今日概览
7
节目总数
6
订阅源
9h34m
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)
Jeff Dean: The 1% Rule for Building in AI
57m
2026/08/01
📝 AI
总结
🎼Alright, should we go, Should we get started, Jeff,, sounds great. Alright, Jack, welcome, And again, thank you so much for being here, especially just got a cold and thank you for being. I'm afraid I've lost my voice. I don't normally sound quite like this, but.😊冇冇俾 we can。So you built map produce big table, Tensorflow, the TPU Gemini. We could spend a whole hour on all the things you've done.😊But what I love is that you're still making bold predictions in public last year.Yes.Last year in May 2025, at AI Ascent, you said that AI is at the level of a junior engineer.😊九十八。A year ago.It's been.How close are we to that prediction?Yeah, I mean, I feel like the models have been getting a lot better at sort of agent based longer running coding tasks. And it seems pretty clear that they are now actually pretty capable. And depending on exactly your definition of junior engineering, It seems pretty spot on, I would say.What did you underestimate from that prediction.啱。I mean, I think the.The ability to do more and more complex tasks has been growing faster than I thought. And I also think outside of coding, these agent based systems are really starting to shine in other domains. And I think.That's, that's going to be an important trend in the future.So.Give us another bold prediction. What do you think is going be the 2027 edition.I think you will see a lot more automation of ML systems themselves, basically getting ML systems to improve their capabilities by running lots of experiments, breaking things down into subproble, you know, running those subproblem in a tight automatic experimentation loop, putting the results together, and being able to then you know, get some improved system out from that sort of fully automated problem decomposition.And and automated experimentation loop, I think that's going to be really exciting. I think that also applies not just to M, but also to other fields of science and engineering. Basically, anything where you can have a measurable objective. I think you can actually make a lot of progress these days.😊Now, let's go back to a little bit in history Back in.Way back in 2001, Google search used to run on hard drives. and you and Sanja did the math and realized that some point, the whole search index would finally fit in all of the Ram of all the computers you had running.And you made that radical realization. And you basically, in few days with Sandy shipped in production a whole new search version that worked in Ram rather than hard drive. And that was the thing that got Google to be so fast. Google searches. So history tends to remix.