Slack’s CMO on AI Agents and the Future of Work
发布时间 2025-07-17 17:13:12 来源
Work is going to change. It's going to change more in the next five years than it has in the last 20 years. Undoubtedly. Jobs are going to change. Undoubtedly. But the human capacity and the human capability, I think what many people miss is AI and things like agent force and a lot of the stuff we're doing with Slack. It's really about giving people a capability upgrade. And with that capability upgrade, they can do so much more than they couldn't do before.
工作将要发生变化。在接下来的五年里，工作变化的程度会超过过去20年，这是毋庸置疑的。工作岗位将会发生变化，这也是毫无疑问的。但很多人没有意识到的是，人类的能力和潜力，特别是通过人工智能、智能代理技术以及我们在使用 Slack 等工具时，实际上是在为人们提供能力的提升。有了这种能力升级，人们可以做到以前无法做到的事情，还能做得更多。
What's up, everyone? And welcome to another episode of the AI Download, your weekly recap of what's happening in the world of AI, including the stories and leaders behind the future. I'm your host, Cheryl Lazar. I'm the CEO and founder of What's Trending. And before we get into the show, a big shout out to our sponsor, this episode is brought to you by Profit, Suite of AI-powered SaaS software and services designed to empower and support the next generation of human-led AI fed communications engineers working in the PR, social, and influencer marketing space.
大家好！欢迎来到《AI下载》的又一期节目，这是您每周一次的AI世界动态回顾，包括未来的故事和领军人物。我是您的主持人 Cheryl Lazar，也是“趋势热点”（What's Trending）的首席执行官兼创始人。在节目开始之前，特别感谢我们的赞助商。本期节目由 Profit 提供支持，Profit 是一套由AI驱动的SaaS软件和服务，旨在赋能并支持下一代以人为主导、以AI为辅助的信息交流工程师，他们在公关、社交和影响者营销领域工作。
You got to learn more. Go to PRprofit.ai, profit human-led AI fed. Also, have later, they're your AI-powered reading assistant bringing you smart, personalized insights. So you can spend less time scrolling and more time doing. Now, today on the AI Download, we're joined by Ryan Gavin, the chief marketing officer at Slack. We're here shaping the future of AI-powered collaboration. Ryan's at the helm of integrating generative AI and intelligent agents into the weight millions of us work every day on Slack.
你需要学习更多。请访问PRprofit.ai，这是一个由人类引导、AI提供支持的获利平台。此外，还有Later，这是一个由AI驱动的阅读助手，为您提供智能化、个性化的见解。这样，您可以减少浏览时间，投入更多时间去行动。现在，在今天的AI 下载中，我们请到了Slack的首席营销官Ryan Gavin一起探讨未来AI驱动协作的前景。Ryan正在引领生成式AI和智能代理的整合，这种整合影响着全球数百万人在Slack上的日常工作方式。
So let's get into the show. Welcome to the AI Download. Thanks so much for being here. I really appreciate it. Oh, sure, man. Thanks so much for having me. It's so great. I'm so glad to be on the pub. Yeah, and this is a very big day week for you all. I mean, some big announcements from Slack. We're going to be getting into later in the show, right? Yep. We've got a lot of cool stuff coming, but yeah, happy to talk about whatever and whenever. OK, that's a tease, everyone. So stick around.
那么让我们开始节目吧。欢迎来到《AI 下载》。非常感谢你来到这里。我真的很感激。哦，当然，伙计。非常感谢你邀请我。真是太棒了，我很高兴能参与这个节目。是的，对你们来说这是非常重要的一周，有很多来自 Slack 的重大公告，我们稍后在节目中会谈到，对吧？是的，我们有很多酷炫的内容即将推出，不过随时愿意聊聊任何话题。好的，这只是个小预告，大家一定要继续关注哦。
First, we've got a quick look at some AI headlines for this week. OK, this is from NPR, a recent high profile case of AI hallucination serves as a stark warning. So this is what's going on. A federal judge just finds two My pillow attorneys, 3,000 each after they submitted a court filing filled with hallucinated legal citations, AKA fake cases generated by AI. Ooh, OK, well, this is the kicker. They didn't initially admit they used AI. And the judge wasn't buying the oops wrong draft excuse. That's like the new My Dog 8, My Homework excuse. Totally. Right?
首先，让我们快速浏览一下本周的一些人工智能新闻头条。这是来自NPR的一则新闻：最近一宗引人注目的AI幻觉案例发出了一个明确的警告。事情是这样的：一位联邦法官刚刚对两名“My Pillow”公司的律师分别罚款3000美元，因为他们向法庭提交了一份充满虚假法律引用的文件，也就是由AI生成的虚假案件。更有意思的是，他们一开始并没有承认使用了AI，而法官也不接受他们“发错版本”的借口。这就像新版的“我家的狗把我的作业吃了”的借口，是不是很像？
And it's actually part of a growing trend of lawyers getting trouble for unverified use of generative AI with over 200 similar cases tracked since just the spring. That is wild. One, I'm sure you're making sure your lawyers don't do this, right? I just said, I'm a note, just like this is all news to me. So I just had like far enough now. Yeah, you should actually copy and paste this like FYI. But you know, AI hallucinations are still a big issue. And for some people, we might not know what that is. And basically when your AI goes out of whack, like it spews out wrong answers that don't make sense.
这实际上是律师因为未经验证的生成式AI使用而陷入麻烦的一个增长趋势的一部分，自今年春季以来，已经有超过200起类似的案件被追踪到。这真是太疯狂了。首先，我相信你肯定确保你的律师不会这样做，对吗？我刚说，这对我来说都是新闻。到现在我已经了解到足够多的信息了。你确实应该复制粘贴这些内容，就像提醒一样。不过，你知道，AI幻觉仍然是一个大问题。对于一些人来说，我们可能不知道那是什么。基本上，当你的AI出现问题时，它会给出不正确且不合逻辑的答案。
So I guess is this something that you're looking at even at Slack? Yeah, I mean, 100% one of the things that we do, and anytime we use AI and Slack, we're actually looking at the conversations that are happening at Slack. So it's the one we're using AI. It's based on those conversations that are happening. And so whenever we have an AI answer, we have the citation for where the model got that context from. So if I see an answer, I'm like, it doesn't look quite right. I get a link right there that takes me directly to the conversation that the model used to generate that AI answer.
