不是我可以理所当然的事情，即使我已经自由职业了大约十五年，我每天早上醒来还是担心这可能是我不再被需要的最后一天。尤其是在我知道当前市场略有放缓、需求略有下滑的情况下，我清楚地意识到自己仍能保持忙碌、同时参与多种不同项目，这是一种难得的特权。这些项目虽然各不相同，但都围绕着同一个核心理念展开，所以我没什么可抱怨的。

Busy, which is never something I take for granted even after whatever fifteen years ish, I think, freelancing, but I still wake up every day thinking this could be the last last day where I'm in demand, I don't take that from especially in the context that I do know that there's a general market slight slowdown, a little dip dip in the roads right now that I know that there's a few people experiencing, I'm very sort of aware of that, kind of privilege to still be busy and still be busy in lots of different things, which means I keep, you know, stimulated because they're all different but they all fit into the same central premise, so can't complain.

是的。

Yeah.

对，你看起来挺忙的。

Yeah, you seem busy.

今天早上我收到了你的通讯简报。

I got your newsletter this morning.

收到这个总是很有趣。

That's always fun to get.

你出了一本新书，或者新版的书。

You've got a new book out or a new edition of the book.

是的。

Yeah.

很多事情在进行中。

A lot of stuff going on.

我想聊聊这本书，但我想我们几个月前在你的播客上已经谈过一些了。

I do wanna talk about the book but I want to we talked a few, what is it, couple months ago now on your podcast.

我觉得大概是几个月前吧。

I think maybe a couple months ago.

我想听听你对这个领域现状的看法。

I wanted to get your sense of where things are in the field.

数据可视化现在的状况如何？

Like what's the what's the state of DataViz?

社交媒体现在分散了，这改变了对话和辩论的性质。

Social media sort of dispersed now, so that's changed the nature I think of of conversation and debate.

你最近对数据可视化整体感觉如何？

Like how are you feeling these days about DataViz sort of generally?

我认为我们正处在一个非常有趣的转折点，我想这个略带保留的说法意味着我并不确定未来的这个转折点是否全然是光明的，那条金光大道是否依然在前方，还是说我们只是达到了某种状态——我认为这并非饱和或完全成熟，但确实正在发生一些变化，比如我们在2010年代初期曾讨论并疑惑何时能解决的许多问题。

I think we're at very interesting junction and I guess that kind of slightly caveated phrase means that I don't quite know if this junction going forward is all going be rosy, you know, the golden path still ahead or if we've just reached a point of, don't think it's saturation or full maturity, but there's certainly something going on whereby maybe a lot of the things that we were having discourse around and wondering about when will this be solved in the early two thousand tens.

也许很多这些问题已经被归类解决了。

Maybe a lot of that stuff has just been boxed off.

是的。

Yeah.

你知道，比如某些工具的成熟度，我们在你提到的那期播客里讨论过Flourish和DataRapane的使用。

You know, so, you know, things like the the maturity of certain tools, you know, we've talked about use of things like Flourish and DataRapane in that podcast that you mentioned.

对。

Yeah.

这些工具与其他工具并存，比如Power BI、Tableau等世界主流工具。

And, you know, those tools are alongside all the others, you know, the the power BI, the Tableau of the world.

它们为人们提供了强大的手段，能够创建他们想创造的更多内容，并以相当良好的实践作为基础进行组织，尽管仍然有大量空间供创意发挥和调整。

They've really given people such a powerful means to create more things that they wish to create and with, you know, pretty good practices as the underpinning of how they're, you know, organized, albeit, you know, still lots of room for creativity and and modifying.

但对我来说，数据可视化领域确实存在着卓越之处——多样性、人才和持续的活力依然存在。

But it does feel to me that there is, you know, there's an excellence in the data viz field in terms of the diversity, the talent, the energy that's still there.

我认为社交媒体带来的碎片化是一个大问题。

I think the fragmentation through social media is a big issue.

我们似乎只看到了一部分，就像拼图只拼了一半。

It feels like we've got a partial view, we've got a partial jigsaw puzzle.

很难看出其中的模式，也很难弄清楚这些事物是如何连接在一起的。

And it's hard to know where the patterns are, it's hard to know where all these things join together.

感觉我们面前是一群群岛式的人群，比如在Blue Sky上出现了一些新兴社区，但仍然有Twitter用户，还有LinkedIn用户，不过后者更多是展示‘我做了什么’、‘我取得了什么成就’。

It feels like we've got a bunch of archipelagos of people, you know, there's a little kind of emerging community on blue sky, then you've still got some Twitter people, you've still got the kind of LinkedIn, but it's a bit more here's what I've done, here's what I've achieved.

