今日概览
8
节目总数
7
订阅源
7h39m
总时长
Top Story
【All for Today】🎙️ 每日播客精要全景图
首条总览
Y Combinator Startup Podcast
(2)
Paul Graham On Startups, Ambition, and Great Founders
21m
2026/09/05
📝 AI
总结
So we're here in Mountain View today with Pg himself at the original offices doing your talk that you do for the YC batch. I looked it up. This is actually the 47th YC batch. You know, I've been trying to calculate that number. I wasn't sure exactly what it was. No wonder I'm not too worried about this talk. You know, I' gonna that many times. Yes, it changes a little bit. Yes, it's the 21st year also of YC. You know, the weird thing is, I make the talk from scratch every time. really, I probably say the same things over and over.But I always feel like I should I think, oh my God. you know, think a couple hours beforehand. oops, I better I'm talking to YC. I better think in advance a little bit about what to say. And so I open up a text file and start typing things. And that's how it always is. Any tidbits from today that are new from winter. No, most of the things most of starting a startup. here, this is an actually interesting point. Most of starting a startup is the same. And most of starting a startup is always the same. right, in, you know, microprocessors or AI or like in the internal.😊Bustion engine.It's always the same stuff.Good.Well, in that case, let's run through.Some questions. Okay, so you recently posted that the startups YC funds now are much more serious than they were in the supposedly good old days. Yeah, what did you mean by that.Well, what I meant by supposedly good old days is people are always saying that YC has jumped the shark.You know, and like, I remember when people started saying this, which is in about like 2008.Right, YC is always this thing that was great, But now it's not what it was, right.Because they can't say it was always terrible. They have to admit it was once great. So if they want to attack it, they have to say.It's fallen.Right, it was good, but now it's not. It's always been.It's always not good anymore.嗯。And so that's what I meant by the supposedly good old days, because the old days are what people claim used to be good.But like, in the old days.There was a lot of things like Reddit.Right, which valuable as it is.Is not exactly.Intercontinental ballistic cargo.Right, which I did office hours with recently. How about that for an idea.Like it's like an ICBM, except instead of exploding, it lands.Ambition and drop stuff off. Yeah, that's that's serious, right, You don't get much more serious than an.And intercontinental balllistic cargo,.What were some of the other serious ones?Well, curing cancer.,There's a startup in this batch. I mean, and there's all sorts of different approaches to curing cancer. And I say.Fund many of them, because it's like.Anyone working would be good. So like one way to approach.Caring cancer is vaccines.That's one end, right.The opposite end is therapies.Right, and so there's a startup in this batch that is.Basically.Doing on demand.Research.For people with cancer.You know.And what I like about it is.They're sort of inflicting on cancer or the death of 1000 cuts.Which may be the way to defeat it because, you know, it's notoriously.It's notoriously not the sort of thing where you just like.Come up with a cure for cancer. I mean, maybe you do if these vaccines work.Maybe the, maybe you do, right, But maybe, or maybe the way you fight it is the death of 1000 cuts. And they're actually really interesting because they helped another. Yeah, right.It's part of the whole YC community.In they help Sid from Gitlab.Deal with his cancer, which he has done very successfully. I mean, he approached it like a startup.Right, and these were essentially his co foundunders.In that defect facto startup. And it's funny because I remember thinking.😊After that, you know.I'm all, I always had this process going of what would be a good startup idea, right, And I remember after that saying, you know, somebody should start a startup to do for everyone what Si did for himself.And now.Not only is this startup happening.It's the people who actually did it for Sid. So it couldn't be better. I was so delighted when I realized that one was happening. Yeah, and I think his essay that I went founder mode on my cancer,.😊That's a very good way of describing what he did.Yeah, using another YC term to attack his cancer.So along that vein, I was rereading the essay you wrote in 2012 where you coined the term frighteningly ambitious.Frighteningly ambitious ideas.A lot of those ideas have actually gotten done since then.Like which ones.Well, a new Google.That was one of them, right, And it turns out, look.都。I, I sort of predicted this, right, that.The way you make a new Google is not by attacking them head on.You have to wait for things to change so much that their model.Is is obsolete. And that's when you can do it. And that's what open AI is.Right, like I realized very early on, like after using open AI like, whoa, I don't use Google anymore.I don't use searches.That what I really wanted from searches was not web pages. I wanted information. Why not just ask the AI for the info...(已截断)