所以，我猜这也是你们在 Slack 中关注的事情吧？是的，100% 的确如此。在我们使用 AI 和 Slack 的任何时候，我们都会查看 Slack 上发生的对话。因此，我们使用的 AI 技术是基于这些对话的。每当我们有一个 AI 答案时，我们都会提供模型获取上下文的出处。如果我看到一个答案觉得不太对劲，我就可以点开链接，直接查看模型用来生成这个 AI 答案的对话。
And I can kind of verify it. I can say, oh, OK, got it right. Maybe I was wrong. Actually, you've interpreted this incorrectly. Here's the actual source. No, and I love that. So thank you for sharing the citation part that's really cool and needed. So let's move on to the story from Reuters, the EU's AI code of practice for companies to focus on copyright and safety. So this is what's going down. The rules kick in August, 2026 for new models. It applies to companies like OpenAI, Meta, and Google, and many others.
我可以稍微验证一下。我可以说，哦，明白了，做对了。可能我错了。实际上，你的解释不对。这是实际的来源。不，我喜欢这样。所以感谢你分享引用部分，这是非常棒且必要的。那么我们来讲一讲来自路透社的故事：欧盟的公司AI行为准则将重点关注版权和安全。这就是目前的情况。新规定将在2026年8月开始生效，适用于OpenAI、Meta、Google等许多公司。
And the signatories must disclose training data and tackle systemic risks. So they are signing up to say, we want to be part of the change and creating solutions, not continuing to keep things behind closed doors. What is your take on this? Because obviously you're a US-based company, but you are also a company that is used by everyone around the world. How are you working with AI compliance? Yeah. Well, I think that particular act is really focused on a lot of copyright issues with the public LLM models that are training on internet data.
签署方必须公开训练数据，并解决系统性风险。这意味着他们表明愿意参与变革，创造解决方案，而不是继续将事情隐藏在幕后。您对此怎么看？显然，您是一家总部位于美国的公司，但同时您又是一个面向全球用户的公司。您是如何应对AI合规问题的呢？嗯，我认为这个行动主要关注的是公共LLM模型在互联网上进行数据训练时的许多版权问题。
Slack's obviously very different. We look at customers data, which is customers own. It's their data. It's theirs to do with what they will. And so our job is to keep that data safe, secure, and private, and do so in a very transparent way. We're never going to train an LLM on your data. And so the idea of your data going outside of the walls of your company just never becomes an issue. Yeah, and love that. Very important. And I think that this stuff that they use doing will set a precedent. It'll be interesting to see how this goes down. And if it's helpful, because maybe it could help the way that we deal with this stuff. I don't know, just saying. Transparency in the space is a very good thing. So the more transparency we have, the better. There's still a lot of lack of trust and a lot of what's happening here. And so that's an important framework for sure. Thank you. You're a great exec.
Slack显然非常不同。我们查看的是客户数据，这些数据是客户拥有的。是他们的数据，由他们自行处理。因此，我们的工作是以一种非常透明的方式来保护这些数据的安全、私密性。我们绝不会用你的数据来训练大型语言模型。所以，数据离开你公司的墙壁这个问题根本不会发生。是的，我很喜欢这点，非常重要。我认为他们在做的事情会树立一个先例。看到这会如何发展以及是否有帮助将会很有趣，因为也许这对我们处理这些事情的方式有所帮助。我只是说说而已。行业中的透明性非常好，所以我们有越多的透明性越好。对于这里发生的许多事情仍然存在很多不信任，这是一个重要的框架。谢谢，你是个出色的管理者。
OK. From CBS News, Indeed and Glassdoor to lay off 1,300 workers as AI shakes up job search business. So this is pretty big. So not only are jobs being cut from companies, but even the hiring business is being hit by AI. I mean, I guess as a company that needs to have people there, humans there to use the product, what do you think about AI taking over businesses and jobs? Yeah, I don't think it's taking over businesses and or jobs. I think that the headlines here can be a sensational thing. But I think obviously work is going to change. It's going to change more in the next five years than it has in the last 20 years. Undoubtedly. Jobs are going to change. Undoubtedly. But this is really about like human capability getting an upgrade.
好的。这是一则关于《CBS新闻》的报道，Indeed和Glassdoor将裁员1300人，因为人工智能正在动摇求职行业。这是一个相当大的变化。不仅公司在裁员，连招聘行业也受到人工智能的影响。作为需要人力来使用产品的公司，你怎么看人工智能对企业和工作的影响？
我认为人工智能并不是在接管企业和工作。媒体的标题可能比较耸人听闻。但显然，未来的工作方式会发生变化。未来五年的变化会比过去二十年更大。这是毫无疑问的。工作岗位的性质将发生变化。这实际上是关于人类能力的升级。
And I tell you what, if you're a small business, this is the most exciting time ever. Because small businesses can start to act like large businesses with that typically we're out of reach because the scale and resources they had, small businesses now have access to that same scale with all this AI agents in the autonomy. So it's pretty exciting time. And all that growth, what does that do? That generates jobs, that generates things for people to go do. Is the small companies get bigger, they need more people. So I think it'll all net out. But there is going to be change for sure. This final story is from payments. Reddit posts reignites debate over AI's rule and medical advice. So a viral Reddit post claims chat, QBT flagged a genetic mutation doctors missed for over 10 years. This person went to multiple doctors X-rays, MRIs did not figure it out, but chat, QBT did. And a doctor later confirmed that. Is this something you do to use AI for diagnostics?
我告诉你，对于小型企业来说，现在是最激动人心的时刻。因为借助AI代理和自动化技术，小企业现在可以像大型企业一样运作，而过去由于缺乏规模和资源，这对小企业来说是遥不可及的。如此一来，小企业也能享有相同的规模优势，真是令人振奋。而所有这些增长会带来什么呢？会创造就业机会，为人们提供更多的工作机会。随着小公司不断发展壮大，它们需要更多的人手。所以我认为最终结果会是积极的。但变化是不可避免的。
最后一个故事与支付相关。一篇Reddit帖子再次点燃了关于AI在医疗建议中作用的辩论。这个病毒式传播的Reddit帖子声称，ChatGPT发现了一处医生十多年来都没发现的基因突变。这个人找过多位医生，做过X光、核磁共振都没发现，但ChatGPT却找到了，后来医生也确认了这一点。你会用AI进行诊断吗？
Oh, dear. Well, man, I am not a doctor and I don't even get to play one on TV. So I'll be careful in my comments there. But listen, I get to see as many of people have in the space observed, AI doing amazing things, cancer screenings and its ability to pick up and early detection things that would be very hard for humans, even like incredibly well-trained humans. So as a tool that we should be using to deploy to help an industry of doctors and nurses that are overwhelmed, overworked, don't have enough resources. Like, amen. Yeah, I agree. I am one of those people. I'll say it that has gone to AI. Like, I'll take a picture of something on my leg. Oh, that's weird. Is that a bug bite? Is that a, you know, I don't know. Exima, what's going on? And I'll take a picture.