😊What is the, It fits the memory moment right now in 2026 that everyone in this room is still.Should be thinking about and designing.Yeah, I mean, it's a little different. But I think you're going to see more and more.啊。High performance and low energy inference hardware systems, because I think everyone is now realizing that inference is the key to making, you know, these agent based systems be available to more and more people and that latency is really important and that specialization of the hardware is a really key way you can make things that are more energy efficient and lower latency than more general purpose computational devices like.Same GPS or GPS.Because I think all everyone here is used to waiting for responses on on the models. So waiting is no fun.Master speed.So you're saying, what if we don't have to wait anymore.Yeah,, I mean, I think we'll imagine what you could do with something where the latency is.You know,50 x better.Interesting thought.Now, what's one assumption that perhaps 6000 people in this room hold that's already false about AI.Yeah, that's, that's a good question. I mean, I think.Probably one thing is people don't quite realize how possible it is to have.You know, agent based system that can run not just for an hour or two hours on a problem you care about, but.For some problem domains and with highly capable models underlying them, you can get them to run for days or weeks and do really, really complicated tasks. And I think that's.You know, starting, some people are starting to see inklings of this. But I don't think everyone has really internalized this. And that's going to be really a pretty big deal. What's the particular task that you have run that has run for weeks. What, What was it, What did the tell What did you tell the agents to solve. Yeah, I mean, I think you can tell.Agents to go off and implement.啊。You know, completely new versions of software in different programming languages that might be, you know, have better safety properties or better performance properties that. And then then they can go off and actually do that in a you know, pretty serious way. that's pretty cool.😊Now, one thing that you've been very well known for is you're really good at napkin math.Sounds funny.😊So one of the stories about you is that back in 2013, when speech recognition started to work at Google, you did the napkin mat...(已截断)
Alexandr Wang: “This is a Once-in-a-Civilization Opportunity”
32m
2026/08/01
📝 AI
总结