The World’s Largest Electric Aircraft Just Flew
14m
2026/09/05
📝 AI
总结
🎼This is the largest electric airplane ever to fly. It's got 100 ft wingpan. It take off weight of $25000, and the electricity to get it off the ground costs just $5. Its successor, the E S 30 will fly up to 125 miles on battery alone up to 500 miles as a hybrid and recharge in about 30 minutes. It's the world's largest electric aircraft by about a factor, too. It's the first clean sheet Airr to be flown in any category, Eric or not in the US.And in the last 18 years, this is Anders Porceland. He's the CEO of Heart Airspace.7 years ago, Anders showed me this plane. It was a 3D printed model You can hold in your hand. It's surreal. Yeah, I came to YC with that 3D printed plane. That was like this, this size. And now it's, you know,100 foot wingspan hurturling down the runway doing taxi testing. So how do you go from basically a toy model to this.😊🎼I visited Heart Aspace's headquarters in L A, to find out.あい?🎼I've worked with Anders since he came to IC seven years ago.Back then, it was Anders and his's cofanic Clara.🎼And that3D printed model today is a 40 person team in L building a real airliner.How would you just describe hard airspace and two sentences, We build hybrid electric aircraft. But our mission is to reduce the cost of air travel. And we're building the technology to support them with a relatively small team They going after something that Boeing and Airbus have not been able to pull off.🎼Heart's plane swapped the jet engine for electric motors.You'd be holding a motor that weigh a few grams that you put on.On a drone and has one moving part. And then you look at the 400 kWts that's just like scaled up.😊So it's this really, really simple thing that has very few moving parts. So it means that you can produce them a lot cheaper.🎼They don't break because they don't really have combustion or.Fluid. So you get basically zero where.Let's break that down.A jet engine has thousands of parts and birds fuel hot enough to melt metal.An electric motor has essentially only one moving part.No combustion and doesn't really wear out.When you're taxing with a jet engine, the rest of it is.By definition, it's noise in this, but you have a constantsatorque profile so you could spin it at like, you know.One10th of the speed is virtually silent. if you stand like.Hundre0 feet away from it.And the power source is 8 battery packs laid across the floor of the plane.It's 8 facts that are like across the floor of the fuselage.It's actually only about four tlas for this one, carrying that.For $40 worth of fuel.🎼To understand why this is such a big deal.Let's look at what's broken about the jet engine.🎼Especially when it comes to short flights, It's as complex to build a jet engine for a 30 seater as it is for a 70 cer. So it costs the same Where's the same whether you fly 100 miles or 1000 miles.So it's pushing the industry into this direction of going larger and larger planes and longer and longer routes.They're extremely inefficient at.タくシー?You can spend, you know, for a short flight, you can spend 10% of your fuels just taxing up to their runway. Takeoff is inefficient. Landing is inefficient.So, that sort of segment.Of flying regionally is really badly served by the technology they have today.And there's already an existing market.Half of all flights in the world are under two hours. There is a replacement market here that's not even driven by cost. The aircraft that we're competing with our 40 year old designs. regional connectivity is not about how far you fly. It's about how cheap you fly. This aircraft would be absolutely best if it was flying the shortest possible route.🎼Think island hopping in Hawaii or a Norwegian F town that goes from being six hour drive to 20 minute flight.In the last year, gone from like being 33% better operating economics overall to 48% just because of this increase price in oil. So why hasn't anyone built this before.It started in Sweden.And I grew up right next to a Air Force base in Sweden. you saw like the Vigham fighter jetest.Going over this soccer field. Now it was always.I'm super excited about that. really into paper airplanes.😊🎼Just been fascinated with planes, you know, rockets and all these things.He did a PhD on jet engines and ended up at MI T.🎼That's where the idea found him.12 years ago, he comes step by tea. This is 12 years ago. Elon shows up, and he's.🎼Talking about.Electric plates, he says.🎼At one point.That, you know, eventually.Al mode of transportation will go electric.Besides rock.He's talking about like 400 wf per kilogram. And it kind of felt like a called arm,.🎼This time, I was like.Working on jet engines during the day, I was like tinkering with drones at the kitchen table at night.Becoming kind of obsessed with this.🎼I was actually doing research on this, being paid by the Swedish government to go around, speak to all the airlines before I even had a company establish the relationships with the airlines in the Nordics before I even like try to build something W...(已截断)