哦，天哪。哎，我可不是医生，甚至也没在电视上演过医生。所以我会小心地发表我的意见。不过你看，很多在这个领域里的人都观察到了，人工智能在癌症筛查和早期检测方面表现出了惊人的能力，而这些事情即使是训练有素的人类也很难做到。因此，作为一种工具，我们应该利用它来帮助那些被压得喘不过气的医生和护士，他们工作负担过重，资源也不足。对此我表示同意。我也是这样的人，我会依靠人工智能来帮助我做出判断。比如，我会拍一张腿上的照片，想看看这是什么，是虫咬吗？还是湿疹之类的。我也不知道是怎么回事，就拍张照片。
That's like dinner, you know, the district. How that? Yeah, you know, they say never Google things because then you get, oh, there's that dark side of the web. Like do not Google your symptoms because you'll find out weird things. I feel like AI is not as bad. So definitely go to AI first, don't go to Google. Google's a downward spiral. AI is a bit more reassuring. Those are your AI headlines for this week. I am with Ryan Gav in the CMO of Slack. And let's get right into your big news because you all have a lot going on. So what is the big announcement? What's going on? Well, you know, there's a lot that we have coming out.
这就像吃晚餐，你知道，那个区域。怎么样？是的，你知道，他们说永远不要查谷歌，因为这样你会发现，哦，网络上有阴暗面。例如，不要查找你的症状，因为你会看到奇怪的东西。我觉得人工智能没那么糟糕。所以一定要先去找人工智能，而不是谷歌。谷歌是一条下坡路，而人工智能更让人安心。这就是本周的人工智能头条。我和Slack的首席市场官Ryan Gav在一起。让我们直接进入你的大新闻，因为你们有很多事情正在发生。那么，重大发布是什么？发生了什么？嗯，你知道，我们有很多新东西要推出。
But, you know, the really cool thing, kind of building up some of the stuff we were talking about, you know, Slack is part of Salesforce. Salesforce has this incredible platform called Age of Force, which is a platform for building out digital labor, building out digital teammates that can sit across your entire organization. And, you know, agents is kind of the buzzword desire, right? Like you kind of have to take a shot every time you hear the words, some say, age. So like I get it, it's all AI agents. But there's something actually pretty cool that happens when you take agents, you take something like Age of Force and you bring it inside of Slack. And this is something that people aren't talking about as much. What most agents get to use is limited data sets.
你知道吗，有件非常酷的事情，就是在我们之前谈到的一些内容的基础上来说，Slack是Salesforce的一部分。而Salesforce有一个叫做“Age of Force”的非常强大的平台，这是一个用于构建数字劳动力的平台，可以在整个组织中打造出数字化的工作伙伴。现在“代理人”（agents）确实是个热门词汇，几乎每次听到有人提到这个词你都想喝一杯。所以，我明白这主要都是关于人工智能代理。但当你将代理人这个概念与“Age of Force”相结合并带入Slack时，确实会发生一些非常酷的事情。然而，这方面的讨论还不多。大多数代理人所使用的数据集是有限的，这一点往往没被充分谈论。
So they can maybe use the public internet data, which everyone has. Or they can use what's called structured data. So like let's say if you're a workday, you know, here's the day you were hired, here's your pay, here's your performance review, like structured data or things that might be in your CRM. But the gold for agents is an unstructured data. Conversations, this conversation, this is gold because it gives agents depth and context and reasoning and insight that you wouldn't get just from that structured data.
他们可以使用每个人都有的公共互联网数据，或者使用所谓的结构化数据。例如，如果你在某个工作日，那天你被雇佣了，你的薪水是多少，以及你的绩效评估，这些都属于结构化数据，或者可能存储在你的客户关系管理系统（CRM）中的数据。但对于代理来说，真正有价值的是非结构化数据。对话，比如这次对话，这些是宝贵的，因为它为代理提供了深度、背景、推理和洞察力，而这些是仅从结构化数据中得不到的。
So what happens when you bring agents inside of Slack, they can get to tap into the long-term memory of your company. They get to see that first Slack message you set as long as it's public and permissioned. They get to see the presentation that was sent to the customer 10 years ago. And you can use all those long-term memory for that agent to learn and reason over so that when it's there to take actions to provide insights to proactively suggest a next step for you to go do, it's all the more relevant.
当您将智能代理引入Slack时，会发生什么呢？它们可以访问您公司的长期记忆。只要消息是公开且有权限的，它们就能看到您最早发送的Slack消息。它们还可以查看10年前发给客户的演示文稿。您可以利用所有这些长期记忆，让代理进行学习和推理。这样当它们需要采取行动、提供见解或者主动建议接下来的步骤时，信息就更加相关。
I guess the big thing people would think of as privacy, like our my bosses is someone gonna know like the private messages between me and the other person if I said something bad. So the way these agents and agent first works inside of Slack is it inherits the permissions that you have. So share if you have, if you're in Slack and you've got access to a whole bunch of public channels in your company, it can see those. And then let's say you have access to private channels that you have, it can see those. But we're really thoughtful about how that agent interactions.
我想大家最关心的可能是隐私问题，比如我老板会不会知道我和其他人之间的私密信息，特别是如果我说了些不好的话。那么，这些代理在Slack内部的工作方式是继承你的权限。也就是说，如果你在Slack中有权限访问公司的一些公共频道，它们可以见到这些内容。此外，如果你有权限访问某些私人频道，它们也能见到。但我们在设计代理互动时非常谨慎，会特别考虑这些问题。
So let's say you've deployed a sales agent inside of a channel with you and a bunch of colleagues. And you wanna ask that agent a question inside that channel. So you can say, hey, can you summarize the IBM account plan? What's the latest opportunities that might exist? Now that agent is going to respond but the agent is gonna just respond to you first. So you sure are gonna be able to see that response privately. And then you can decide whether you wanna share that agent's response back into the channel.