🎼All right, full rock star treatment for Alexander Wang, everyone. All right.So why don't we start out backstage, we're saying, you know, one of the cool ways to think about this event is like.Yeah, this room is actually full of people who are just like us. But when we were 18 or 20 or, you know, there are some 16 year olds in this audience. You know, let's jump to your story. I mean, you got it came up always really smart. like Ma Olympia. like, jump us to, you know, the Alex of that time. Like, what were you feeling, What were you thinking and what drove you down this road.Yeah, well, I grew up in New Mexico, Los Sallams, New Mexico, which now Oppenheimer famous, but it really was the middle of nowhere. and I remember I did all these math competitions, all these computer science competitions, but then I knew I wanted to do really big things. And it was like not exactly clear how or what the exact paths to do that would be. And I had a friend who.Was really into programming and you know after high school got an internship in the valley I think his first internship with that palanter and and he you know he was kind of this influence for me and so after I finished high school I ended up working at Qra here in Silicon Valley and then I worked there for a year I took a gap here to work there and then I went to MIT and this was I was 19 I worked at quora.It was 18 when I went to MIT and there was 19 when I started scale, and I remember this period from like 17 to 19.It was.I felt like I was constantly changing like, you know, exactly what I want to do was constantly changing, you know, I was learning so much just from the people around me. and it was just like, I felt like I was drinking from the fire hose pretty constantly during that time. and I would definitely recommend, you know, the two things that were really important. One is, I think working at a company was really valuable because like I think from the outside end, you have no idea how companies work. You have no idea what it looks like to actually build something. you have no idea what it looks like.To iterate on something. you have no idea what it looks like for groups of people to make decisions. And so I thought that was really important. And then going to school at MIT was actually really important because it just gave me a lot of opportunity to explore what was interesting And so it was at MIT that I started training my first models and that I like played around with TensorFlow, which had just come out that year at MIT and where I like ultimately came up with the idea of scale and then after one year of MIT, I applied to YC.You know, I felt like kind of like a miracle to get in at that time. And and whyC was was really critical to my entrepreneurial journey. Like, I don't think like YC is this amazing blend of, you know, they're very supportive and they obviously want you to succeed. but they also give it to very real. and they tell you when you're being a dumbass, which I think is you know, that's what we all need in life. So, yeah, that was I think the story till there's 19 star scale and the rest is history. I guess.