AI西经东译
(1)EP150:深度采访OpenAI高管Tibo-揭秘颠覆行业的产品文化密码
40m
2026/09/05
📝 AI
总结
🎼大家好,欢迎收听西京东易。如果你和我一样,厌倦了在同质化的AI二手信息中打转,渴望第一时间了解彼岸的优质内容,但又苦于语言的壁垒。那么这个播客就是为你准备的。我是主播杰瑞,一位深耕AI领域的产品经理,很荣幸能为你搭建这座跨越语言的桥梁,带你零距离追踪全球一手的顶尖博客,欢迎关注同名公众号,获取更多内容。本期播客将深入探讨人工智能领域的最新趋势和突破性进展。我们的嘉宾将分享他们在AI行业的独到见解,从技术创新到商业应用,从挑战到机遇,让我们一起来听听这场精彩的对话吧。我可以在任何我觉得合适的时机按下那个按钮。😊我不太倾向于去过度关注竞争。我真正关注的是。😊我们能做出哪些独特且出色的东西?我们的价值观是什么?😊以及我们如何最大限度的加速实现这些目标。也许再过一两年。😊现在这样的速度就会成为常态。即便不是默认标准,也相差无几了。你看。如今的计算成本下降的非常惊人,科技总有办法随着时间变得极其高效。我们非常专注于提供广泛的接入。😊并且不断优化,希望能让你从中获得最大的效用。我听说现在真的有一个实体按钮了。是的。确实有我可以拿给你看,非常非常酷。😊四步。非常感谢你今天能来。😊当然。很高兴来到这里。😊是的,我非常期待和你的对话。我想从你在谷歌工作的时期开始聊起。当时。你在deep mind团队。在chaGPT出现之前。谷歌有一个叫做。LDA chat项目。你当时发推文说。谷歌因为太过紧张而不敢发布它。还说。deep mind被限制。无法推出那些可能会颠覆谷歌自身业务的产品。我时常会想起这句话。在那个时候。当你还在参与这些项目时。你在想些什么?毕竟。那是在chaGPT真正改变世界之前很久的事了。嗯。😊那确实是一段非常激动人心的时期。deep mindd是一个极具创造力的地方。我当时主要负责的专业领域是基础设施和产品。用来加速研究。所以。当时确实有一个团队在研究语言模型和规模化问题。而且他们已经取得了一些相当不错的成果,很自然的。大家就会开始思考我们能不能把它变成一个可以聊天。😊可以用于各种场景的东西。于是。😊像LMDA chat这样的想法就自然而然的出现了。他最初是一个内部项目。后来大家也希望能把它打造成一个公开可用的工具。这大概是哪一年?差不多就在chGPT发布前一年左右吧。当然我们当时也在开发很多其他的东西,我在这里就不细说了,但那的确是一个创造力迸发的地方,只是deepmin的组织架构并不是为发布产品而设计的。从这个意义上说,open AI是一个非常非常不同的地方。在这里研究和产品部门的协作超级紧密。我们一起工作,共同设计很多东西。我们有强烈的发布产品的倾向也倾向于将成果开放给大众使用。这一点我非常喜欢,也正是这里的使命,人才和人才密。吸引我来到这里,说真的,open aI有太多闪光点了。当时你参与LMDA chat项目的时候,是否已经意识到它是一个特别的东西,或者未来会变得很特别。当时的感觉确实非常特别。就好比你第一次发现模型居然能生成连贯的有帮助的文本了。虽然说实话,最初的版本更多是搞笑,而不是真的有用,但后来它就变得越来越强大了。你说你经常思考这件事。我完全理解,我觉得在很多方面,谷歌确实是自己爱了自己的事。那么你从中学到哪些经验,并把他们带到了open AI呢?是的,我经常思考这个问题,主要是从团队文化,openI自身文化的角度来想,哪些好的部分需要保留,以及哪些事不能做。open AI的文化非常自下而上,是一种非常赋能的文化。大家可以提出各种各样的想法,然后聚在一起,非常非常快的把产品做出来。对。最新的产品想法,公司内部基本上没有什么阻力,这让人感到兴奋又有趣,而且一切都是为了给世界带来积极的影响。所以对我来说,保护好这种文化非常重要。但另一件同样重要的事就是不要把事情搞得一团糟,对吧?你不会希望产品功能杂乱无章,缺乏整体方向和连贯性。所以这种文化需要一种对简约和产品质量的自豪感来平衡。我觉得chatGPT的IOS应用可以说是市面上最好的应用。😊でいい?我们希望能保持这个水准。😊所以我们在愉悦感、性能、效率和简约性这些方面投入了大量精力。总的来说就是在这些大原则下,同时赋能每个人去尝试新事物。😊并且快速交付。如果让你给一位创业者提建议。告诉他如何培养这种文化。open AI内部。有哪些更具体的做法,或者说元素?可以给他一些启发呢。我认为首先是要有坚定的信念。😊然后找到一种方法。能让你接触到用户。并根据他们的反馈,快速迭代。此外。😊还要愿意自我颠覆这一点对创业者来说可能没那么切身。但对于像open AI这样的公司来说,却非常重要。我们总在不断的提出新的研究和想法。你必须能够判断什么时候是投资他们的正确时机。😊即便这意味着要从你的主营业务上重新分配资源。这一点至关重要,但做起来真的很难。😊不过,能够做到这一点是超级重要的。这不正是你刚才描述的。谷歌没能做到的事情吗?说句公道话。那也是一个宏大计划的一部分。但在我看来,那并不是一个合适的路径。😊在open AI或任何公司。随着公司日渐成熟。要维持那种快速交付和自我颠覆的文化,会不会变得更难?是的。😊尤其是当你已经有了一个像印钞机一样稳定赚钱的业务。而旁边又出现了一个可能很酷很有创新的新东西时。维持这种文化会不会变得特别有挑战?我们是一家非常。😊非常向前看。和人工智能的未来,他将呈现怎样的面貌?😊以及他将如何造福全人类。这些宏大的命题并不会真正在意我们接下来一两个月所建立的任何东西。😊我们在这里未来1到3个月建立的任何东西其实都无足轻重。所以说。我觉得非常重要的一点是。😊要积极拥抱变化。并且。要对这一切的走向保持清醒的认知。然后你要想清楚如何定位自己。才能抓住下一波浪潮。你知道吗?即使是在open AI内部。😊我们也是先训练出模型。然后才在探索中发现他们的能力。基准测试并不能告诉你所有事。我们必须花大量时间。😊却和模型本身互动。才能真正意识到。😊比如。我们突然发现。Oh。我们好像没想过能用这种特定的方式来利用他的能力。😊或者。哇,原来他还能做这个。😮然后你就会恍然大悟。这意味着我们对产品的思考方式需要发生转变了。举个例子。就像我们最近发布的用声音对话的新功能。跟他交谈的感觉超级棒。非常自然。而且现在他也具备了使用工具的能力。这就改变了很多事情。现在我花在和他直接对话上的时间多得多了。我另一件一直在做的事。😊就是语音输入。因为他的听写质量实在太高了。😊作为一种输入方式,比打字提示词要高效的多。所以每天早上。我就拿着手机坐在那念叨几句要为chGPT做的事情。😊然后他就直接去执行了。因为他能访问我所有的工具。这种体验在拥有真正顶尖的语音模型之前,是完全不可能的。所以你看。😊他就这样突然彻底改变了你对产品的思考方式。让我们继续聊聊新模型和新的应用方式吧。几周前。我想从你的另一条推文开始。因为你的推文总是能引发热议。你说codeex在两三个月后就会显得相当初级了。我们即将迎来又一次。重大的技术飞跃。下一代模型所需要的。将远超你笔记本电脑的范畴。我们先从应用层面聊起。随着模型变得越来越好。在哪些应用领域仍然有巨大的创新空间?太多了真的太多了。😊我刚才提到了语音。还有一点。😊如果你现在是一个codeex或者其他任何编程AI的深度用户。你或多或少已经习惯了。😊他操作上的一些不太顺畅的地方。对吧。你需要管理所谓的技能文件。用这种方式来教他一些东西。但很多人也发现。😊长时间下来,这套东西维护起来其实挺麻烦的。记忆功能算是有,但他并不总能记住所有事情。如果你用到子袋里,你就得操心子袋里的配置。构建一个小的网络结构。但在你和他互动的各个环节。😊那种沉浸感就很容易被打破。而你真正想要的。😊是一个能深度理解你。理解你的目标,了解你的日常工作。😊甚至还清楚你团队动态的伙伴。他能以最佳方式做出反应,甚至主动的在日常工作中帮助你。并且不会打破那种沉浸感。让你觉得。这就是你身边那个完美的无缝协作的伙伴。这就是我们正在努力的方向。另外,当你拥有了非常非常强大的模型之后,你还会意识到另一件事。这么一来,你的笔记本电脑本身就成了一种限制。😊你想想。一台笔记本电脑能处理的工作量。其实是为人类的节奏设计的。他的设计初衷是为了匹配你能产生的输出,你打字的速度,你思考的速度,以及你需要同时打开多少个应用程序。所有这些其实都是基于人类能力的限制。但模型没有同样的限制。比如说模型完全可以同时处理100个打开的应用程序。未来这可能都不是问题。所以在资源获取方面。很明显。未来的模型需要的算力。将远远超过你笔记本电脑所能提供的。我猜你这里指的是云端代理之类的技术吧。当我们拥有像ulttra fast。这样超高速度的技术时。我们稍后会聊到。token生成的速度。可能是现在的14倍。这时候。带宽的限制就变了。或者说。瓶颈转移了。CPU本身反而成了新的瓶颈。比如工具调用。网络工具调用。协议站里的任何一点开销。都会成为限制因素。不过话说回来,你也可以通过并行处理多个任务来弥补这一点。没错,所以你可以想象模型可以一边在探索,一边写测试。同时还在编译。甚至验证一个新的假设。😊所有这些。都可以同步进行。这样一来,瓶颈就再次转移了。因为你能够并行处理更多的事情。模型也就可以非常高效。非常快速的。进行思考。以目前的token速度,我经常会同时启动10个。15个代理并行工作。但要不断在他们之间切换。对我来说。认知负担真的挺重的,因为启动任务后,你得等上三四十分钟才能看到结果。现在有了超高速度。这个工作流会发生巨大的变化。我想我可能就没法同时管理十几个代理了。这也许是件好事。可能同时处理三四个就够了。你认为一个独立开发者的工作流未来会如何演变?我觉得。如何管理你的注意力,或者说如何让技术对你的注意力更友好,这是我们非常关心的一点。😊毕竟。我们是在为人类构建工具。我们希望打造的是。😊能最大程度赋能人类的技术。这就要求我们必须围绕用户的多任务处理能力来设计。要考虑你希望如何管理自己的注意力。也是希望马上得到反馈。还是等30分钟后得到结果更好。😊而当你拥有了超快的速度。再结合语音交互。你就会突然发现。😊这个系统能以和你相当,甚至比你更快的速度运行。这样你就能一直沉浸在心流状态里。😊不断构思,不断看到原形。你可以实时生成一些小的报告。那种及时反馈的感觉。真的非常棒。你会突然意识到。跟以前那种同时开着10个智能体,手忙脚乱的状态比起来。😊你真的完全不想再回去了。所以。我们正努力创造一种体验。他不仅非常自然。😊更重要的是。他是为你量身打造的。最终实现的目标是。你不需要去适应技术。而是技术。来主动适应你。说起来,过去几个月,大家讨论了不少关于智能体编程的技术。一开始是循环。loops模型很火。现在我听说two graphs模型。也开始流行起来了。你刚才提到一个词,我特别喜欢,叫做对注意力更友好。那么。这些不同的技术。是不是都是为了帮助独立开发者?更好的。管理自己的注意力呢。我觉得可以从两个不同维度来看待这个问题。第一个维度。是打造一个最顶级的个人AGI。或者说个人智能助理。这个助理会和你无缝协作,能主动提出重要的创新想法。他能极其高效的完成你交给他的任何任务。😊无论是技术难题,还是研究咨询这类的工作。他都能胜任。并且是为你深度定制的。这一点非常重要,因为它深深植根于对你这个独特的唯一的个体的理解。😊这是我们正在全力推进的一类问题。而另一大类问题,我称之为全流程自动化。它的核心是构建能够处理复杂流程的智能系统。这些流程过去可能需要相当高的智能水平才能完成。看起来非常复杂。举个例子。😊系统可以自动去分析生产环境的日志,然后进行性能优化。或者他可以去检查代码回归问题,并自动生成补丁。在网络安全领域,我们也能看到类似的应用。比如当扫描器发现一个漏洞。系统能自动打上补丁。把风险窗口期几乎压缩到0。整个过程几乎不需要,或者说只需要极少的人工干预。可能只是在一些高风险操作上,需要人来点击一下批准按钮。但系统本身已经高度自动化了。对你来说,其实并不需要对他进行直接的事无巨细的控制。那我想稍微换个话题,我们看到。在过去几个月里。chatGPT和codeex正在逐渐走向合并。所...(已截断)
