挑选 AI 赢家的法则 | The a16z Show

Anthropic and OpenAI are adding more revenue per month than Meta, Google, or Microsoft. Anthropic 和 OpenAI 每月新增的收入，已经超过了 Meta、谷歌或微软。 And I wouldn't be surprised if the combination of those two companies is doing [music] 200 billion of revenue run rate. 我不会感到惊讶，如果这两家公司加起来，营收运行率已经达到 [音乐] 2000 亿美元。 Between 2020 and 2024, top 1% exit started at $10 billion. 2020 年到 2024 年间，头部 1% 退出的门槛是 100 亿美元。 We updated those numbers in February this year, $20 billion. 今年 2 月，我们更新了这个数字，是 200 亿美元。 We just updated them yesterday. 我们昨天刚刚更新了数据。 It's now at $32 billion. 现在是 320 亿美元。 So, we've 10x 所以，我们已经 10 倍了 Yeah. 对。 over the space of kind of 24 months. 在大约 24 个月的时间里。 models get really good and the products that get built around them get really good, you see this take [music] off in usage happening. 模型越来越强，围绕模型构建的产品也越来越强，你就会看到使用量 [音乐] 呈爆发式增长。 Thought we had an AI bubble. 以为我们处于 AI 泡沫之中。 I feel pretty confident saying that we're not in a bubble right now. 我很有把握地说，我们现在不在泡沫里。 The one thing that could shift that would be I can't think of a time in my career where I have changed my mind about things at a faster clip, which is good, but is also humbling, right? 唯一可能改变这一判断的情形是，在我的职业生涯里，我想不起哪个时候我改变想法的速度比现在更快，这是好事，但也很令人谦卑，对吧？ Two big areas are scale and value capture. 两个大方向：规模和价值捕获。 So, scale on the scale side the world kind of changed in November as it relates to our business and I think sort of productivity in in the workforce 就规模而言，11 月前后这个世界发生了根本性变化，影响到了我们的业务，我认为也影响到了劳动力生产率 um you know, the way that we thought about much of the AI work that was happening before that was sort of you know, a sort of like nebulous promise in the enterprise, but we probably were contextualizing it around things like the cloud you know, in software companies and in productivity enhancement. 你知道，在那之前，我们对 AI 的理解，是企业界一个模糊的承诺，大概被我们框定在云计算的语境下，软件公司、生产效率提升之类。 And then on the consumer side, you know, you could think about AI companies like a consumer business, like you know, how many users they have and and what the price is and how big that can get. 在消费端，你可以把 AI 公司想象成消费者生意，看有多少用户、价格多少、最终能有多大。 And by the way, I think that's going to be much bigger than people expect to, which we could which we could talk about, but as of November, I think all of our priors shifted around what is actually going to happen in the enterprise. 顺便说一句，我认为那块会比人们预期的大得多，我们可以聊聊，但从 11 月开始，我觉得我们对企业端真正会发生什么的认知，已经全面更新了。 But just maybe to contextualize what's happened since then basically, Anthropic and OpenAI are adding more revenue per month than Meta, Google, or Microsoft. 先简单说一下从那时起发生了什么，基本上，Anthropic 和 OpenAI 每月新增的收入，已经超过了 Meta、谷歌或微软。 They are already at that scale of revenue getting added and actual diffusion of this technology into the real economy is tiny. 它们在收入增量上已经达到了这个量级，而这项技术在实体经济中的实际渗透率微乎其微。 It's like less than 5%. 不到 5%。 Yeah. 对。 Now, within coding and in tech-forward companies, yes, it's it's much more advanced. 当然，在编程领域和技术前沿公司里，渗透率要高得多。 Um but as it relates to every other function in the enterprise, um you know, full sort of utilization of the capabilities, we're nowhere right now. 但说到企业里其他职能的完整使用，我们现在还远远没到那一步。 So, if you pair that up with the fact that they're already getting bigger, you know, in terms of revenue added than the hyperscalers, and you're at less than 5% diffusion into the economy, I think the outcomes are going to be extraordinary. 所以，把他们的收入增速已经超过超大规模云厂商这一事实，和经济体渗透率不足 5% 放在一起看，我认为最终的结果将会非凡卓越。 Um so, the thing that we've started to try to look at to gauge, you know, what can possibly happen, like what's the upper bound, is enterprises are going to have to pay for this somehow. 我们已经开始试图评估，什么可能会发生，上限在哪里，企业最终总得为这个东西付钱。 Yeah. 对。 And so, if you just look at the Fortune 500 or the S&P 500, they're actually pretty close. 如果你只看财富 500 强或标普 500，它们的规模其实相当接近。 Um it's they generate like 2 trillion of profit per year at the collective. 它们合计每年大概产生 2 万亿美元的利润。 Um and I wouldn't be surprised if the combination of those two companies is doing 200 billion of revenue run rate by the end of this 我不会感到惊讶，如果这两家公司加起来，到今年底的营收运行率达到 2000 亿美元 Yeah. 对。 Not to mention people using open source, other vendors, so like you can add even more on top of that. 