是的。

Yeah.

就是那种‘恭喜我’类型的东西。

Congrats to me sort of thing.

Instagram上都是酷炫的内容，但很难把这些碎片拼成一幅关于当下正在发生什么、谁在其中的完整图景，不过我们之所以这么说，是因为我们身在其中。

The Instagram, the cool stuff, but it's hard to sort of piece all that together into a cohesive picture of what's going on right now and who's in there, but it still feels, I mean, it's easy for me and us to say because we're inside it.

仅从与那些在普通组织中从事数据库工作、但不属于这个领域的人交流的轶事来看，要达到‘我们已经彻底掌握了数据库这一实践’的地步，还有很长的路要走。

It still feels just from the anecdotes of working with people who are in normal organisations doing this, not part of the field but doing database, that there's still a huge way to go before anyone can say, yeah, we've nailed that, we've mastered database as a practice internally.

也许有了这些工具、这些参考资料以及我们所有人写的书籍，现在人们觉得他们不再那么需要外部的帮助了。

Perhaps with all these tools and with all these references, all these books that we've all written, maybe there's a sense now that they don't need as much help from outside.

是的。

Yeah.

也许内部的人就能解决这些问题。

That maybe people inside can solve these things.

但接着，我们总是会问一个问题：这到底带来了什么不同？

But then, you know, the question that we always ask is, so what's it's what's what's the difference it's making?

它产生了什么影响？

What's what's the effect it's having?

在一定程度上，除了公众层面的内容，比如新闻媒体，那些始终难以捉摸、并将持续如此的内容，是因为它们隐藏在企业活动的四面高墙之后。

To a certain degree beyond the public stuff, news media, for example, that is the stuff that always has and will remain elusive because it's behind the four walls of corporate undertakings.

是的。

Yeah.

对。

Yeah.

我同意所有这些观点，我也在想，数据可视化领域又有哪些不同的部分呢，因为数据可视化领域确实包含很多方面。

I I agree with all that and I and I wonder where the data viz field again, there's so many different parts of the data viz field.

所以很难一概而论说‘数据可视化领域’，对吧？

So it's sort of hard to say like the data viz field, right?

因为这里有研究人员、实践者，还有……是的。

Because there's researchers, there's the practitioners, there's the Yeah.

是的。

Yeah.

计算机科学。

Computer science.

我只是好奇，这些不同的群体如何看待这些问题，他们认为未来会走向何方。

I I just wonder where all these different groups see things, where people see things going.

对吧？

Right?

我们是否需要，你知道的，有没有什么新的进展或工具需要开发？

Do we need you know, are there are there advancements and tools you have to make?

我们是不是应该更关注实践者、自由职业者或培训师这一群体？

Is it should we be focusing maybe on the on the practitioner or freelancer or trainer side?

我们是不是应该更关注数据准备这一部分？

Should we be focusing more on the data prep side?

我的意思是，我不确定你怎么样，但这一部分我通常都会避开，就是你拿到一些数据，我们直接谈怎么可视化它。

I mean, I don't know about you but that's kind of the part I've generally sort of avoided, like, kinda like, you've got some data, let's talk about visualising it.

在它背后还有大量工作要做。

A whole boatload of stuff that goes on behind it.

完全正确。

Absolutely right.

我的意思是，我想我们之前聊过，之所以这部分要么很难做，要么容易被忽略，是因为首先每个场景中，对数据的处理方式、转换方式、准备方式和配置方式都是独一无二的，以便为使用做好准备，抱歉。

I mean, I think part of the reason why, you know, we were chatting before, but part of the reason why that is either hard to do or easy to not do is that first of all, every context is unique in the treatment that you need to give the data, the way that you need to transform it, prepare it, configure it, ready for its usage, sorry.

因此，要培训人们来做这件事很难，很难提炼出两三个可靠的案例研究，来应对大多数情况。

And therefore to train people to do that is hard to narrow down two or three reliable case studies that will cope, you know, give people most situations a way to cope.

是的。

Yeah.

所以我通常会以非常概括的方式谈论数据收集和转换，但在我心里，当我在这门课的30张幻灯片中讲这些内容时，我知道实际上这部分需要300张幻灯片才能讲清楚。

So I generally talk in quite general terms about data collecting and transformation and I know that in my mind when I'm saying all this stuff across, let's say, 30 slides of a course that that is something that needs 300 slides.