Modern Wisdom
(1)
Couples Therapist: “The One Rule Every Relationship Must Live By” - Stan Tatkin -#1146
1h 9m
2026/09/05
📝 AI
总结
Hello, friends. I'm bringing a brand new live show across the UK and Ireland this October. It's stories, psychology, audience Q And As and whatever happens when thousands of overthinkers voluntarily leave the house and end up in the same room together. I'd love to see you there. The last tour are completely sold out. So make sure that you snag tickets now using the link in the description or heading to Chris Williamson dot Live. Dublins already sold out. Los of venues are limited. Tickets are very limited. So get yours now. Chris Williamson dot live. You're great. Your work is fantastic. You rule.😊Thank you if I was to, if I was to try and summarize your worldview on relationships, allowow me to step into your skin and wear it for a second. I would summarize it as love by design, not desire. Is that a fair like it thought, not felt as relationships that design versus desire thing. How far away am I, I think I prefer earned love, earned love on a daily basis.By knowing how to hack the brain.The neuro endocrine system, which we can talk about.By by.嗯 you know.Designing the relationship.To protect feelings.To enhance feelings to.To.嗯.😊Allow two different people under stress to maintain a collaborative and cooperative.Attitude.That is a hard get.Because that.That takes a certain amount of planning.Preicing and planning and preparing.嗯。What could go wrong between human beings.hy up.Romantic attachment relationships.Miurochemically.So hard to navigate.Well, they're hard because in the attachment world, if you.If you.Understand or buy the idea of a primary attachment relationship, or primary attachment relationship is one that is.Is.Tend to be dyadic.With human primates, we pair bond and herds.And.The romantic adult.Primary attachment relationship.Is most like the earliest one.Which is predicated on survival.So.The attachment system is actually an existential.You know, glue.That holds us together it's the I can't quit you biology, we confuse it for love.And because it goes back to the earliest.You know, survival relationship.That's why breaking up also is hard to do. That's why.It has a very long memory because it reminds us of that earliest dependency relationship.Whereas other relationships tend to not do that.So we're reliving our childhoods.In a way.We're living our relationships.How we felt with all these important originals that we were around.In this.Relationship that we call a romantic adult relationship.But it is 12 in before.That is generally how I will think of it.Is that we have the parents in our heads that we had.Around 12 and and before that time, even if our parents.And they do.Changed a great deal for the better.Or for the worst.We still draw upon that earliest.You know, because that's the root.Of our experience.Through thousands upon thousands of interactions starting from infancy.Of.嗯.So powerful.Is love.Enough.To maintain a relationship.Absolutely not.😀哈哈.はははははははははは。😊No feeling is because feelings come and go. And besides, who knows what your definition of love is, it may be very different than mine.And even if I were to describe my definition of love.In action, my expectations of love may be quite different than what I say, so there you already have a problem, but if you're basing your life on emotion, which is like a roller coaster and your life is like a roller coaster so I think the smartest and the wisest and the way that it's always worked in terms of long-term relationships between individuals of equal power adults, is that you,ha to have first a shared purpose, a reason for being in this relationship, not based on feelings and a vision of where are we headed, what's the end point, are we right or die, is it all the way till you know I'm wiping your butt you're wiping my butt in old age know where are we headed and why what's the point of us, who or what do we serve only purpose that's based on a selfish reason?Will carry the day.Because we're two.Unreliable as a species.Our species is.Notably, historically.Mooy and fickle and.Impulsive and.Unpredictable.Sometimes, opportunistic, warlike.Xenophobic.Easily influenced by groups and other people.So we're not great over the long run.In unions.Therefore, we have to a.Okay.Couch.For going through time together. And that would be going through some changes. So we need.An invention we needed to design from the ground up, which is.A made up thing, right, All relationships are made up.That would sustain us in the long run.Would account for.All of that willable room necessary as we change through time.Right, otherwise.We're going to.You know.Yeah.We're not going to make it somehow.That's issue one issue two is we have to understand the forces that push us around namely our survival instinct because it's always on and because this relationship is unlike any other on the planet because of its memory most like that earliest one, based on survival it has a long memory and we tend to be proxies for everybody who came before and we tend to, it's a very p...(已截断)