😊You work with Jared Friedman at that time, and you came in with actually a very different idea than what ended up becoming scale.Yeah, so we wanted to build like an AI agent enough for Doc to help people like get medical care and it was like the it was a great example of an idea that I think will ultimately exist like I think we're even seeing it now like AI agents to help people get medical care or very real but it was the wrong timing and and we worked on it for about a month or two before Jared pulled us aside and we're like guys this is I don't know if this' is going.anywherewhere. And and that's exactly what we need to hear. And it was at that time when like, you know, I had studied AI at MIT, I like trained models and we thought we sort of went back to the drawing board, thought deeply about where the opportunity was and came up with scale. I guess selling data at the time, you know, large language models were not even had had not really come to the for yet. But selfdriving cars were sort of coming up and and computer vision suddenly became. So that was sort of the first.😊Is that right? Yeah, so the story here is that like I was when I was out my team, I did a bunch of projects like train train models of various forms. And these were like, you know, by comparison today, they're like little toy models. And and I remembered to train a model. I needed three things. I needed a GCP account. like I needed an account on some cloud service to get compute. I needed the code to run to actually train the model. and I needed data. I needed a.said and for two out of these three things, you could just press a button online and get them. And then for the last one for data, there was like no effective way to get data for training these training these models. And so it felt incredibly obvious that this was going to be the future that there ...(已截断)
乱翻书
(1)
270.大厂押注AI办公,飞书和钉钉却先成了配角
1h 37m
2026/08/01
📝 AI
总结
🎼hello hello,大家好,欢迎来到乱漫叔,看到这个新闻,然后飞叔应该可以认为飞书被拆封了吧。这也是一个超出多少人,至少在上个月说飞书要被拆封,这个点还是很让人意外的吧。其实不止飞书啊,你看隔壁钉钉其实也是相似的命运。我们今天就是打算把飞书钉钉,然后还有BAT的AI办公串在一起大概来过一遍。聊聊聊今天当下这个市场都在发生什么。,大概的主题就是聊BAT的AI办公战争,就是怎么间就突然打响了。这半年前春节的时候,大家不是还在讨论che bottle吗?不是还在讨论元宝派吗?千万发红包吗?讨论这个豆包两一日活一气绝尘吗?这怎么着突然间哎世道变了,对吧我觉得这问题谁能想到呢?就是其实从飞书你可以想到另外一个点啊,就是飞书更需要豆包还是豆包更需要飞书啊,然后就由此展开豆包。现在有这个两亿人活这个巨大这个优势,它到底是资产还是负债嘛?当然这个都是我们今天需要讨论的话题啊。那sry做个简单自我介绍吧。艾瑞克,大家好,我是erric,我之前是在飞书工作了4年,然后两年是在飞书的商业化团队,然后两年在飞书的产品团队,然后现在是一个全真奶爸,你好。😊大家好,我是屠龙之术的主播明浩,也是乱翻书比较常见的嘉宾之一了。然后一直观察行业比较多。对,然后今天正好看到新闻,我就第一时间转到了那个我们几个的群里。然后就有可能有了这场直播的由头贾亚达,大家好,我是亚楠,但是也是从业互联网十几年的老兵,之前也的确在字节跳动工作过。不过今天的这个新闻呢,主要还是以一个行业观察者的身份。因。因为本来最早也是从事媒体的,所以现在也算回归老本行。就从这个公开报道的角度,我们去看一下有哪些可以深挖的点。大家现在对于这个当红炸子机热点都很感兴趣。对你要知道在半年前当红的榨汁机还是豆包,是千问是那个元宝,以及是飞书和钉钉都在开他们的各种大会,要去做AI转型。