是的。

Yeah.

但总是这样，别担心，我们直接跳到酷炫的部分吧。

But, you know, it's always like, well don't worry about it, let's just get to the cool stuff.

我也觉得，你可能会对自己的能力感到自满，认为凭借多年积累的肌肉记忆，就能轻松地把数据从A点处理到B点，比如快速把Excel表格整理好。

I also think though that it's something that you can be complacent about in your own abilities to get from a to b with just the built in muscle memory that you've developed over years of how to quickly get a table in Excel into shape.

如何把它重新配置成Tableau所需的长而窄格式，或者适应Flourish中每种图表类型所需的独特数据表布局。

How to reconfigure it for Tableau, long and thin or into Flourish with all the unique sort of data sheet layouts you need each chart type to be plugged into.

对。

Right.

你往往忘了，并不是每个人一开始都知道如何快速复制一列并粘贴到表格底部，进行重复操作，做VLOOKUP，甚至写一点VBA代码来自动化循环。这里存在一种‘知识的诅咒’，我确实感到内疚，因为我在自己的材料中一直忽视了这一点，需要加以改进。

That you kind of forget that not everyone does know initially how to quickly copy and paste a column and stick it at the bottom and do the repeating, do V lookups and even just write a little bit of VBA code to automate a loop and so there is a curse of knowledge there which I do feel guilty of that I need to address that in my materials.

但我也同意你的观点，我认为这或许正是下一个需要关注的重点。

But I also, to your point, I do think that that is perhaps where the next focus has been.

要知道，多年来我们一直谈论那些小的可视化图表，我自己也包括在内，关注的是那些细微的细节。

Know, we We have for many years talked about the little viz, certainly myself included, about the minute details.

而或许在当下这个时刻，不仅在数据可视化领域，整个世界都该暂时把这类事情搁置一旁。

And maybe at this point in time, not just in data viz, in the world as it is right now, that stuff needs to be just put aside for now.

让我们关注一些更重要的事情：我们为什么要做这个？你想让人们了解或明白什么？也许更重要的是，那些隐藏在冰山下的庞大工作——数据处理的重头戏。

Let's care about some more important stuff which is why are we doing this, what's the thing you're trying to tell people or let them understand and maybe the bigger chunk of all this hidden iceberg which is the heavy lifting of the data side.

是的。

Yeah.

这很有趣。

It is interesting.

我不是Python程序员，但我一直用ChatGPT帮我写代码，比如抓取数据，用来做冰球相关的分析。

I use I'm not a Python coder but I've been using chat GPT to like write code to like scrape out of like, I, you know, like pulling up to do hockey.

这纯粹是出于兴趣。

This just got them for fun.

我不禁想，AI会不会改变人们对这一切的考量方式。

And I do wonder whether AI changes that calculus for people.

对吧？

Right?

你不一定非得知道，就我个人的经验来说，我会说，嘿，我想从hockeyreference.com的这个页面上抓取这个表格。

You don't necessarily know need to know, well, from my own experience, You know, I'll say, hey, I wanna scrape this table off of hockeyreference.com on this page, you know.

然后，我会给它一个提示。

And, you know, I'll give it a prompt.

不。

No.

好吧。

Okay.

这是代码。

Here's the code.

复制，粘贴，运行。

Copy, paste it, run it.

我刚开始的时候甚至都不知道该怎么运行Python代码。

I'll I'll say, even the beginning, was like, how do I even run Python code?

对吧？

Right?

好的。

Okay.

告诉我怎么运行它。

Tells me how to run it.

完全一样。

Exact same thing.

是的。

Yeah.

对吧？

Right?

运行它。

Run it.

好的。

Okay.

这里有一个错误。

Here's an error.

哦，我遇到了这个错误。

Oh, I got this error.

好的。

Okay.

这就是修复方法。

This is how you fix it.

哦，明白了。

Oh, okay.

现在可以运行了。

Now it runs.

你知道，人们当然可以用VBA做到这一点。

You know, people can do that obviously with VBA.

他们可以用所有这些工具做到，而关于ChatGPT的有趣之处在于，你只要说‘修复这个’，它就会修复并解释它所做的所有更改，但其实我根本不在乎这些。

They can do that with all these tools and in the funny thing about chat GPT is you'll say fix this, it'll fix it and then it'll explain to you all the changes that it made and it'll be like, I don't I don't actually I don't actually care.