井户端会议
(1)
我喺上海捱世界,唔使挂心:和广东籍沪漂音乐剧&脱口秀演员闲聊
1h 17m
2026/09/06
📝 AI
总结
hello,大家好,我是范宇如,欢迎收听本期的警户端会议啊。这一期的警户端会议比较特殊,其实是更建议大家用视频的形式观看。可能有一些听友已经关注到了小宇宙现在已经可以有看视频的一个功能了。如果大家还没有感觉到的话,可以更新一下自己小宇宙的新的版本啊。其实这期节目是我在两个多月前录的,因为可能比较了解我工作室的举阵的听友可能会知道我原来有。一个戏剧类的博客呃,但是已经有呃两年多没有更新了。我因为我一直在想说做一些内容上的一些调整和改版啊之类的。然后其实在两个多月前,我就试着录过一期想做的那一种新的方向的一个尝试。但因为这个中间过程中也有各种各样的问题和忙于一些别的一些事儿吧。然后这个新的改版一直没有得到很好的一个推进。但是这期节目其实我觉得录的还挺好玩的。所以说今天放在警湖端这个。专辑里边跟大家见面,同时再次提醒大家可以看视频的形式啊。呃,大家可以感受一下消宇宙的视频功能啊,因为为什么推荐大家更看视频呢?是因为。呃,这一期我们是别着小蜜蜂的麦克风来录的,不是手持麦,然后录的那个过程中更像一种视频综艺聊天的那种感觉。所以说视频的呈现形式可能会更。丰富一点。呃,光听音频的话,你可能有些地方听不清,或者有些地方不知道我们在笑什么。😊呃,都是在视视频的这个画面里边。如果大家看的话,可能会感受到的更清楚,好吧。所以说这期节目就推荐大家,如果你实在眼睛解放不了的话,你可以听音频。如果有这个时间和这个兴趣的话,大家可以关注一下肖宇宙的视频功能。好吧,那我们就进入正片。🎼.扳手剥壳工作。Sure.🎼世界太高端,我爱近护端。hello,欢迎大家收看收听。这来个掉链子不是。我是没什么还面哎呀。😊其实有点受线下喜剧播客的一个影响。我就发就说其实如果把音乐剧演员抽离出他戏剧的部分,嗯,让他换到他生活中别的一些有兴趣的事儿啊,或者说他的背景。成长背景一些好玩的事,其实他们还是可以聊天的嗯,那我就想好聊还是可以聊天。😊聊天聊人,然后我就尝试我就尝我就尝试嘛,我说我说是不是可以呃做一做这种这种方向嘛。所以说今天第一次。非常巧,就上海正好来了。一个戏这个戏呢其实。戏本身因为我看过他原版的,但是这次来的是一个粤语版。然后我就跟那个那个出品人就说,哎,我说可不可以让你们的演员来聊一聊粤语和广东话,然后文化嘛,就是大现在叫什么大湾区的大湾区文化。然后今天就传了这么一个局嘛。所以说呃先铺垫完了,先介绍两位音乐剧圈的嘉宾,好吧,你们要不先用广广东话自我介绍一下,我下点日语蹦出来,我举这个真回到语。😊お前がも。あめの話。😊いこ。あああ。😊哼嗯。来来来。挡住啊呀来来挡住你什么东西说话结合结合一下,结合一下结合结合结合一下一下。诶,自我介绍一下。hello 大家好,我系音乐剧演员林志铭。呢位系刘丁松,多谢你 hello hello hello hello hell啦啱好,系诶呢次咧,系我哋粤语版嘅小说诶系跟住一齐过嚟嘅,系首先啱,啱,系结束嘅。😊冇冇错准备系七月份啦左右,系翻翻广州冇错大家多多关注下啊,冇错多多关注啊,多多支持啊。はいしょう。词会讲你好没错大概大概的,我好像大概能听懂的一点点。对,就简单的能听懂。对对对对,因为7月初对7月初,然后要回到好像要7月初回到广州去演是再次回。因为这次我们是就昨天你们是啊对周日。😊上个周日在等于是刚演完在那个上海大剧院嘛,是吧,然后后面是等于是转战一下,然后要到广州去演这个这个戏,对吧?然后今天除了这两位音乐剧圈的嘉宾呢,还有我就是我觉得这个场面就是一定要。热闹一点,我特别怕音乐剧演员。包我哋有家换啫嘛,请得,你发生多顺嚟啊。我心有个明嘢啊。😀好啦好啦你俾先好冇嘈冇嘈我有我话筒他容。人家就是胸肌比这个还堡OK关于胸肌的部分待会再聊。我们聊,盯着看好久了。我超我很想Q你这方面的问题。我我觉得上次狂喜播客节的时候,我就发觉我们常值得聊一聊,有有意思意思先先介绍那个剩下的三位脱口秀圈啊剧圈漫财圈的嘉宾。😊せ太多。れ。呢个。两位乔艺跟那个穷小峰是来过我很多次井箍端的。然后我们每年我们固定要聊一次那个M one漫彩啊那些话题。然后你们你们不是有那个slogan吗?来。好。大家好,我是我先生。😊先生其出家你有死咯嘅。说一下名字。,大家好,我们是荒岛,我是曹毅,我是穷小峰,自否感觉是快来看我们的专长。现在有可以官宣的专场。呃,但是目前还没有现在可哎,能透露你们在录节目这个事儿吗?可以也可以官宣宣,现在是等于是拖有第三季有第三第三季,然后。😊火热进行中,火热进行还没被淘汰。呃,这可以了这个不能提的吧。OK好好,那如果大哥还在录之中,那大概就是因为我们这期上线应该不应该还没就是第一集还没播吧。如果第一集还没播的话,大家是能找到你们两个人。呃,可以可以说明第二集就没了,对吧?第二集可能还在,因为前两集前两集我告诉下太。对对对对对,那欢迎那个小峰跟乔野,然后第一次来我播客节目的啊,那个常叔来。哎,大大家好,我是常叔,然后啊也是个脱口秀演员啦。😊然后最近小弟也搞了个自己的播客,局外客厅啊,大家来可以关注一下。上次我们广喜播客节,我有让那个昌叔来的嘛,就是局外客厅等于是也是跟我们有个集体的一个亮相。对对对,然后上次我还传了你们两个人的那个广播话的那个粤语粤语的那个那个那一场嘛,对吧?我刚刚想我想说昌叔那天嘛,我说他们他们不是录那个播客那一场,我就在场地里边。因为我们当时是在电影院里边影厅里边录播客,观众都是做那个电影院的位置。很舒服坐着,然后台上嘉宾就坐的非常不舒服。那种大家起量舒服。然后当天话筒还不够,因为平时我们都是四只话筒然后呢,他们那天是6个人嘛,然后就是有要两个嘉宾可合用一个话筒。我记得你边上坐的是老刘吧啊,对吧?啊对对对,坐在边上做刘刘宏伟嘛,然后他跟刘宏伟两个人就合用一个话筒,刘宏伟不是长得就是像那个那个那个就是非常怎么说。😊。就是流氓啊。长得不像好人是吧??😊然后仓叔就在他坐他边上就是非常柔弱,你知道吧?然后我刚进去的时候,仓叔说着说着。说到一个很好笑的事情。然后哈哈哈这样笑,然后靠在老师身上。然后我边上一个妹子说磕到了,哇塞,我说我说昌叔怎么那么软,他说他说啊仓说么他说昌叔好软萌啊,我就我就边上一个妹子就这么说,我就当时就说哎,我说。气质上嗯适合来聊一对对,你你想让一下我想就是坐胎他。对,刚才刚才刚才我想。😊ちら。よれたでいいだ。对吧是吧你现在你现在对于那个。😊肢体上会会是会有这种对吧?是对于壮态是有向往的对吧?对对对,对对,因为我自己有在健身,我健不到这种程度,所以我还挺你有在健身的,我有在健身,但我做不到像他那样子那种。你的你的你的乳沟来看一下。这健身是不见证是吗?不是说你要健身是健指呃,而且那个那话说回来,你这个是练了多久了?😊我就亚周期的话,从我同。3岁就举铁对,我是大周期是什么叫大周期啊?就是我不懂总时长健身总时长是由16年开始啊,16年11年了,但是我是那种三天打月,两天筛黄行的。所以这十年间也不是说那种不间断的,不停歇的改练。就是可能偶尔病了,可能就歇半个月。然后对对对哎,大家可能听口音听出来,其实讲普通话的没有那么广东感因的是合肥。对对你跟大家介绍一个,其实你并不是广东籍。😊我当然不是了,我是河北籍,我是河北籍的,河北哪里?石家头啱啱才讲酒六过用个钱你打电话会北话不开差多不多他自己冇你做翻嘅。我仿讲河北他不会河北话,他没用口音,口音就是标准普东话,每次遇到个河北人先听对方什么口音,然后马上学起来的。山西也可以学学习那就很好奇了。哎,比人说一般怎么称呼你比较好,比如叫你全名好像有叫松就好叫叫丁松对吧?丁松就,就是在丁松,你怎么会身为一个河北石家庄人怎么会就是讲那么好的广东话这件事情,其实也没有那么有点捞嘞,点他知道捞叨捞唠叨唠是什么意思?捞就是语里面形容。😊的意思。就是上面说洋津包呃就是你有点哎有点外地有点土,就是外地的意思唠头对啊,就是口音没有那么正啊没有那么正。😡然后我们刚才在说什么,就是你为什么会捞。😊😀あ。因为我是河北人,但是说这种话。你说那种话不说广东话,你说说太清。😊那说广东话我堪中不老,那那你好像为什么会就讲广东话这件事?讲广东话是因为什么呢?是因为我我这个老生常谈啊啊,我我没听过说东西。对我因为我从小就很爱听广东歌,我听了十几年广东歌,然后我对这个广东歌呀,还有一些这个影视作品是有一些情节的哦,然后最早听什么?😊最早海阔天空啊,最早的话我记印象中的第一首歌应该是谢霆锋的那个。如果只得一星期。啊,我唔会啊。こや今たや先系なななななな。那那那那那那那个时候我。真的是完全听不懂,然后就来。😡就是有一种很莫名其妙的,我觉得异域风情很好异域风情。广东的肯定不是异域风情嘛。对啊,哎呀好吧,这台是,这个太恶了。😊あ。越来越捞了,捞到外国去了,是说。😊か?气谢那那就是粤语歌,然后加广东的那个香港电影。对,就是港港产片呀、粤语歌呀,就是构成了我对于这个广东文化的最早的一一部分认知。,然后再加上前。前年嘛是我们去年去年。去年去年对去年的上半年,刚好有一个澳门那边的一个。呃,音乐剧在那边做戏,然后就在大湾区嘛,然后跟着就。等到了岳小说而越小说我说越小说就粤语版小说的意思啊。对对对,因有很多黑户,今天的这个哎就是我们今天要渲我们这个7月份会回到广东啊小说小说哎哎岳小说。😊对,然后就跟着就。有越小说的这个面试,我说。那就试一下吧,因为当时。有很多我。音乐剧嘛他也是有那种一唱到底的那种。我当时面试的时候,我也有问导演,我说咱们这个戏。文本的地方多不多呀,就是台词的本。我说台词这些东西你唱不怕,你怕台词,我会对我会不在唱,因为我不会。你像那些。嘢啊乜嘢好好字旁啊嗰啲嘢啊。我。我都不认识的,我不知道什么意思哦,其实那些就反那些很奇奇怪怪的那些子台词都是用粤语搞给你看动画的一些纯粤语写啊剧本都这样哦。对,因为他要求你口语也要用那个广东式的那是的是的,那我们看的都会有点费劲,是吗?那纯粤语的文本,我们平时会说的,你要我们看着完全按照他的念其我们会有点反。他们如果打字他会打一些就是音比较相似的其他的字乱写对对。一个低,对对对好低嗯7M就就就就你比赛你比较处这个东西其实唱就还好。😊唱因为我当时。其实后来到了组里之后,才发现唱也有很多地方是需要。哈的是要需要收弛收正音正正正正正一下,我我现在眼神就是有可能是这样,然后这样。😊ち。回是交给我们交给我们正音啊,要正音要正音。对,有很多地方要正音的。然后后来但是。😊呃,我当时因为唱是唱完多年了,我觉得可能还。可能大差不差好好好好处理一些。但是后来呢。我到了组里之后,才发现就是。因为也当时面试就给我是一段台词。但是台词量不会太多,当然还是以唱为主。我说哦,那行,我试一下嘛,然后发了一段台词,全是刚才我说的那种,我说更好。就是这首台词讲的是什么?明明是中文的,这种词看不懂,我能讲的是什么。😊他说哦,他说你完全不会那个什么吗?我说。呃,我说我我我是我说不太会,我唱,他说那你怎么会唱歌,那你要不就面试先唱歌好了。我然就唱歌哎哎你唱歌还可以啊,唱的很很可以啊,然后就是。😊但是你那个。你就一点都不会读吗?我说。😡我我就不理解什么意思,但我插一个问题啊,就是身为一个就是即便喜欢广东歌或广东文化的一个人,就是下集有决心要去演一个粤语的音乐剧,其实要下很大的一个决心啊嗯。你当时是什么实年?就是火少嘛?😡彼。哦尬聊啊尬聊啊聊啊,嘛。呃,说实话说实话说实话,当时港期确实也是空着的啊,当时也是空着,但是这些都不是重点。OK下定这个决心呢,还是因为我刚才说的是有前面的这些情怀情怀,对情怀占很大一部分,然后。我说有这么一个机会,我是想试一下,但是我有担心自己能不能 handlele到,也就是能不能把这个事情很好的完成。然后最后面试过了面试过了就很不也不我的意思就是说就中间其实没有中中间中间比如说他们那个主创有没有沟通沟通过,为什么后来就也就选你了呢?我实在是实在是这这方面的演员的人才储备太少的意思吗?😊我觉得可能。各方面原因都有嗯,就是如果是我我当时听导演的这个他的原话是什么呢?就是是综合考量他说的是用了一个综合考量这个,但如当天面试也是穿着背心去的吧。即我位你千万别生讨嘢我得。😊もじ。違こやってようか。あて。我大概这你...(已截断)
Lightcone Podcast