假设你在一个与同事们共同使用的频道中部署了一个销售代理。你想在这个频道里向代理询问一个问题。你可以说：“嘿，你能总结一下IBM的账户计划吗？目前有哪些最新的机会？”随后，这个代理会进行回应，但首先只会将回应发送给你。这样，你就可以私下查看这个回应，然后决定是否将其分享回频道中给其他人查看。
What if I'm a smaller company, right? How would I use this in Slack as well? Yeah, I mean, let me take an example, EasyCator, which is a business catering company. They're looking at the use Slack, they're growing their company every single day. Part of their concerns is like, hey, how do we scale without adding just a ton of overhead as a small business like that's always a concern? And so they have a number of times when customers are requesting pricing checks or refunds or anything like that and they have to go process that and that's a human that's to go do that work.
如果我是一个小公司，该如何在Slack中使用这一工具呢？我来举个例子，比如EasyCator，这是一家商业餐饮公司。他们每天都在通过使用Slack来发展壮大自己的公司。作为一个小企业，他们总是关心的问题之一就是，如何在不增加大量管理负担的情况下实现规模扩张。在业务过程中，客户常常需要查询价格或申请退款，这些工作需要人工处理。
Well, they built an agent with Agent Force, they deployed that in Slack and that agent can then handle those inbound requests and they've given it a set of knowledge and data and they've empowered it with a set of actions that it can go take. So when those inbound requests come in, this agent can now shield it. Now what's really cool is this agent will come back with a recommendation to the human employees of EasyCator. And because it's built in Slack, it's human.
他们使用 Agent Force 构建了一个智能代理，并将其部署在 Slack 上。这个智能代理可以处理传入的请求。他们为代理配备了一套知识和数据，并赋予了它可以执行的一些操作。所以当有请求进来时，这个代理能够进行处理。有趣的是，这个代理会给 EasyCator 的人类员工提出建议。而且由于它是在 Slack 上构建的，显得更加人性化。
Part of Slack is just this delightful way to work. And so if the agent comes back with a recommendation that maybe a little bit too high or maybe out of bounds where they wanna go do, the human employee can just respond with like a thumbs down emoji. They can just use an emoji to tell the agent actually know that's a little too high once you go back and rework that. So you can interact with these agents and like in very, very human ways and very fun ways but our ways that are allowing you to scale your business in really meaningful ways.
Slack的一部分魅力在于它是一种令人愉悦的工作方式。因此，如果系统给出了一个可能有点过高或者不太符合需求的建议，人类员工就可以简单地用一个"👎"表情符号来回应。这样，他们仅用一个表情符号就能告诉系统，建议有些过高，需要重新调整。你可以用这样非常人性化和有趣的方式与这些系统互动，同时这些方式还能让你的业务以非常有意义的方式扩展。
Other, you know, they're not as small now but anthropic as you know is a company that's using HFORC and Slack for their deal desk. So how do they accelerate the deals? How do they get the deals that are coming in with human and agent collaboration? And they saw like a 60% improvement in the speed of the amount of deals that they could process when they deploy at HFORC and Slack.
其他公司，虽然现在不是那么小，但你知道，Anthropic 是一家使用 HFORC 和 Slack 进行交易处理的公司。那么，他们是如何加速交易的呢？他们如何通过人类和代理的合作来处理进来的交易？当他们部署 HFORC 和 Slack 后，发现处理交易的速度提高了大约 60%。
So whether you're a small company and midsize company or a big company, every department can benefit with this like digital labor that can be deployed right aside long side humans. Let me say, is the onboarding you're like in the platform and then do you just open it up? Is there somewhere you open up and say like onboard my AI assistant? How easy is this?
无论您是一家小公司、中型公司还是大型公司，每个部门都可以从这种可以与人类并肩工作的数字劳动力中受益。请允许我说明一下，您在平台上的入职流程是怎样的？是否只需打开某个地方并选择“启用我的AI助手”？这有多简单呢？
Yeah, it's remarkably amazing in terms of, you know, the agent force platform from Salesforce is this low code. You can go through and build out your agents. You can wire them up with your data, et cetera. And then what would happen in this case is your company's admin or your IT staff would deploy those agents down to Slack. And so when I open Slack in the morning, just like I could see you as a teammate, I can see all my other teammates, I have a tab that says agent force and I can click on that and I can see all the agents that are available to me.
是的，Salesforce 的 Agent Force 平台真是令人惊艳，它是一个低代码平台。你可以通过它来构建你的代理，并将它们与数据连接等。在这种情况下，你公司的管理员或 IT 人员会将这些代理部署到 Slack 上。这样，当我早上打开 Slack 时，我就像看到你和其他团队成员一样，可以看到一个名为 Agent Force 的标签。我可以点击这个标签来查看所有可用的代理。
And what's really cool is those agents have skills that are identified with them. Just like I would know that, oh, you know, sheer as an expert in communications and being unbelievably charming and being, you know, I mean, nominated, you know, so, you know, there, so I could see that the agent, this is a sales agent. This agent is an expert in understanding account plans, helping to summarize account opportunities and generate account briefing docs. And so I was like, I can go through and browse.
真正令人兴奋的是，这些代理具备与之对应的技能。就像我会知道，哦，你知道，某某人是沟通方面的专家，极具魅力，并且多次被提名等。所以，我能够判断这位代理是一个销售代理。他在了解账户计划、帮助总结账户机会和生成账户简报文档方面非常擅长。所以我就可以浏览这些信息。
We have engineering agents, you know, we have onboarding agents. So then I can just deploy those agents just like a teammate. And that's what really makes it kind of special in Slack or not even kind of special, makes it very special is, you know, we have this principle called don't make me think. And it's one of our design principles of Slack. It's like, it's just work. And so when we design agents that run Slack, they work just like teammates. I can at mention them, I can add them to a channel, I can ask them to take an action in natural language.