更别提还有用开源模型、其他厂商的人，这些都能再加上去。 So, we're already talking about like a 10% profit, you know, into the Fortune 500. 所以，我们已经在讨论相当于财富 500 强 10% 利润规模的数字了。 And so, I think the upper bound is going to be where the dollars going to come from. 因此，我认为上限取决于钱从哪里来。 And one of the implications, you know, like to buy this stuff. 为了购买这些东西所带来的影响之一。 Like and um you know, one of the implications of this is we had all these theories why open source and local were going to be really important. 我们原来有很多理论，说开源和本地部署会非常重要。 And it turns out that like cost is going to hit us in the face and make them really important sooner than we thought. 结果证明，成本压力会迎头打来，让它们比我们预想的更早变得至关重要。 scale we've updated our priors to to get, you know, really pilled on this on this outcome thing, on the on the size of the prize, um and the scale. 在规模方面，我们已经更新了认知，对这件事，对回报的规模和整体量级，变得非常笃定。 Um and you can see the early signs of it in the numbers. 你已经能从数据中看到早期迹象了。 But basically almost no diffusion into the real economy. 但基本上，渗透到实体经济的部分几乎为零。 It's going to get great for all these other functions. 它会在其他很多职能上大放异彩。 By the way, what's happened in coding, you can kind of start to see it in some other white-collar jobs. 顺便说一句，编程领域发生的事，在其他一些白领工作中也开始出现了。 So, like it's starting to happen in legal. 比如法律领域已经开始了。 Um, you know, the legal space is is you know, much smaller obviously than coding. 法律领域显然比编程小得多。 Um, but you know, when the models get really good and the products that get built around them get really good, you see this takeoff in usage happening and I think it's going to happen in a bunch of different functions in organizations and verticals uh over the next 12 months. 但你知道，当模型足够强、围绕模型构建的产品也足够强之后，使用量会呈爆发式增长，我认为未来 12 个月里这会发生在很多不同的企业职能和垂直行业。 And how much of that do you think's going to be native kind of AI applications? 你觉得其中有多少会是原生 AI 应用？ Cuz 因为 I kind of always go back to Chris Dixon's point around like the first three or four years you kind of see these skeuomorphic applications that kind of come in. 我总是会想到 Chris Dixon 说的那个观点，就是头几年你会看到那些仿形应用涌进来。 And and you know, we've we've seen that at the at the minute, you know, most people are using AI to do their existing job in a way that's more efficient, faster, you know, Um, but we're kind of starting to see some of the native applications come in with the you know, particularly around the generative AI. 我们现在看到的确实是这样，大多数人在用 AI 更高效、更快地做原有的工作，但围绕生成式 AI，原生应用已经开始出现了。 How how do you think that alters the landscape? 你觉得这会如何改变格局？ So, I think the big thing that's going to change in enterprise is we're kind of nowhere on how companies are run differently today. 我认为企业端最大的变化，是公司的运营方式，我们现在几乎还没开始。 And so, um, you know, the most cutting-edge companies I I happen to think that um, you know, what's happening with some of the layoff things that we're seeing is is kind of like trimming of previous fat. 在最前沿的公司里，我倾向于认为，我们看到的那些裁员，不过是在削减之前积累的冗余。 Like I I don't think it's actually efficiency gains. 我不认为那是真正的效率提升。 And by the way, there's some a really interesting thing that's happening inside these companies where um, most of the resource devotion at least for really good companies is actually on product and new things as opposed to like automatic automating the way they're run. 顺便说一句，这些公司内部正在发生一件很有意思的事，好公司把大多数资源都放在产品和新的方向上，而不是自动化自身的运营方式。 So, like they only have so many resources and the best ones know that the size of the prize of getting something right on the product side. 资源就那么多，最好的公司知道，在产品端做对了，回报的规模会有多大。 And by the way, the the best people at those companies, best engineers want to work on that side of things. 而且，那些公司最好的工程师，也想在产品这边工作。 Um, the size of that prize and and the best people are going to work on that. 这块回报的规模，以及最好的人才，都会涌向那里。 And so, that's kind of where most of the work is happening. 所以，大多数工作都在那个方向上。 Um, you know, the more mature companies would be the ones who probably would be better suited trying to automate the way their business is done internally, but they're the slower adopters. 成熟企业或许更适合尝试自动化内部业务流程，但它们恰恰是慢的那批。 Um there's kind of this latent opportunity that we see in our portfolio companies to get more, you know, drive efficiency gains and stuff, but it's not the best people working on it. 