(1)
Open Models Change The Economics of AI
N/A
2026/09/05
📝 AI
总结
🎼Cost is by far the largest pain point that open models can jump in and solve. But, you know, every business has a vision of getting better control over AI and customizing it for their business. And that's really their North star. You know, cost is something they can solve in the short term. But then it enables them to then go and customize these models for their unique use because early 2024. There was lots of interest in finding using your own custom models. Then it sort of went away and or that will just be wasted effort or get sted by the next model release.😊Seems like it's coming back now. You have a front seat to all of it. Do you think we're going through like another cycle or is it here to stay this time.Welcome back to another episode of the Lightcone. Today, we're talking to Jeffrey Morgan, co foundunder and CEO of Oama, the easiest way to run open source AI models locally and in the cloud. Oama is used by 9 million developers, has 178000 Github stars.😊And is used by 85% of the Fortune 500.Which means Jeff.Knows a lot about the state of the art of AI. What models score highest on benchmarks and what developers actually download and keep.Using.Jeff.😊Welcome to the Light cone.Thank you for having me.😊We're down to here. Like, what is the state of the art, What are you seeing out there.Well, I think the biggest thing we're seeing is a shift to open models, especially in enterprise.And that's.From a mix of US and Chinese origin models.And it's predominantly driven by coding agents.And also, AI assistance, more cowork use cases like open claw and Hermes. And because you sit in the token flow of like so many tokens.You have really good data on what models people are actually using and how it's changing. What are the trends that you're seeing.Yeah, you know, Obama started as a way to run open models on your MacBook or other hardware Nvidia, M D, Intel.And earlier this year.We launched Alma cloud.And what we're seeing there is that it's predominantly Chinese models right now of Chinese origin.But they're being accessed by businesses all over.All over the world.Especially US and Germany is actually a.Big source of where.Open model tokens are being accessed.Is it all about cost, Is that our enterprises coming because they just want to get the cost down or is there anything more.To it.Cost is by far the largest pain point that open models can jump in and solve.But, you know, every business has a vision of.Getting better control over AI and customizing it for their business.And that's really their North star. You know, costst is something they can solve in the short term.But it then enables them.To then go and customize these models for their unique use case.Is there a particular large enterprise that you can name.知ったる?Has done this.I think there was a great article in the information yesterday, from AT And T.And it ends up that they've already shifted 40% of their token consumption to open models.And that's right now, predominantly through.US and Europe models.But they're also evaluating the Chinese models.What kind of workflows do they run.Predomminantly coding agents. I think what we've seen just from the.Extreme growth, and.You know, per developer or per user token usage has predominantly been from coding agents.And then earlier in March and April, we saw open Cla takeoff and subsequently the Hermes project, Hermes Age Project take off.Which has then opened up that.Ability to automate a huge chunk of work over a long span of time to non developers, too.Whether it's like finance or support or marketing or sales.You had this actually very cool graph on the.😊Take off.Exponential for for open claw.Yeah, so earlier this year, you know, this is a graph of.Token usage.By developer on Obama's cloud.You know, the average amount of tokens they use per week.And we can this is.Per developer. So like, it looks.This graph looks like it should be an aggregate of allama growth. that this is actually the per user exactly. This is on an individual user user basis. How many tokens are they using a week.And so there's kind of like two big inflection points. One is that initial run up at the start of the year, which was driven by coding agents. So we saw Kimmy.The GLM models, mini Max launch.Finally, we had open models that could power coding agents.And then in April, we saw this increase.Inredible growth from.Open claw, really, which was.Then not just developers, but the rest of the world could take a hard problem. Give it to.An open model. and let it go complete the task, which obviously consumes a ton of tokens as it's figuring out what tools to use.What data to go fetch, We went from a context window of 1208 K to a million plus with open models. And so all that enabled this explosive growth.So went from a.Roughly 5 x. So under somewhere around, I don't know,15 million tokens.Before.Before all these co.Tyle of use cases.I think that's about ride for that open claw jump we saw in April.Obviously, in aggregate, it's.You know...(已截断)
高能量
(1)
Vol.233 媒体、社区与投资:与张鹏聊极客公园的发展史和宇树投资往事