我们有工程代理，以及入职代理。我可以像对待队友一样部署这些代理。这使得Slack真正与众不同，也非常特别。我们有一个设计原则叫“不让我思考”，这也是Slack的设计原则之一，就是让事情变得简单。所以，当我们设计在Slack中运行的代理时，它们就像是队友一样工作。我可以用@提到它们，可以把它们加到频道里，可以用自然语言让它们执行某个动作。
And I don't have to learn a new pattern in practice. I don't have to knew a new way of working for just there. I can see their skills and I can deploy them just like I would talking to any other teammate in Slack. Yeah, I mean, I like that doing me think, I am that person. Like sometimes I'll just start working on something, especially with tech or AI where someone's like, try this out, you know, when you see it online, it looks so easy. And then you're like an hour in what the, what's going on? This is not as easy as they acted like it was.
我不需要在实践中学习新的模式，也不需要为此学习新的工作方式。我能够看到他们的技能，就像在Slack上与任何其他队友交流一样轻松地调动他们的才能。对，我是指，我喜欢这种感觉，我就是那样的人。有时我会直接开始做一些事情，尤其是在涉及到科技或AI的时候，有人会说，试试看，你知道的，在网上看到的时候看起来很简单。然而，当我实际花了一个小时的时候，感觉到底发生了什么，这并不像他们说的那么简单。
Yeah. And so anything where I could just wake up when it already has my to-do list, it's already drafted things and I could just tell what to do and I could even do it on a walk. Like that's my, for me, that is my ideal. I want to get into the fact that Slack did just release some new data showing there's a 233% surge in daily AI use and that daily users are 81% more satisfied at work, which is really cool.
是啊。所以任何那种我醒来时已经准备好待办事项清单、已经起草好一些东西并且我只需告诉它该做什么，而且我甚至可以在散步时完成的事情，对我来说就是理想的状态。我想谈谈一个事实，那就是Slack刚刚发布的数据显示，日常AI使用激增了233%，而且使用AI的用户在工作中满意度提升了81%，这真的很不错。
But I want to know, are we moving into that from busy work to deep work or just getting better at multitasking? Yeah, you know, it's kind of crazy. You know, if you look at all the tools that have been released over the last like 20 years, you know, all the new SaaS applications, you see them talk about them every week, you know, it's mind-boggling and it's been exponential to growth number tools.
我想知道，我们是从忙乱的工作转向了深度工作，还是仅仅在变得更擅长多任务处理？是啊，你知道，如果你看看过去约20年发布的所有工具，所有新的SaaS应用程序，每周都会听到关于它们的讨论，真是让人难以置信，工具的增长是指数级的。
But if you go and you look at, there's so much research out there about employee productivity. Like employee productivity has actually improved very little. And you know, 30 to 40% of people's time is still spent looking for stuff. Like I can't find the information I need for my job. You know, the average person in the enterprise is spending, you know, between 11 to 13 applications that they're swivel-charing from day to day.
但是，如果你去查看，会发现有大量关于员工生产力的研究。实际上，员工生产力的提升其实非常有限。而且，30%到40%的时间，人们仍然花在找东西上。就像无法找到工作所需的信息那样。企业中，普通员工每天在11到13个应用程序之间切换。
If you just think about your own, like I go from this app to this app and I get distracted, or this app, I gotta go find it in this app. And all that context-sushing is wasted time. So, you know, we still are operating in a pretty broken system of work if we're just really honest with ourselves. And even this last year with all this explosion of AI, like 70% of all EA projects failed.
如果你只考虑自己的情况，比如从这个应用切换到那个应用，容易分心，或者在这个应用中找东西。所有这些切换情境的过程都是浪费时间。所以说，如果我们诚实面对的话，我们仍然在一个相当不完善的工作系统中运作。即使在去年，尽管AI技术迅猛发展，但大约 70% 的企业人工智能项目仍旧失败了。
They failed to meet the expectations that the, you know, the teams that were deploying them were set out to. And the reason why is because we're just layering on stuff on top of a broken system. It's like adding a chaos monkey. You know, it's like, hey, you got a broken work system. Let's add in a bunch more and assume it'll make it all better. And so, to answer your question, you do really have to step back to like fix this problem.
他们未能达到部署团队所设定的期望。原因在于，我们只是在一个有缺陷的系统上不断叠加东西。这就像是引入了一个“混乱猴子”，就像是在说：“嘿，你的工作系统已经坏了，让我们再加入更多东西，并假设这会让一切变得更好。” 所以，要解决这个问题，你确实需要退一步，从根本上进行修复。
You do have to reimagine work. And one of the things I like to talk to my teams about, one of the things I talked to customers about a ton is I said, okay, whatever process you're doing, whatever company you're running, whatever line of business you're responsible for, how would it look different if you built it today than the age of AI? How would it be if you started it right now and you weren't constrained by any of the things of how you used to do it or all the things or all the code that you've written today?
你确实需要重构工作。我经常和团队以及客户讨论的一件事是，无论你在执行什么流程，无论你在经营什么公司，无论你负责哪个业务领域，如果你在当今人工智能时代重新建立这些工作，它会是什么样子？如果你现在就开始这个工作，不受过去的工作方式或现有代码的限制，它会有所不同吗？
And if you can kind of take that beginner's mindset, it really does force you to reimagine a lot of these systems and patterns and practices. Now, don't get me wrong. I don't think anyone's getting a three day work week anytime soon. Like I haven't seen even the most advanced companies be like, cool, everyone gets to go home early. We will fill the time, but we will fill the time with more productive, higher value, greater return for the business type of activities when we can successfully deploy this technology. And we see the promise of really getting out of this broken work system that has plagued us for the last two decades.
如果你能以一种初学者的心态来看待这个问题，确实会迫使你重新想象很多系统、模式和实践。不过，不要误会我的意思。我并不认为任何人很快就能实现每周工作三天。即使是最先进的公司，我也没见过它们说“太好了，大家都可以提早回家”。我们会填满工作时间，但我们会用更有生产力、更高价值、更能为企业带来高回报的活动来填满这些时间，当我们能够成功地部署这项技术时。我们看到有希望真正摆脱过去二十年来困扰我们的这个糟糕的工作系统。
Yeah, I agree with a lot of what you said. I think that it's hard when new things happen, you tend to, even from a smaller business angle, my company, what's trending, we create content daily. And even that is a ship that I need to move very slowly. I have a very small team. And still, I have a hard time just saying, we're going to scratch everything we're just going to go to here because you have people working on stuff. They have their agreements. Like there's so many layers to so go through at an even a small company level that I love the idea. And I've done this in meetings where I have said, if we didn't have our workflows, we weren't doing what we were doing, we had to start and scratch what we would do. And it does take a mindset. It's really difficult actually. It's more difficult than you think for people to think from scratch like that.