我们的投资组合公司里有一些潜在机会，可以提升效率，但做这件事的不是最好的人。 And you know, it's not where the incremental dollar is going to go just yet. 而且，增量资金短期内也不会往那个方向走。 Um you know, the the most cutting-edge folks inside those companies who are trying to do this that I've talked to are kind of in the documentation phase, uh which is just like turn everything into markdown files, you know, have, you know, as much context capture as you can possibly get, uh and then see, you know, where you can kind of still manage your business appropriately, not make sacrifices on customer experiences, um but drive efficiency. 最前沿的内部人士，跟我聊过的，基本都处于文档化阶段，就是把所有东西都转成 markdown 文件，尽可能多地沉淀上下文，然后看在哪里能适当管理业务，不牺牲客户体验，同时提升效率。 So, we're very, very, very early in that. 所以，我们还非常非常早。 Um I would say that the, you know, the native AI companies run themselves totally differently. 原生 AI 公司的运营方式完全不同。 Like the founders are just built different. 创始人就是不一样。 Um 嗯 [clears throat] [清嗓子] one of the things that, you know, we've observed about the previous generation of founders, like if you look at you know, SaaS companies, for example, I've written about this. 我们观察到上一代创始人的一个共同点，比如看 SaaS 公司，我写过这方面的东西。 Like we didn't realize how inefficiently they were running until you know, until until much later. 我们当时没意识到他们跑得有多低效，直到很久以后才发现。 It's like the the gap 就好像那个差距 more quickly they could grow as well. 或者跑得有多快也是。 Yeah, or how or how much more quickly they could grow. 对，或者跑得能有多快。 Um 嗯 And by the way, like, you know, it turns out that the the magnitude of their market it we're already seeing is just is so small compared to what we've seen in the models. 顺便说一句，事实证明，他们市场规模的量级，和我们在模型这边看到的相比，简直小得可怜。 Like the model companies are adding more than the entire public software universe in terms of revenue added, you know, combined. 模型公司新增的收入，比整个上市软件宇宙加起来还多。 Um and so um so yeah, they're they're not particularly tightly run, but they had great business models. 他们跑得并不特别高效，但商业模式很好。 And so they could grow and they could do well and everyone had a mandate to to buy more software and you know, head count grew and so so everything kind of worked out. 所以他们能成长，能做好，每个人都有扩充软件采购的任务，招人也在涨，一切都顺理成章地走通了。 The new companies are very lean, very aggressive, and they work all the time. 新公司非常精简，非常进取，他们随时都在工作。 Mhm. 嗯。 And so, uh you know, it's fun to see like the most cutting-edge companies 所以，看到最前沿的公司，是一件很有意思的事 when you go in, you know, all their researchers are sitting there 走进去，所有研究员都坐在那里 and they're they're whispering in, you know, to the 他们在低声说话，你知道，对着 Yeah, so they're not typing agents. 对，所以他们不是在敲键盘指令 agent。 They're not even typing. 他们甚至根本不打字。 Like they're they're are efficient they're like whispering in and they're running, you know, swarms of agents and um you know, I think that's kind of going to be the future. 他们非常高效，低声说话，运行一群 agent，我觉得这大概就是未来的样子。 It's just really early. 只是现在还非常早。 Um you know, this this I think the skew market is to be the skewmorphic phase is the, you know, I would say it's like everything that is reactive today. 我认为仿形阶段，就是一切都是被动响应的今天。 Like I think there's going to be a shift to proactive engagement both in consumer and in enterprise. 我认为，消费端和企业端都会向主动式参与转变。 Yeah. 对。 Um and, you know, we're starting to see it in some of the cutting edge 我们已经开始在一些最前沿的 [snorts] [哼笑] early stage companies that we're doing, but it's it's really, really early. 早期阶段公司那里看到迹象了，但还非常非常早。 Yeah. 对。 When when I think of our prior 10 to 12 months ago, you know, there's there's a couple of things that I think have have kind of changed. 回想一下我 10 到 12 个月前的认知，有几件事已经发生了变化。 One's been re-reinforced, which was, you know, we always thought that the largest companies were going to continue to be an order of magnitude larger than we'd seen in prior cycles. 有一件事得到了强化，就是我们一直以为最大的公司会比以往任何周期都大出一个数量级。 Yes. 没错。 And if anything that's accelerating. 如果说有什么变化，那就是这个趋势还在加速。 So, you know, you know, we've we've put out some data around the size of a top 1% exit doubling every 5 years or so. 我们发布过一些数据，说头部 1% 退出规模大约每 5 年翻一倍。 Um so, between 2020 and 2024, top 1% exit started at $10 billion. 2020 年到 2024 年，头部 1% 退出从 100 亿美元起步。 Um we updated those numbers in uh in February this year. 今年 2 月我们更新了这个数字。