我不是想学Python。

I'm not trying to learn Python.

我只是想完成一个任务。

I'm just trying to accomplish a task.

是的。

Yeah.

是的。

Yeah.

而且你说得对，很多这种小可视化更多是关于审美、设计、主观和人性的，对吧？

And to your point, a lot of the little viz is more aesthetic, design, subjective, human, right?

而AI其实无法回答其中一些问题。

Where AI can't really answer some of those questions.

是的。

Yeah.

这也是主观性存在的地方，对吧？

It's also the place where the subjectivity exists, right?

因为在这个领域的某些部分非常开放，所以不存在明确的规则触点，可能存在多个正确答案。

Where the rules don't have a touch point because it is so open in those bits of our world for there to be several right answers.

是的。

Yeah.

问题是，当你使用ChatGPT这样的干预工具时，你能用自己的眼睛验证结果吗？

The thing is that when you've got that chat GPT intervention is are you able to validate it with your own eyeballs?

所以你可以回看那个你抓取的原始冰球数据表，做大约10个快速检查，确认它看起来是否正确，我觉得应该是这样。

So you can look back at that original hockey table that you've scraped and you can do like 10 quick checks if that's looking right, if that's looking right, I'm pretty sure.

别误会，即使人工操作，人类也会出错。

And don't get me wrong, humans will make mistakes even if they do it by hand.

所以问题不在于我们是否完美无缺，而是当你引入其他工具大规模处理这类任务时，风险可能更大——比如，想象一下表格中出现了合并错误。

So it's not that, you know, we're flawless, but when you then introduce such assistance by some other tool to to do this on mass where there could be more jeopardy of it getting, you know, let's imagine there's a there's a merge sale in a table.

你知道，这种错误确实会发生。

You know, there's a Yeah.

这类错误总是会发生的。

These are always happening.

它无法正确处理这种情况，而且已经磨损了，结果所有东西都磨损了。

And it doesn't handle that correctly and it's worn out, and then everything's worn out.

你有

Do you

对。

Right.

有足够的能力来检查这些事情吗？

Have the the scope to be able to check those things?

所以这里仍然存在某种犹豫，但这并不意味着我们不应该使用这些工具。

So that's where there's still that, I guess, that hesitancy, but it shouldn't mean that we shouldn't use these tools.

我用Python也做过完全相同的事情。

And I've done exactly the same with with Python.

从来没学过，但现在我觉得我找到了一条入门途径，而不需要去。

Never learned it, but now I feel that I've got a route into it without having to Yeah.

学习它。

Learn it.

是的。

Yeah.

但AI最合理的位置就在这里。

But but it it is the logical place for AI to live.

对。

Yeah.

我认为在随意尝试和认真应用之间是有区别的。

And I think there's a difference between messing around.

我想画一个亚历山大·罗维奇金进球的折线图，但你知道，做那种会影响人们生计或福利获取的分析，对吧？

I wanna make a line chart of Alexander Rovichkin's goals and you know, doing an analysis that's gonna affect you know, people's livelihoods or how they get their benefit, right?

比如，这里面有一点 triviality（琐碎性）在

Like, there's there's a little bit Triviality on

这个话题上。

the topic.

是的。

Yeah.

是的。

Yeah.

没错。

Exactly.

你看，我搞错了又怎样，毕竟ESPN会做得对。

Like, you know, who cares if I get the thing wrong because ESPN's gonna do it right.

或者

Or

或者我就是当个爱好来做，而你以前可是和阿森纳合作过的，我记得。

or even just like I'm doing it kind of as a hobby whereas you've worked for you worked with arsenals, I recall.

是的。

Yeah.

阿森纳。

Arsenal.

对吧？

Right?

所以你和他们合作时，情况有点不同，因为你提供的分析会影响团队的决策，是的。

So like your work with them, it's a little bit different when you're like working with the team because they're making decisions based on your analysis and that's Yeah.

这是完全不同的情况。

That's a different story.

我想知道你目前对这个领域中的批评有什么看法。

I wonder what your take is right now on critique in the field.

因为当我们刚进入这个领域时，推特数据可视化正风靡一时，当时总有很多讨论，关于批评应该保持怎样的尺度、温度和态度。

Because when we were sort of coming in and the data the the Twitter data viz was all rage like, you know, there was always a lot of conversation about what is the right sort of level and temperature and attitude around critique.