是的，我同意你说的很多内容。我觉得当新事物出现时，处理起来会很困难。即使从小企业的角度来看，我的公司“What's Trending”每天都在制作内容。即便如此，这艘船也需要我慢慢操控。我的团队很小，但我仍然很难说，我们要放弃一切转向新的方向，因为每个人都在各自的岗位上工作，他们都有自己的职责。这就意味着，即使在一个小公司层面，也有很多层次需要处理。我喜欢这个想法，我也曾在会议上说过，如果我们没有现有的工作流程，没有在做我们目前做的事情，而是需要从头开始，我们会怎么做。这确实需要一种特定的思维方式。实际上，这比你想象的要困难，因为要人们从头开始思考是非常不容易的。
I mean, I guess what tips do you have for people to do that in this new era? Because clearly you've been able to do it. You run teams here at the CMO level. Yeah, it is not easy. And it's not easy. And there's companies that will specifically, and we work with many that will specifically kind of warden off a small part of their company to go off and say, hey, figure out how you would disrupt us. The company is going to continue to run. But you go figure out how you would disrupt ourselves. And you're unencumbered by anything that's happening.
我的意思是，我想问在这个新时代里，你有什么建议给大家吗？因为显然你已经做到这一点。你在这里带领着首席营销官级别的团队。是的，这并不容易。而且确实很难。有些公司会特意这样做，很多我们合作的公司会专门分出一小部分团队，让他们去研究如何颠覆自己的公司。公司会继续运行，但这部分团队的任务就是不受公司现状的束缚，去设想如何自我颠覆。
So there's techniques out there certainly for doing that kind of at scale. In some cases, you just need to create the space for people to reimagine and encourage them that, hey, I wanted to try to reimagine this process. And it might fail. It might fail three or four times, and that's okay. But if we get it right, this could really be transformative. So you've got to create some safety for people to try something different that does feel uncertain. And that is part of the challenge with all this whole period is everyone's approaching AI with a different starting point.
当然，已经有一些技术可以大规模实现这一目标。在某些情况下，你只需要为人们创造一个重新想象的空间，并鼓励他们大胆尝试，说：“嘿，我想重新想象这个过程。” 这个过程可能会失败，可能会失败三四次，这没关系。但如果我们做对了，这可能会带来变革性的影响。因此，你需要为人们创造一个安全的环境，让他们去尝试不确定的新事物。这就是整个过程中挑战的一部分，因为每个人面对人工智能的起点都不一样。
There's some people, I bet you a lot of people who are listening to you, are AI enthusiasts, love it, leaning into every tool, excited about it, or whatever. But we have to be honest. There's a set of people out there that are nervous about it. There's a whole set of people who feel like, is it okay to use it at work? Am I going to, is this cheating? Am I, if I say I'm using AI, does it mean I'm less valuable? And maybe I'm not as important to the company and that puts me at risk. And yeah, a lot of EQ emotion that comes with this technology that we have to just be aware that everyone has a different starting point in this leaders.
有些人，我敢打赌，很多正在听你的节目的人都是人工智能的爱好者，他们喜欢人工智能，积极使用各种工具，对此充满期待。然而，我们必须诚实地面对现实：还有一部分人对此感到不安。有不少人会觉得，在工作中使用人工智能是否合适？这算不算作弊？如果我告诉别人我在使用人工智能，这是否意味着我就没有那么有价值了？也许我对公司来说不再那么重要，这会不会让我的处境变得危险。此外，这项技术带来了很多情感因素，我们需要意识到，不同的人对待这项技术的起点是不同的，尤其是领导者们需要注意这一点。
Part of our job is to create that safety for them to go and experiment and try these things and know. And my team knows, hey, if they're not using these tools, they're not meeting my expectations. So I have things where I'm like, hey, if you're bringing me a piece of content maybe or something you've produced, I want to know what the grade you've got and Gemini was or Chattu VT. Like, how did you grade it? And what was the grade come back? And like, how did you refine it based on that? Like, that's the expectation. I expect everything to be created by AI before we do anything externally with it.
我们工作的一部分是为团队创造一个安全的环境，让他们可以去尝试和实验这些新事物，并从中学习。而我的团队知道，如果他们不使用这些工具，就无法达到我的期望。所以当他们带来某个内容或他们制作的东西时，我会问他们，"你用Gemini或ChatGPT评分了吗？结果是什么？你如何根据评分进行改进？" 这是我的期望。在我们对外发布任何内容之前，我希望所有东西都是由AI创建的。
So let's get into the market inside of it because I feel like every company is marketing in new AI, how do you stand out? As a CMO, what is your strategy here to make sure people know about you versus all these others? I mean, there's so many from big companies to, as we know, like new companies popping up every day. Yeah. Well, I think there's really two tiers of it. One, it's Slack. One of the things we've done recently is we built, once now well over a dozen new, like what we call native AI capabilities that just flow into Slack. And we've integrated those into all of our pricing packaging plans. So as you purchase Slack and as you kind of get higher and higher plans, you get more and more of these native AI capabilities.
那么，让我们深入市场，因为我觉得每家公司都在利用新的人工智能进行营销，你如何才能脱颖而出呢？作为首席营销官，你在这方面的策略是什么，以确保人们能够关注到你而不是其他公司？毕竟，从大企业到我们知道的每天冒出来的新公司太多了。是的，我认为这里主要有两个层面。首先是Slack。我们最近做的一件事是开发了十多个我们称之为“原生AI能力”的新功能，这些功能直接融入了Slack中。我们还将其整合到所有的定价计划中。因此，当你购买Slack并选择更高的计划时，你会获得越来越多的这些原生AI功能。
And these things just show up and they just work. And so, so you are earlier kind of conversation. I think the expectation in the future is all these incredible tools are just going to be AI enabled. And we have taken this point of view of like AI that doesn't make you think. So we have AI search right inside of Slack. It's an enterprise search capability. I can search not only all my Slack, but I can search all my connected applications, Google, Microsoft, Gera, Confluence, GitHub, et cetera. And it just works. I don't have to learn anything new. I just go up to the search box. I say, what I'm looking for? And then I get not only an AI answer, but I get all the related documents. I get all the related conversations. It's just there.