2h 27m
2026/09/05
📝 AI
总结
🎼欢迎来到高能量,我是李翔。这次我找过来一起聊天的是我非常好的一个朋友叫张鹏。张鹏呢是极克公园的创始人,也是就媒体圈里面样非常老炮的一个人物,他跟极克公园可以说是陪伴了很多创业者的成长吧。在极克公园历史上其实有过很多后来大家津津乐道的场景。比如说我记得应该是雷军在发布小米手机之前,他其实现在杰公务员做了。一个关于手机的演讲好,现场也做了一些调研。2014年的时候,当时还没有成名的张一鸣跟还没有成为事业首富的马斯克在集克公园有一个同台的讨论。2017年就在乌镇张鹏组织了一个聚会。当时刚刚创业没多久的王星星,像雷军跟王星演示了语树的机器狗。然后其实杰克公园也是语树非常早期的一个投资人。这次请张鹏过来呢,就是说。我们其实会聊聊,就从他的角度。从移动互联网到AI,包括巨深这一轮创通行业的变迁以及。他作为一个观察者和某种程度参与者在其里面扮演的一个角色吧。那彭主要不您跟大家打声招呼,也可以补充介绍一下自己。好的,感谢李翔。大家好,我是极构公员的张鹏。客观讲,这李张老师录一个播客,我还是稍有点不适应。因为过去好像一般都是我来说这个开场的他说歌。😊介绍提这个问题,但今天可能要配合一下这杨老师。反正我们一起聊一聊吧,能够去看到从移动互联网那个繁荣的十年到中间,可能看起来有一段这个在创业投资,包括科技创新。所谓的布朗运动,然后到今天可能。大家都感觉到AI呀聚身呢带来一个新的时代。他背后有一些什么样的东西,反正公园作为一个长期在这个领域里的观察者,跟这些创业者都在水里泡着。也许能给大家分享一些有意义的东西吧,希望我能完成这个使命。对对对。其实我们开始交流的时候,我印象还蛮深刻的。张鹏会跟我讲说,他说哎,我们要不要从语数开始聊,因为他说我会担心说大家不太知道杰克物源是干什么的。然后张鹏是什么样的人。哎,我其实有点奇怪,因为在我印象里面。无论是集科公园。包括张鹏他自己的那个职业经历。应该是在。创业群里面还是蛮熟悉和蛮知道的。就是为什么会有这种打引号的不自信在里面呢?😊确实客观来讲,在疾工公园所在的这个行业里边。你看到一代一代的优秀创业者长起来,就真正的明星都是这些创业者。然后真正赚到钱呢,那投资人,当然有这些创业团队的人。像咱们这么多年的媒体人。我们在里面更多的是一些记录者、旁观者。当然有时候我们希望也能有所贡献。在这个大潮里边,我们还是。不是那个在里面的那个决定性力量素,我并没有觉得公园因为年份长,所以大家一定可以记住我们。其实今天我觉得很多年轻人甚至可能也真的未必知道寄公务员。因为咱们刚才说那些记忆都。😊多少年前都十几年前的事儿了。呃,雷军在那个地下室里边那个访谈。我觉得那时候好多同学们可能还在上小学。上初中这都是有可能的,可能都没有关注到。还有一个就是我觉得这也是今天聊天以前没聊过的。😊さき?就是其实我一直大家都认为结婚员是个媒体。我非常长的时间里特别。😊想去跟大家说,哎,金融园不是个媒体,金融园是个社区。然后后来我就放弃了,后来我就想明白了,我就说哎,我不需要让大家非得要去接受警公员是谁的概念。😊我觉得你自己最终。是谁?他会体现在你做的事儿上和结果上。所以聊到语书这个事儿是,因为作为公园的这个投资的一个成果。算是有了一个闭环吧。对。所以那个点呢我觉得他可能就能印证说公务员他不只是一个媒体,或者说其实媒体是他的某种能力。但这家公司或者这个组织。他有一些自己不同的目标,就是你真正骄傲的是那个目标,你真正想要的是那个目标。但反正摘不下媒体老师的身份,我觉得也没有问题。那我就努力的也做一个合格的媒体老师,给大家去输出一些好的内容认知。我觉得我也已经想通了。对,但你看这个其实就蛮奇怪的。你比如说A6Z就安森霍洛维茨。😊今天应该是全世界最大的风险投资机构了吧,他管理的资金已经超过了红山。然后他们就老说他们自己是个媒体公司。あ?😊缺什么就说什么。咱们坦诚讲都是这样,就是你反而觉得自己。😊有的大家都认同的东西,未必是他exactly覆盖了你想做的那东西。其实对他来讲,他的媒体能力反而是他的那个特点。那可能相对于媒体呢呃公务员的社区啊,投资啊,可能是他的特点。我觉得坦诚来讲,真的就是缺什么,就是说啥。是的,这个就是最真实的状况。是的是的,之前我也想了一下,突然跟张鹏我们认识也蛮长时间了。其实我们也经常在一起聊天。😊但是我回想了一下,我们其实没有很认真的去聊过。就比如说你认为自己不是媒体。然后这里面背后的思考啊等等之类的。比如说你是从什么时候开始就跟。创业圈,然后发生这种高频的深度的链接的。我之前也做报纸杂志嘛。我的理解就是之前其实你做IT经理世界嘛,因为IT经理世界当时其实也是。非常有名的一本杂志。那个时候接触的应该都是大公司,成熟公司是吧?当时对创业者有这种。概念吗会有兴趣吗?类似于这样的。我跟李翔老师,咱俩一直都挺聊得来,其实我觉得骨子里。都是媒体人的出身。然后我觉得对内容啊对于有意义的这种认知。我们这方面聊的很多。这次比如说聊聊公园自己他是怎么回事儿?以前咱们都是私下就聊了。😊但今天你说要认真捋捋呢。我觉得也是个挺有意思的事。既然你说从这个起点开始,就比如I迪经理世界。我估计今天可能很多听众。得有跟他年龄差不多的,大家才能有这个印象。当年有一本杂志叫IT经历,这个杂志现在还在吗?应该已经停刊了啊啊,那个杂志98年创刊的,我98年毕业的。在那个杂志里,就是我第一份工作。那个杂志他其实长在一个叫信息化的年代。就IT,所以他叫IT经理事件嘛。说个有意思,我毕业的那个通讯录上。写的叫工厂经理世界。我说我要去哪时候,他们那时候不知道什么叫IT。大家就看我写了一个东西,可能我写比较连笔。他认为啊叫工厂经理世界。😊所以IT当时还是一个挺先进的词儿,然后也是刚出现的这么一个概念。我当时就算比较幸运吧,就在新的这个领域里。那那个时候确实。你核心的要去讲IT的。那个年代就是IBM。呃,SAP.oracle国内就是联想神州数码。如果你那时候再延展一点,到那个家电领域,那个时候TCL还跟今天的小米一样如日中天呢。就那个年代其实。是一个跟今天我觉得已经完全不同的年代。😊那名字可能好多人大家都不一定能了解。我呢。😮就是本科毕业,当时的编辑部里边大部分都是硕士博士。我呢也没有什么经验,就是分配我去说你盯一个小一点的领域。叫互联网。那98年的互联网。就是在美国很热闹,国内还刚萌芽,当时有门户嘛,雅虎对对对对对对,热的公司是那个时候就国内其实啊刚萌芽。所以我那个时候就去。盯一个所于新兴的领域啊,那那个就是的起点。虽然我在这个爱丽经理世界里,确实那个时候。我们理解一些大的公司在推动信息化时代。他们是主体。但是我可能就一方面接触他们,另一方面。也比较多的去看了一些早期的也有。靠谱不靠谱的互联网创业公司。😊就那个时候可能就从这儿作为一个起点了。嗯,当时的创业公司应该。主要还是互联网相关的。还是也有其他的,我不知道比如说有硬件什么有吗?呃,那个年代其实主体就是互联网互联网啊。因为你想98年的时候,在中国。还没有风险投资的正经的概念,是非常非常早期。我记得那个时候有一些个别的机构IDG软IDG是算比较早的软银是不是一当时也有啊对,有一些这样的,但就非常早。你想98年、2000年,红杉都还没有在中国真正展开呢,他得零几年之后了,所以国内那个时候是一个呃创业的非常非常早期和混沌的状态,基本都是张朝阳啊,从海外回来,只有那一类创业者。当然那个后头有。像腾讯啊等等这样的本土的。但我觉得在早期其实那是一个创业生态非常蛮荒的阶段。所以我那时候只能看美国,就看着美国很热闹。这儿有一个公司做了一个饭模式,我可能工作了两年吧,就是另一个教授,我们一起整理了一个美国的77种互联网商业模式。这本书其实没什么书的意义,他更像是一个做了个合集,又整理了一下。那时候对你说这个我印象还挺深刻的。但时应该最有名的记者是刘任是吧?对且知识英雄。对。刘感志啊,也写雷军邱伯君这样的人。对,那早期那个软件呀、PC啊,就那个时代是中国的创业者的阶段。就是我们今天理解哈,就是说如果你是一个比如大学毕业或者工作几年之后,你自己想要做一件事儿,能给你投资,你可以去做,那个时候还是挺难的。你那个时候的创业者大概率都要之前有一些渊源,有一些契机,他才能完成这一步,他还不是一个普遍的创业的年代,对我印象中当时中国最。名的创业者应该就是王志东,可能大家已经不知道这个名字了。然后还有田素宁。对,,还有张朝阳,就对这群人我觉得当年都得是挺超人的人。就是在那个环境下,他身上有一个特别的carroer。你很难看到所谓叫平民创业者,他过去的履历没有什么特别的积累。那个时候要你的初始能量还是蛮强的。说到这儿,我就想起来,就是我看过一本书,叫中国破晓签了当。他是一个特别有名的记者,叫大卫谢夫。写的就写田树宁、王志东,他们创业的过程,以及说非常早的风险投资是怎么进入到中国的过程。但卫信夫我们这个圈子里面可能对这个名字不熟悉,但是一定看过他的作品,就是他是当时为花花公子采访乔布斯的那个记者,他是个非常有名的记者。是的是的对,但是看那个我就觉得还挺感慨的,就是非常好玩。嗯,其实那个时候我也没有把创业当成一个象限去看。嗯嗯其实主要就是看的是呃一些技术环境在变化。然后有一些新的事儿。可以做了,而且这些事可能是有意义的。其实本质上这个是底层,就当年就是因为互联网有了之后,你确实有一些新的东西就可以做了。而如果你没有互联网之间,比如说你做PC做硬件,做那种技术的投资,做交换机,做通信。那我觉得确实不是一般的人在里边你就可以把它去展开了。有了互联网之后,你发现说你可以站在软件的层面上,站在连接这件事上,连接就是一个大的红利。然后你可以在这上面。😊去做自己一些设定的方向的优化,然后去解决问题。我基本上可能前十年去看国内的这个创业,就依托于互联网年代的创业,就基本都是这一波红利还是挺明显的。对,嗯我不知道比如说你们当时在杂志的时候,你们会接触和报道风险投资嘛。就今天大家其实已经非常熟悉了吗?会在零几年之后逐渐互联网的2000年,经历过一次泡沫的破灭嘛。到0203年,其实又开始。复苏包括国内你也看到就所谓的门户时代也开始确实创造价值。虽然那个时候的价值跟今天的互联网巨头比呢还是很小很小的那种价值,但基本能够看到他的那个价值循环了。然后那个时候风投也开始逐渐出现了。从之前的IDG红杉等等一系列的这样的这种风险投资也开始出现了。呃,我们也会有关注,因为会跟他们去聊,就他们怎么看这个事。因为有的时候一个新东西出来,你不是太看得懂。所以后来我们就找一个逻辑说,那。钱的人他总得看懂点啊。对,因为我们觉得记者其实是比较敏感的看变化。但我们毕竟也没有投过钱,没有创过业,没有做公司的老板呃,甚至很多业务你都没有很了解。所以那个时候就你很容易质疑。但其实你的质疑未必是基于你看懂了的质疑。所以那时候去跟风险投资聊更多的是出钱的人你到底是什么逻辑。但后来你知道有意思的是,我从他们这也没听出,他们有特别明确的逻辑。因为好多最后那个程度。也不是那个逻辑成的这个是我当年印象非常深刻的对,有可能他压中了一个很好的项目,但是他可能也不是因为他认为的那个原因。那个项目他认为的原因到后来也不是他成功的原因。他早就可能迁移了,这个也是特别当时给我印象深刻。就我的那个基模训练里面,因为接触了这方面的数据,我就特别能接受说创业者今天说的话呢,哎,你就先听在这个阶段make sense,但可能下一个环境变化,他就得跳,不是。😊说他忽悠你,也不是说他错,他就是得变。这个是当时我印象挺深的。嗯嗯,我其实对风险投资最开始有概念。我今天回想起来就是2005年的时候,当时应该是洪山啊,KPCB啊,他们开始陆续的进入中国嘛。当时有一个杂志叫环球企业家。然后他们做了那个m克就是红山的当时全球的管理合伙人。嗯,然后也呃做了他的一个封面,就是应该是mris跟沈南鹏,还有当时洪山中国张帆,可能现在都已经退掉了。包括做了那个约翰杜尔,就KPCB当时的管理合伙人,当时全世界最好的投资者了,做了他们封面。然后这个我是有影响的。呃,另外一个就是如果我自己的话,就是应该是0708年当时开始接触了GGV就富继讯。对,当然他是从新加坡到上海去做G...(已截断)
All-In with Chamath, Jason, Sacks & Friedberg
(1)