我不确定，社交媒体的碎片化是否改变了你的考量？

And I don't know, has the fragmentation of of social media, has that changed your calculus?

你现在对这个问题的看法有不同吗？

Do you think of it differently now?

有一点。

A little bit.

我认为部分原因是，碎片化使得新兴的社群规模更小、更亲密、更个人化，而且目前仍然友好、融洽且具有建设性。

I think partly because the fragmentation means that the communities that are popping up are a little bit smaller, little bit more intimate, a bit more personal and at this point, still friendly and still convivial and still kind of constructive.

想想当事情变得如此庞大时，我不是要专门挑这个例子，但你可以想想Reddit。

Think when things become so huge, I mean not to pick on this but you know, think about Reddit.

当一切变得如此庞大、用户流量如此之高、且高度匿名时，人们的态度自然会比与熟人互动时更加放得开。但就批评本身的性质而言，我认为我们已经走了很长一段路，学会了放手。

When everything's so huge, so voluminous in traffic of people and so largely anonymous, you know, the kind of, the gloves are off a little bit more so than when you've got, you know, a person that you're interacting But in terms of the actual nature of the critique, I I think it's I do think we've come a long way in letting things go.

即使我们不一定认同所看到的选择，但经历了这么多，我们大概也明白，他们这么做很可能有其原因。

Even if we don't necessarily subscribe to the choices that we witness, we probably done enough laps of this world to know that there's probably a reason why they've done that.

是的。

Yeah.

与其我说，‘我讨厌你选的这个红色’，

And rather than me say, oh, I hate that red that you've picked.

我们或许可以想，嗯，他们可能被限制在一个有限的配色方案里选择。

We can probably think, you know what, they've probably been given that as a limited palette to pick from.

很可能有设计上的考量，他们可能已经用某种印刷方式测试过了，不管怎样。

There's probably a design criteria, they've probably tested it with a certain print, whatever it may be.

所以我觉得，当你在这一行摸爬滚打过一段时间后，就能稍微放松一点，心想：如果我能理解这件事背后的真实背景，大概就会觉得没问题了。

So I think when you have been around the block a few times, you are able to sort of just step off the gas a little bit and think, you know what, I'm sure if I understood the real context of that thing, that would be fine.

那将是次优的解决方案，而不是最优的。

That would be the least worst solution rather than the best.

所以在我看来，几乎可以肯定的是，我们不再那么急于扑向某样东西并指出它哪里可以改进了。

So I think there's just almost a certainly from my perspective, there's a little bit less of let's pounce on this thing and start to point out the ways it could be improved.

但我认为刚才那一点很重要，而且一直以来都是如此：如果你要批评某样东西不够好或不符合最佳实践，你就必须说明原因，并指出你自己会怎么做。

But I think that last point there is important, that and it's always stood the same way, is if you're going to have a go at something in terms of it's not good or it's not showing the best practices, you've got to back up with why and what you do differently.

而‘你具体会怎么做’这一点，必须考虑到当时的背景，我的意思是，十七年过去了，我们还在说‘这要看情况’，而且有充分的理由，因为确实如此——你并不知道人们经历了什么。

And the what you do differently therefore needs to be cognizant of what the context was, I mean, you know, seventeen years into this film we're still saying it depends and for good reason because it does, you don't know what people have been encountering.

是的。

Yeah.

但回到工具上来，拿走别人的设计就变得容易多了。

But back to the tools then, it becomes a lot easier to take someone's design.

我还没这么做过，但我猜现在把设计丢进像MidJourney或AI图像工具里，说‘把所有红色改成蓝色’，要简单多了。

I haven't done this, but I assume it's easier now to drop it into one of these, you know, like mid journey or AI image tools and say, hey, change all the red color to blue.

是的。

Yeah.

所以在某种程度上，这种批评——我不确定这是好事还是坏事，因为这种批评不一定需要你深入思考。

And so in some ways, the critique I don't if that's a good thing or a bad thing for critique because it it doesn't require you to think as hard necessarily.

当你构建那个仪表板或信息图时，有哪些限制条件呢？

Like, what are the constraints when you build that thing, that dashboard or that infographic?

你可以直接用工具快速完成它。

You can just sort of use a tool, to do it quickly.

但 presumably 它在某种程度上改变了权衡的逻辑。

But but presumably it changes the calculus a little bit.