这些东西会自动出现并且能够正常工作。关于你之前提到的讨论，我认为未来的预期是所有这些令人惊叹的工具都会具备 AI 功能。我们采取了一种不需要你思考的 AI 视角。例如，我们在 Slack 内置了 AI 搜索功能，这是一个企业搜索能力。我不仅可以搜索我所有的 Slack 聊天记录，还可以搜索所有连接的应用程序，如谷歌、微软、Jira、Confluence、GitHub等。它就是这么好用，我不需要学习任何新东西。我只需要打开搜索框，输入我想要找的东西，然后不仅能得到 AI 的回答，还能获得所有相关的文档和对话记录，这些信息就在那里。
Like I can now search my entire enterprise just from Slack. And that stays right in the flow work. You know, I'm about to go on vacation here in a little bit. I know and I do, I'm going to have hundreds of channels that I'm going to have to catch up on. And I can literally just go and hit recaps. And I can say I told the channels, full Slack with channels are most important for me. I can't recaps. And I get this beautiful like Monday, Sunday morning, coffee update of all the things that have been happening in channels. And I can just read it in my leisure's couple pages done. I've caught up on like 300 channels with AI recaps.
就像我现在可以直接在 Slack 上搜索整个公司的内容。这让我的工作流程更加顺畅。你知道，我即将去度假一小段时间。我知道，当我回来时，我将有数百个频道需要跟进。我可以直接使用摘要功能。我可以告诉系统，哪些频道对我来说最重要，我不需要每个频道的详细信息。而且，我会收到一个很棒的总结，就像周一或周日上午喝咖啡时的更新，告诉我这些频道里发生的事情。我可以在闲暇时阅读，只需要几页就完成了。通过 AI 的摘要功能，我便跟上了大约 300 个频道的进展。
And so these are not about marketing and going through and making big slash. This is about bringing AI just right into the flow of work. Let's also just talk about the idea of invisible AI and why that's the next big shift. Well, yeah, again, part of it is what's the increasing consumer expectation? I think you said it in three, five years so now, you're just going to kind of expect this to be there. We're like, we talk about AI is a very named thing right now. It's kind of like, I don't know what a good analogy is. It's back when the internet started. And we would say www.linkedin.com. Or like no one says www or so.
这段文字的意思是：这不仅仅是关于营销或大张旗鼓地进行变革，而是把人工智能融入到日常工作流程中。我们还应该讨论一下“隐形AI”的概念，以及为什么这是下一个大的转变。事实上，一部分原因在于消费者预期的不断提高。可能你说过，再过三到五年，你就会自然而然地期待它的存在。现在，我们谈论AI总是把它当做一个特定的名词。有点像，以前互联网刚兴起时，我们会说网址的“www”。现在没人再说“www”了。
So if you know, these are, these are things that are applied because it's new and we're looking to understand it. But we look at very simple things like a new feature that we haven't quite rolled out yet, but I'll give you the exclusive. But it's kind of a fun one inside of Slack. It's like I can click on an individual inside of Slack and I can see a profile summary. Now, what's cool about that is every organization has theirs or charts, here's what people do. But how often times you've asked yourself, like what does that person work on? Or what are the things they're doing? And turns out they're sharing their work in Slack.
如果你知道的话，这些都是因为新的原因而正在应用的东西，我们正在努力理解它。不过，我们也在关注一些非常简单的事物，比如一个我们还没完全推出的新功能，但我会给你一个独家透露。这是Slack中的一个有趣功能：我可以点击Slack中的某个人，然后看到他们的个人简介。有趣的是，每个组织都有他们的组织结构图，知道每个人具体的工作内容。但是我们常常会问自己，那个人到底在做什么？或者他们正在进行哪项工作？结果是，他们在Slack中分享了他们的工作。
They're having conversations in public conversations in Slack. So I can click on your name, Shira, and I can see a summary of who you are, what department you're in. And then I can see, here's the latest things Shira's working on. Here's what's on her mind. I can see an update of that. And it's like, wow, you know, super helpful because instantly I feel more connected to you. And I didn't have to go and reach out and ask a bunch of dumb questions. I can just go, hey, I see you're working on this. This is incredible. I'd love to partner with you on that. So little simple things. And again, I don't have to interpret, oh, this is AI, doing a bunch of whizz-bang wizardry. It just works. It's invisible, just there.
他们在 Slack 上进行公开对话。所以我可以点击你的名字，Shira，然后看到一份关于你的摘要，包括你所在的部门。我还能看到 Shira 最新的工作内容和她的想法更新。这真的非常有帮助，因为我立刻感觉与您更紧密地联系在一起，而且我不必去问很多无聊的问题。我可以直接说：“嘿，我看到你正在做这个，真了不起。我很想与你合作。” 简单的小事情。而且，我不用去担心这是不是由 AI 操控的，它就是这样自然而然地存在。
And we have feature after feature where we've rolled out inside of Slack that again just helps make people more productive in the flow of work without having to make them think. So I guess to wrap this up, where do you see the future of work in the next five to 10 years? Boy, that's always a fun one, tough one. Something I feel pretty certain about is that more and more of the workforce will move from doers to orchestrators. And I think that's a really cool thing. If you kind of, I don't know if I wish my job and anyone I love my job, it's fun. But I do a lot of meetings. And most of my job is spending time meeting with really thoughtful people and teams, and reviewing, and providing input, and direction, some strategy, and spending a lot of time with customers, and partners, obviously, wonderful people like yourself.
我们已经在 Slack 平台内推出了一个又一个功能，这些功能可以帮助人们在不需要投入额外思考的情况下更高效地工作。总结一下，您觉得未来五到十年的工作形式会如何发展呢？这是一个有趣但不简单的问题。我比较确定的是，越来越多的劳动力将从"执行者"转变为"协调者"。我觉得这是一件很酷的事情。虽然我很热爱我的工作，但是我工作的很大一部分是参加各种会议。我主要的工作内容是与富有想法的人和团队进行交流，审阅工作，提供意见和方向，制定一些策略，并花很多时间与客户和合作伙伴互动。当然，他们都是非常优秀的人，就像您一样。
But most of my job is orchestration, kind of at my current level. Now, what happens when orchestration can go down in the organization? And you can have more of the people today who are feeling like, hey, I'm just doing a bunch of work. I'm doing a bunch of activity. But they can, in turn, to your point earlier, orchestrate work. They can ask those agents to go take care of, hey, document that workflow. Hey, go ahead and recap that podcast. We just had it with Ryan, and pull off the five key messages so I can easily write a LinkedIn post about it.