GPT-6 Hits AGI? Tech Euphoria 2.0, SF Mansion Shortage, NYC Bans AI in Schools & Venezuela Oil Deal
1h 31m
2026/09/05
📝 AI
总结
🎼Alright, happy birthday, Chiath, we're taping on your 50th birthday. Super exciting for you, brother. Thank you. It's really great. I got a birthday letter from the big man. Oh, wow, if you would ask my parents,50 years later, you're gonna have a kid that gets a letter from the then sitting president of the United States. They would have thought absolutely not possible. Wow, look at that. Which White House intern wrote that. I wonder.😊Let your winner slide. Ra man, David.🎼We open sources to the fans, and they've just got crazy with. queen.Well, actually, speaking about interns, it was an intern. In this case, it was my son. The story was that, you know, Bradden worked at the White House all summer as an intern. And at the end of the summer, you know, he used all of the political capital he built and put in a request to get this letter from the president. and he did it. Obviously, I want to thank.Pots present sendinging me a letter about, I really want to thank my son because it was a really thoughtful, thoughtful thing that he did for me. So really appreciative of that. That actually is impressive, because.😊Yeah, it takes time out of the president's schedule to sign things so.That's why I mean something, you know.Alright, listen, breaking news, breaking news. It's a breaking news story here on the all in podcast, Open AI.Is rolling out chat GT.6.😡AK, A Astra, A S TRA folks, This is their new flagship model. It's going to be available to a limited set of organizations in the coming days.Y'all will be able to get it if you have a credit card in your chat G account, I.e., the plus pro business and enterprise users, You're gonna get it soon.Greg Brockman says he believes Open eye has entered the AGI era.And we've talked about that here before Sam Altman is.Well, he's back to being a little bit.APformative slash alarmist, or he sincerely believes.That this is the one, here's what he said at Wednesday's G20 summit.Quote.Much, much.Much,3 matches.More capable models are coming soon.The next generation of models are going to be sobering for everyone. And his final quote, I think no one intellectually honest can look at what's happening and not feel.The weight of responsibility in front of us, my Lord, a little bit of drama there added to his.As quotes, people are freaking out.About the benchmarks which are off the charts, some people are saying.😡This is the one.Your thoughts, Jaath.I don't think normal humans know how to read this chart and infer anything useful from it. The more important takeaway, I agree with Greg.I think AGI has basically been here since the beginning of the year. There are.Models that are incredibly performant.Inside of the closed frontier lab companies.And I think right now we're trying to figure out the right cadence with which to release it.I think it's very important.I also think that these capabilities will be matched or close to matched by.A handful of closed source alternatives.And.Open source alternatives.嗯。😊Probably within the next.3 or four months.So I think the right way to think about this for most people is.We are seeing.Incredible intelligence capabilities.Be broadly available.And we are seeing the cost.Of that incremental unit of intelligence.Being driven further and further down.So it's a really positive trend.Now, I think it's just the messy middle part of organizing this properly for companies and for people and building things that are useful and.Showing an actual R Y. And I woulding and tackling., I would just keep calm and carry on.Okay, Sa any ducks here on.Breaking news today.And then Fber, I'll get your reaction. But by the way, it is very exciting. And congratulations to Open AI. And I'm sure Anthropic will have something as.😊Groock, you know, Grooc still has yet to drop 4,6 is coming. That's going to be a blockbuster.The meta models are really good. Je and I just dropped something yesterday that was exceptional. So the great thing is, we have a lot of choice.😊And the even better thing is the cost of that choice is becoming cheaper and cheaper and cheaper. That's very sex. Let me tee it up you this way.Open AI seems to get its groove back now.After enthropic was running away with it.Cock and rock bot are doing things at those two other frontier models.Simply can't do so the execution, whether it's the harness or the underlying model.Or the instructions set.Or curses magic sauce.clearly, GroockBt is competitive, if not better for most users.So.Vibrant.Vibrate in competition.With leap progging occurring at a cadence, I would say of two to four weeks right now.This is what.You told us.You were going to do instead of setting up the architecture here when you took over as.AIzar.Can this continue and what's it going to be like when Jensen gets in here.With Laguna.And his access to all these open source models. I mean, this, this could be.Fate are complete. No that we win the AI race here. Look, there's a lot of innovation going on. I think American companies are competing v...(已截断)