确实如此，而且如果你还能访问数据源的话，你知道，越来越多的人会公开他们所使用的数据——虽然不是普遍现象，但确实有不少作品会附带数据集，相当于说：‘检查一下我的工作过程吧，这是原始材料，你来试试看能做出什么。’

Well, it does and and if you've got also access, you know, to to the to the data sources, mean, you know, increasingly people do publish the data that they've used for, not universal by any means, but there's a good fraction of works that do come with the data set to say, you know, almost like, check my workings, you know, here's the source material and almost like, have a go, you know, see what you can do.

但这很有趣，因为回到AI，我认为如果真能实现的话，一件很棒的事情是：即使你只有一张图表的PNG图片，也能……

But it's interesting thing because like going back to AI, I do think one of the I don't know if it exists yet, but one of the great things, if it did exist, would be that if you could take a chart, even just as a PNG of a chart Mhmm.

有没有办法从这张图中提取出数据？

Is there a way to scrape the data from that?

提取出这些数据的数值？

The values of that data?

请。

Please.

某样东西。

Something

通过编码反向转换成是的。

Through the encodings reversed into a Yeah.

我的意思是，如果真能实现，那将会是非常酷的一件事，我相信这是可以做到的。

I mean, that would be a that would be a super cool thing to have and, you know, I'm sure it's achievable.

它的准确性会更高，但作为起步，作为起步，我会

It'd be even more fraud for accuracy, but as a starter, as a starter I would

我的意思是，确实有个叫 Plot Digitizer 的工具，我从来没用过。

mean, there's this there's there is that tool plot digitizer which I've never gotten to

没有。

No.

我还没用过。

I've not got.

对。

Right.

但是，是的，你知道，哪怕只是随便做个原型，就说嘿，我想试试这种不同的图表，或者纯粹出于好奇。

But, yeah, you you know, even just to mock up mock something up, just say, hey, I would like to try this as a different chart or, know, just just out of out of curiosity.

所以你刚才的评论让我想到另一件事，我想问问你。

So so your last comment makes me think about one other thing I wanna ask you about.

让我们回到《信息之美》这本书。

So let's go back to the information is beautiful words.

你身后摆满了所有的奖杯。

You've got all of your trophies lined up behind you.

而且我们还有DataVisuals协会的年度会议Outlier即将在六月举行。

And we've got the DataVisuals Society has their conference outlier coming up in June.

他们现在负责颁奖了。

They have their they they now run the awards.

我很好奇，我知道我们过去也讨论过这个。

And I'm curious, and I know we've talked about this in the past.

我现在很好奇，你对数据可视化奖项有什么看法或想法？

I'm curious now how you feel or think about DataViz awards.

我要先说明一下，因为你刚才提到的数据问题，是的。

And I'll caveat that because of the point you just made about the data Yeah.

你并不知道人们在制作可视化之前对数据做了什么处理。

That, you know, you don't know what people have done with the data before making the visualisation.

所以这某种程度上是

So that's sort of the

是的，是的。

Yeah, yeah.

一个激励性的问题。

Motivational question.

这是个很好的观点。

It's a great point.

我正在努力回想，我记得弗朗西斯·加农也一直在强调这一点：如果你仅仅根据美学来评判这些作品，那就应该把美学作为奖项的评判标准。

I'm just trying to think, I think from memories from like Francis Gagnon as well has been pushing this button as well which is if you're judging these things just on the aesthetics, then kind of make that make that the goal of the awards.

如果评判标准是作品本身，那就需要更根本地判断：它是否忠于数据，是否在掩盖事实。但对我来说，我一直认为，对于这种高度主观的产品或作品，奖项本身并不值得过于纠结。

If it's about the work itself, then it needs to be a more sort of fundamental judgment of, you know, is it faithful to the data, is it disguising things, but again, think the awards to me I've always learned more on the side of not getting too wound up by the notion or the nature of awards for very subjective products or or works.

而且，我确实觉得，虽然你从不希望落选，或者连长名单都进不了短名单。

And I and I think really, although, you you never want to be, like, not winning or left out the long list into the short list.

但我认为，对大多数人来说，最好的态度或许是放松一点，简单地说：对于获奖者，这确实是一件很棒的事，就让他们享受这份荣誉吧。

I think it's played best for most people's buying tests to probably just chill a little bit and just say, well, for those who win, it's a really nice thing and let them have it.

对于评审者来说，这却是一种折磨，因为你得快速浏览这么多作品，时间有限，刚开始前五件你还很认真，但到了第十五到二十件时，你就想：天啊。

For those who are judging, it's a pain because you're rattling through all these different works, you're limiting time, you start off very thorough with the first five and then you get into the 15 to 20, it's like, oh, god.