哎,今天你看这5个全部都在今天来看哈,全部都应该怎么说是明日黄花吗?对,没错,明日黄花。这个半年前你看这这5家可能贡献了这个市场一半的那个公关稿的预算。但在今天哎。突然间都变得不重要了,都让我于AI办公了。哎,这个到底怎么回事?我们来聊一聊,我们先从这个具体的新闻开始聊吧,就是从那个飞书分拆这个事实开始聊,就是飞书这个应该可以认为是拆分,就是他的产品团队进入了豆包,他的商业化销售相关的团队进入了火山引擎。然后他的研发那一块,据说是向金框那条线去做CTO那条线去做汇报。然后我们看到一个点就是。是你看之前都是飞书跟钉钉跟企位在打。哎,但你看今天就是飞书不重要了,钉钉也不重要。钉钉呢为什么不重要,我们后面可以往后面再聊啊。但我们就是大家看到这个新闻,就是我们应该怎么来看这个事情呢?就是可以认为它是飞书再一次全面的拥抱AI,但我感觉诚意啊。因为今天飞书是被分拆,不是一个以飞书的独立主径往前走的,还是说是呃是AI拥抱飞书啊,是AI拥抱飞书就是还是说。再往前来推,就是不是自己的发现豆包C端模式走不通之后,开始重新来组织拼装手里面的这一些to B的资产。我们先先从这个点来说,豆包更需要飞书还是飞书,更需要豆包。因为白天其实乱胸发过一条朋友圈是说这个先进产品啊,某前先进产品,现在成了这个甩不掉的包袱。其实哈我个人觉得,当然这也是个人观点。我还是觉得它是。😊是豆包更需要飞书。因为其实我们从这个AI的变现模式来说,它实际上和互联网完全是两个产品。这一点上,我们用传统的互联网。是的,现在互联网已经成立一个传统行业。我们用传统互联网的思维来理解它是完全不对的。之前互联网的思维是说把规模做大,把边界做的越广越好。然后最后用边际效应去覆盖它前面的这个相当于牵一发而动全身的成本。是OK的。但是现在AI不是这样子,AI是有一个用户就多一份计算的成本。其实大家根据公开报道就可以算出来哈,做一个非常简单的算账就可以算出来。今年春节豆包投春晚的广告之前可能还稍微好一点。但是投完整个的那一套之后,用户体量起来了以后,就尤其是2月份之后,豆包几乎已经成了。不管是主流市场还是下沉市场,非常几乎它就是AI的代。名词了,这种情况下,他的用户成本,他一天大概得有百万亿级的token使用量。我们好歹算一下,就知道就是简单的大家算一下账,就应该能知道豆包现在的这个C端用户,虽然现在已经全网见不到了,但是大家都懂得都懂得。梁文峰梁胜的说法是C端用户简直是甩都甩不掉。然后他现在的情况就是百万亿级的token的消耗的话,换算成成本。大概一年也得有百亿级的人民币的消耗,而且是大几百亿。然后如果是这种情况下,他还没有变现的途径的话,实际上是非常岌岌可为的。我们都知道自节的这个之前,比如说不管是像朝夕光年呢,还是像这个大力教育啊,都是需要尽快的去证明自己是能够变现的。虽然今年梁汝波说是O inAI,但是这个欧 in他他他。其实也是没有那么高的耐心的。所以从这个角度来说,我个人认为绝对是豆包更需要飞书。它急需一个类似于A社那样能通过to B的方式把它进行变现的入口。我我个人是这样觉得。另外两位有如果有不同意见就说啊,那我补充一下吧。我觉得豆包to B更需要飞书这个入口,包括飞书的数据,然后飞书的商业化更需要豆包的AI的这么。从故事还是底层的模型。所以整体而言,我觉得两边是互补的啊,就是飞猪今天也处在一个尴尬的境地。尤其是当那个,你发现哎对面那个小1000人吧,多少人,你看差不多小几千吧,应该现在应该也小几千了,对你看小几千人,别人创造了一年那个大几百亿美金的那个营收,哎呀,就这怎么能忍呢?看到我们这个人效低的令人发纸,对吧?实际上就。😊就是因为只有to B用户在AI时代才是有价值的。to C用户就是属于梁盛是靠量化的那个钱养着的,要不然他实在是觉得留着也就留着吧。所以就从这个角度来说的话,A社的那个他从一开始就定位自己完全是从to B进行商业化的这个现在已经完全验证了,是非常成功的。海外的其他的类似的产品,不管是gockjay还是最十祖的祖师爷T都是这样的,就是他只。因为它本质上AI是一个还是个人认为哈,它是一个卖挖金子的铲子,等于它还是工具,工具还是用来提效的。所以谁有这个卖工具的需求呢,就是提高生产力的需求呢,肯定是从事生产的人,他本身就是生产力的人,才有提高生产力的需求。所以他只有B端只有A社的这个路从一开始才是对的。其他的C端的用户真的在AI时代是没有价值的。这一点,我们完全不能用传统。互联网的思维去理解它。嗯,的确我们发现索的收入已经超过了openAI但只有这两家。你像就是前面两家大模型公司,他们的收入超过了500亿美金,但是剩下第三名可能就类似于咳,只有小几十亿美金。我看那个今天那个洛一航他们那贵星人不是发了一个类似于AI没有应用这样的一个文章吗?对,就有可能就是AI没有应用,然后只有模型OK那我们还是围绕着今天这个新闻继续往下面来去做拆解啊。就是我们还聊到围绕飞数本身而言,你看。作为一个独立的业务来说,它其实有两项就是重要的权利嘛。一个就是你怎么去定义产品。另一个就是你怎么去拥有用户。但今天飞猪的产品是进入豆包体系,它的市场销售客户他进入火善引擎体系,就是说飞猪还剩下什么呢?对于飞猪的产品来说,就是这次调整,他真实的变化又是什么呢?要不我我先来说一下吧。因为我作为飞猪一个前员工。然后我今天呢还特意的跟钱同事也打了个电话。但实际上我觉得这个就是如果从内部去看,我觉得他跟就是跟新闻看到可能很多完全是相反的。我可以从商业化,然后产品,还有研发三个角度去看吧。因为基本上就是这三个团队了。啊,从商业化去看的话,首先第一,其实因为他们面对的其实是同一不管是火山引擎还是飞书的商业化,面对的是同一波客户,同一笔预算,但是呢对接的是两个不同的销售团队。这个时候呢。其实就相当于原来比如说火山引擎的销售,他只能卖火山引云,卖Cd,卖豆包的API的 token。然后就有客户说我要买豆包的企业版,然后但是豆包就其实没有一个企业版这样的一个产品。然后呢,那豆包就要把哎豆包最后什么大家看到新闻稿里面有一个叫做豆包企业版的内测功能,实际上跟这个产品根本不存在。它就是一个飞书的企业版里面加了一个机器人,那个机器人的名字叫豆包。然后呢,你们公司买了豆物包企业版,我送你飞书企业版。所以就相当于是说火山引擎的销售是这么去卖豆包企业版和飞书的啊,就是买豆包送飞书啊,因为老板更愿意为AI付钱。你跟老板说呃,我要买一个什么办公协同,办公协作,老板说不重要。我们公司变成AI native的公司。我们是AI native的团队,我们要去买豆包企业版,买豆包的token,所以就是火山引擎,他们要去卖豆包企业版的时候,他需要他需要。有一个产品去作为一个形式去去去卖给,最后都变成就是买豆包送飞书了。