大部分工作对我来说，目前主要是进行协调。那么，如果这种协调能够在组织中下放会怎样呢？这意味着现在感觉自己只是在进行一堆工作的员工，也可以进行协调。他们可以像你之前所说的那样，安排他人去处理一些任务，比如记录工作流程、总结我们刚与瑞安录制的播客并提取五个关键信息，以便我可以轻松地写一篇LinkedIn帖子。
Hey, I'm going to go for a walk. Go ahead and create me some creative that's going to look great for the website update, and make sure it's builds off the last three years creative. Like, you're going to move to orchestrating work as individual. And people who learn how to be great orchestrators of work are going to scale their capacity beyond what's, honestly, even imaginable today. And we're going to see that broken work system feel incredibly fluid. I love that. Orchestrator, I feel like one. And I feel like that's kind of worked against me in the past sometimes because people are like, you don't want to actually do the work. You just want to tell people what to do. But you know what? In the age of AI, it works, and I will lean into it. So appreciate that. Very empowering. And it's very specific as to what people should be looking out for in terms of skill set.
嘿，我打算去散步。你可以帮我创建一些看起来很棒的创意，用于网站更新，确保这些创意是建立在过去三年的基础上的。就像你将要成为一个独立工作的协调者一样。那些学会如何成为优秀工作协调者的人，将会把他们的能力扩展到今天甚至无法想象的程度。我们将看到那种破碎的工作系统变得异常流畅。我很喜欢这一点。协调者，我感觉自己就像其中之一。在过去，有时这似乎对我不太有利，因为人们会觉得我不想实际做工作，只是喜欢指挥别人。但你知道吗？在AI时代，这种方式行得通，我会抓住这个机会。这让我感到很有力量。这也很具体地指出人们应该关注哪些技能。
And for those who I guess are worried, right? There's so many people that are worried. They don't know what to do. What advice do you have for them? How do they lean into their orchestrator side or make sure they build the right skills for this next era? Yeah. There's probably no magic stiller blood. But I would start with by being curious. Be curious. Find time to start to play with these tools and capabilities. So many of these AI tools offer free versions that you can just go play with on the weekends. Come in with an idea. I'll tell you as someone who gets to hire people, it's always amazing when someone on your team comes in and was like, hey, I worked on this. And I tried this new thing with this new tool. And I want to show it to you. And like, that's just awesome.
对于那些担心的人，我猜你们可能会很焦虑，对吧？有太多的人感到忧虑，不知道该如何应对。你有什么建议可以给他们呢？他们应该如何培养自己的组织能力，或者确保他们在新时代中掌握正确的技能？其实，并不存在什么神奇的解决方法。但我建议大家从好奇心开始。要保持好奇心，找时间去尝试这些工具和功能。现在有很多AI工具提供免费版本，你可以在周末的时候试用一下。带着想法去尝试看看。作为一个有机会招聘员工的人，我想告诉你们，当团队中的某个人走进来并说：“嘿，我在这个工具上做了一些尝试，我想要给你看一下”，这总是让我感到非常惊喜和欣赏。
And it doesn't have to be great. It doesn't have to be something they end up using. But like that kind of curiosity, that, you know, figure out and rethinking stuff, that's the mode we need to be in in this period. Like there's going to be so much is going to change. You know, the crazy stat of like chat to you, T, got to like 500 million users in like 17 months and Netflix got to 100 million in 10 years. Like the pace of change is dizzying and none of us can keep up with it. So if you can come along and be part of the team that's curious, that's helping people to see what's possible and play with new things, like you've already got yourself like 80% of the way there. And so just stay being curious.
不一定要做到完美，也不一定非要做出能够被最终使用的东西。但是，我们需要保持那种好奇心，去探索、重新思考各种事物，因为在这个时期，一切都在快速变化。就像那些劲爆的数据，ChatGPT在大约17个月就达到了5亿用户，而Netflix用了10年才达到1亿用户。变化的速度让人眩晕，我们谁也无法完全跟上。如果你能加入一个充满好奇心的团队，帮助他人看到新的可能性，尝试新事物，那就已经成功了80%。所以，请继续保持好奇心。
Well, Ryan Gavin, thank you so much for joining us today. So appreciate you, your thoughtfulness, enthusiasm. And obviously great work you're doing over there. Thank you again. Thank you. I love that shout and hope to do it again soon. Okay. Yes, definitely. Well, we're not on Slack, but I'll see you out there. And of course, you could find Ryan on social media, LinkedIn, we'll put all the links in the show notes and the description of the show.
好的，Ryan Gavin，非常感谢你今天加入我们。我们非常欣赏你，以及你表现出的周到和热情。显然，你在那里做得很出色。再次感谢，谢谢。我喜欢这样的呼喊，希望很快能再次合作。好的，当然。虽然我们不在Slack上，但我会在其他地方见到你。当然，你可以在社交媒体上找到Ryan，比如LinkedIn。我们会把所有链接放在节目的注释和描述中。
And to big thanks again to our sponsor, Profit AICom's tech tools and services custom built for marketing and communication leaders looking to replace guests with knowing. Learn more at PRprofit.ai. And high blader, we've got the AI download collection, all the links from today's headlines and from previous episodes. So it's easy to read more about these stories or share them with your friends and colleagues. Click the high blader link in our show notes to get smarter. If you've loved this episode, subscribe, rate, and share it. It helps me keep growing the show independently. And we've got so much more coming your way. So see you next time on the AI download.
再次特别感谢我们的赞助商Profit AICom，感谢他们为希望用知识替代猜测的营销和传播领袖定制的技术工具和服务。更多信息请访问PRprofit.ai。此外，我们还为大家准备了AI下载合集，包含今天头条新闻和往期内容的所有链接，让您可以轻松阅读这些故事，或与朋友和同事分享。点击节目备注中的high blader链接，让自己更智慧。如果您喜欢本期节目，请订阅、评分并分享。这将帮助我独立壮大节目。我们还有更多精彩内容即将呈现。下次AI下载节目再见！