是的。

Yeah.

我已经看够了。

I've seen enough now.

而且，公平地说，你提到的那些奖项，确实付出了巨大努力，让大量人关注这些作品。

And and I and I think when, you know, to be fair, the the awards that you mentioned there, they do go to a really great effort to get a load of eyeballs looking at these.

不是每个人都会看每一件作品，但这种方式确实能缓解‘只有一人评判’的印象。

Not everyone looking at everything, but certainly a nice way to just cushion the idea it's not just one person looking at a piece.

所以，再次强调，用这种说法，它是最不坏的方式，是的。

So, again, with that language of it's the it's the least worst way do Yeah.

如果你没赢，别在意。

This And if you don't win, don't worry about it.

如果你赢了，就把奖杯放在高架上十年左右，然后再问自己是否需要把它拿下来掸灰。

If you do win, stick it on a top shelf for about ten years and then ask questions about whether it needs to be dusted off and put away.

好的。

Alright.

这是柯克船长的睿智建议。

Sage advice from Captain Kirk.

所以我们已经讨论了，稍微谈了一些关于工具的事情。

So we've talked about we've talked a little bit about tools.

我们已经谈到了当今数据可视化创作过程中的许多不同方面。

We've talked about, you know, lots of different aspects of sort of today's data viz creation process.

我们还没谈到的一件事是，在这个过程中做它——我要直截了当地说。

The one thing we haven't talked about is doing it within this is a very I'm gonna say right up front by the way.

我现在这个过渡非常自然。

This is a very good segue that I'm doing right now.

我们还没有讨论过在团队或组织内部创建数据可视化的问题。

We haven't talked about creating data visualizations within a team or within an organization.

而你的书正是试图解决这一点。

And that's what your book sort of tries to

我帮你看清。

I help you see.

这就像一座桥梁。

That's what a bridge that is.

这很不错，对吧？

That's pretty good, right?

我也是刚想到的。

It just came to me too.

刚才没想起来。

Had not lit Yeah.

第三版。

Third edition.

第三版。

Third edition.

是的。

Yeah.

版本。

Edition.

实际上确实是。

It was actually.

想象一下，想象一下。

Imagine imagine

想象一下。

Imagine that.

你得让它支撑住。

You've got it to hold up.

看看这个。

Look at that.

当然，听众们看不到这个，但他们可以想象，因为大家都拥有它，那美丽的彩色漩涡。

Now folks listening of course can't see that but they can imagine it because they all have it with that lovely colorful swirl.

圆形的东西。

Circular thing.

是的。

Yeah.

那么，第三个是什么？

So what is what is the third?

什么是好的。

What is okay.

所以这是个两部分的问题。

So two two part question.

第三版有什么作用？

What does a third edition do?

第一版中有没有删掉一些内容，让人觉得我必须把三版都买齐？

Is it is it have you left things out from the first edition that people should be like, I need to get all three?

你是单纯为了增加收入才出第三版吗？还是说，第三版到底在做什么？

Like is this now now now are you just trying to like just gain revenue or like, yeah, what's the what's the third edition doing?

这有多 cynical（愤世嫉俗）呢？

Well, what level of cynicalness is it?

这正是问题的核心。

That's the thrust of the question.

其实，大概三年前，我被问过要不要做第三版，当然，这都是在一个小众市场背景下发生的，但出版方塞尔吉奥一直很友善，和他们合作非常愉快。

You know what, so I was asked quite a few marches, I think it might be three years ago, was asked agree to do a third edition and listen, this is all in the context of a niche marketplace but the publishers, Sergio, have always been really nice, really enjoy working with them.

他们之所以提出这个请求，是因为在他们的评估标准里，这本书销量不错。

They asked me that because within their parameters, it's seen as a good seller.

我从不说它是畅销书，但从他们的角度看，确实值得更新。

I'd never say best seller but it's worthy of a refresh from their perspective.

嗯。

Mhmm.

但当我被问到时，我在想，我真的不知道有什么新内容足以支撑第三版。

When I was asked though, I was thinking, well, I don't really know what is new to warrant a third edition.

但我在第二版时也做了同样的事。

But I add that as well with the second edition.

第二版的情况通常是，不妨说第一本书像是初稿，这次我们能做得更好吗？

Now the second edition that case is often a little bit of a, well let's say that the first book was a first draft, can we do it properly this time?

我们能否消除一些低效的表达？回头看看，我确实觉得天啊，我当时浪费了不少词句。

Can we eradicate some of the inefficient language, which I definitely look back and think, Jesus, I was chewing up some of those words there.

但第三版很有趣，因为当你从远处看时，距离第二版已经过去五六年了，我感觉在这五六年里，我对数据可视化的理解一直在不断演变。

But the third edition was interesting because actually when you look at it from a distance, is five, six years since the second edition and I felt that in that five or six years, I've continued to evolve in my lens of what data viz means to me.

是的。

Yeah.

你总是希望用新的例子和新的参照点来更新内容，以反映最新的工作和新涌现的从业者，这一点确实存在。

And you always want to update things with new examples and new reference points to reflect the new work and the new people doing the new work, so there is that part of it definitely.

但当我去年一月着手修订时，距离截止只剩两个月，我发现每一页都让我看到改进的机会——要么重写，要么删减，要么添加一些让我觉得更能表达新想法的内容。

But what I found actually interesting when I got into it last January when I had two months left to do it was that every page I was looking at was seeing opportunities to either rephrase stuff, remove stuff, or add stuff that just felt new ways to express things.

而我对这一切的总结是，对我来说，这仍然是同一个房子。

And and the way that I've kinda concluded all this is that it to me, it's the same house.

如果这本书是一座房子，那它还是同一个房子，同样的房间，这些房间的功能依然不变，厨房还是厨房。

If the book is a house, it's the same house, the same rooms, those rooms still save the same purpose, the kitchen still exists as the kitchen.

嗯。

Mhmm.

但我拆掉了许多橱柜和电器，换了墙纸，换了地板。

But I've ripped out a lot of the units and the appliances, I've changed the wallpaper, changed the flooring.

我拆掉了那个棚子，这是我唯一去掉的部分。

I've got rid of the shed, so that there is one bit I got rid of.

我们把棚子拆了，因为它没什么用处，整个事情几乎就像是重新设计、重建一样，但对我来说，这代表着我在思考这个领域以及如何表达这个领域方面向前迈进了一大步。

We got rid the shed, that wasn't serving any But purpose yeah, the whole thing is just being almost, you know, re engineered, reconstructed but it feels to me like, you know, a bit of a leap forward in terms of how I think about the field and how I want to express the field.

我想说，最大的一点是，我从不希望任何人因为已经拥有我的书而再花钱购买，当然，有一些收藏爱好者我很欣赏。

I would say that the biggest, you know, I never want anyone to spend more money on books from me if you've already got the books or one of the There are some completists out there who I love, daily.

但我建议，如果你拥有第一版，那么如果你希望扩充你的书架，第一版和第三版之间已经足够不同，值得考虑入手第三版。

But I would suggest that if you've got the first version, sorry, the first edition, there should be enough clear water to warrant possibly a third edition if you've got a desire to kind of grow your bookshelf.

如果觉得有什么缺失，我总是建议先买别人的书。

I would always say buy someone else's book first if you feel there's something missing.

也许从外部视角来看，第二版到第三版之间的差距并没有我感受到的那么大。

Maybe the second edition to the third edition to an outside lens isn't as big a gap as I see it.

嗯。

Mhmm.

但我从不希望对人们如何花他们辛苦赚来的钱购买我的产品做出预判。

But, you know, I never want to prejudge how people want to spend their hard earned money on my products.

但对我来说，更重要的是那些刚进入这个领域的人。

But to me, it's more just those arriving into the field now.

那些第一次购买我书籍的人。

Those for whom this is the first book that they buy with me.

我有信心，是的，这一版是我大脑中最新的版本。

I feel confident that, yes, this edition is the latest version of me in my brain.

对。

Right.

对其他人来说，是否要升级到新版本由他们自己决定。

For others, it's up to them if they want to decide to jump onto a new edition.

但这确实是一个有趣的挑战。

But it's an interesting challenge.

你已经出了第二版，对吧？

You you've done the second edition, right?

我没有。

I have not.

哦，你有啊。

Oh, you've Yeah.

我提过要出《Excel》这本书的第二版，但那是另一回事。

Talked about And I've talked about doing a second edition of the of the Excel book but that's a different game.

这是因为技术和工具都变了，Excel 本身也发生了很大变化。

That's because the the technology and the tools have changed and Excel itself has has changed a

很多，是的。

lot of Yeah.