然后飞书的销售呢其实也是一样的,因为飞书的销售它只能卖飞书,但是比如说过去的我看到新闻比比如说过去90%的新签客户里面都买了AI的功能。然后呢,那飞书的老客户就更不用说了。所以其实对于飞书的老客户而言,他们也的增购也主要是AI相关的功能。所以对于飞书的销售,他们的主要的销售。不管是老客户增购还是新客户签约的主要的驱动,也是AI。所以那现在就变成就是站在客户的角度。我刚才说嘛,就是不管是买豆包还是买AI都是同一笔预算,同一个团队。然后这个时候我要去对接自己的两个销售团队,没有必要啊,两个销售团队还要拿两笔的那个就所谓的销售的提的提成。而且互而且互相的销销售都会互相的打架,就会你抢了我客户,我抢了你的客户,其实没有必要嘛。那所以所以当然有我看。网上说火山引擎说飞书的商业化团队并入了火山引擎,其实我就不重要。站在张一鸣的角度,就是两个销售团队合并了啊,你们效率提高了,你们不用再互相去抢客户了。然后豆包卖飞书,也完全就是没有任何的门槛了啊,其实就相当于比如说字节的巨量引擎,你要去找自节投广告,不管是投精人头条还是投西瓜视频还是投抖音,你都去买巨量引擎的广告就行了。你都充值巨量引擎就行了。然后你先。😊那要买什么AI呀买办公系统,你就找火山引擎嘛。然后这两个团队,这两笔一算在公司里面是它是完全两个不同的团队的。然后从产品团队里面的话,就是实际上完全是相反,产品团队相当于是豆包的企业版,并入了飞叔,汇报给了谢星,不是飞书的产品团队并入了豆包是相反的。因为啥呢?因为首先第一豆包的个人版叫内部叫做flowlo那么一个团队啊,包括之前的齐俊源过去的,也是也是那么一个团。然后包括我现在老板。很早也去了flow,然后呢,豆包的企业版实际上是从去年开始呢,有很多飞书的产品经理转过去的。包括我们之前那个就是北京这边的负责IM的那个产品经理,它就转去了豆包的企业版,就相当于你可以理解成,因为实际上豆包企业版这个产品不存在。它实际上就是在飞书里面加了一个A那个呃豆包的机器人。如果你要去做这个产品最主要东西是怎么样,因为底层的模型是一样的。那个模型。两边都不断那个模型其实是C的团队去做的那我要做的其实豆包企业们要做是怎么把飞书里面的各种各样的数据,把它的接口做成上下文接入到这个模型里面去,作为这个模型的一个上下文啊,那这样当然是飞书的产品团队去做更顺手啊。所以你可以让豆包企业版的产品团队,相当于原来从飞书的产品团队借调的人过去的。就从谢星手下借调的人过去,做豆包企业版。然后现在呢这帮人现在又汇报给谢星的,其实就。是就是还是同一个团队。但唯一的变化可能就是说那谢欣现在是更专注于产品。他研发。其实研发的汇报线没有任何的变化。原来飞出的研发负责人登刚,现在还是汇报给定坤,没有任何变化啊,研发没有任何变化。然后产品其实就是借调过去,然后又回来。然后谢欣现在专注于产品。然后商业化的话就相当于两边合并了。对,然后研发其实除了产品研发之外,另外就是模型研发,模型的研发跟豆包和。所以说都没有关系,他是C的团队去做的。所以你从这样去看,其实这个两个团队做这样的调整是非常非常合理的。比如站在张一鸣的角度吧。OK有没有豆包企业版,这个我们不知道,反正我看到豆包企业版,我其实打个问号,因为我只知道豆包在做付费,然后那个付费的渗透率非常费这低,就是低到这种发质的一个程度。哎,那我们就是再往下跑了嘛。因为今天他这个分拆和这个合并它暗合的一个趋势,就是今天所谓的AI。顶跟这个长...(已截断)
没理想编辑部
(1)
Vol.236 虽然没挣到钱,但起码累着了
53m
2026/08/01
📝 AI
总结
🎼 你你万镜子。だれ。🎼Yeah.Oh.🎼Oh.我是是最开始。🎼の。お?🎼お。让我摔在这这世。🎼hello,大家好,欢迎收听美理想编辑部,我是林兰,我是布理,我是乔木,我是思珍。今天是我们久违的这个姐妹唠嗑,已经很久没有坐在这儿跟大家聊聊天什么的,突然发现我普通话又标准一点,什么叫唠嗑,这个儿话音非常的标准。所以今天就是纯正姐妹闲聊啊,聊一下我们最近的这个生活,如何挨过混乱生活,不如我们先来请情。最近过。这流浪地球般的生活的吴师傅来分享一下,好久不见,所有人都是两周左右没有见过吴师傅了,其实超过两周两周多了,最近一直在一个项目拍摄,接着另外一个项目拍摄,所以是个大出差的状态。我今天再来公司的路上,跟着我的日历数了一下,我7月份总计在北京待12天,你后面还走吗?不走了,现在只剩下三天是7月,介于现在7月已经快过完了。。😊我7月份总共在北京只待了12天,总共出差出了得有个5趟吧。如果我没记错的话,看吴师傅以及其他的商务组的同事的这个出差的强度以及干的事来说,不知道以为我们赚了两个亿,其实可能赚了2万块钱吧。虽然很劳累,但是挣的那个钱跟这个劳累程度是完全不匹配的。嗯,没挣着什么钱,起码累着了。我看你回来之后,还发了一个出差。好物分享里面有18张还是20张图,就是有非常多的好物,就感觉你的这一趟已经出成了一个项目,变成一个出差博主。出差博主有点这感觉吧。就是我的箱子在最近一段时间在我屋里是不会把它掏干净并合上的,就很怪又会拿着他出门,并且这种比较需要下乡拍摄的差,他跟那种出去只是。😊因为活动或者拜访客户的差还不太一样,他不仅日夜颠倒,并且没觉睡,没饭吃下一些也没有厕所的地方,所以就会非常的艰苦啊。然后并且还是在7月份这么热的情况下,要做很多额外的准备,的确是比较混乱。出差期间,你大概平均你的睡眠可以有多久呀?在此非常敬佩各位做这个影视及影视相关行业的周边从业者们,他们都。都非常的牛。因为咱这个虽然是可怜的乙方,但是在这个实际执行中,下面还是有执行方,他们是丙方,他们肯定会比我睡的觉更少更辛苦。他们是真的纯体验劳动,我的睡的觉能比他们稍微多一点,但是时长也只能睡三四个小时而已。哦,天哪,每一天都是啊,不是不是,有的时候睡的比较少的话,可能是三四个小时。然后有的时候也能睡个五六个小时。对,但是对于他们来。说他们会睡的更少,并且他们在现场还非常的精神,我就觉得非常的敬佩。同时他们在现场会没饭吃。因为现在这个成本都紧缩着,所以要会要在一天可能拍16个小时拍很长的量,所以就没有空吃饭。,我当时在现场跟化妆师聊天。那个化妆师说他之前是画短剧和综艺,所以现在才会转行来接这种更加轻松一点的单子。没错,这个对他来说是轻松一点的单子。嗯,所以我觉得大家真的都很不容易。那他们是怎么活下来好问题,我也想又不吃又不睡,那咋活呀?所以他们是一种筛选下来的超觉人类。我觉得能做影视民工的,这个都是有两把身体素质在的。我们就大概也来接着形容一下最近生活的混乱程度吧。最近我的生活混乱程度其实不算特别特别混乱,主要就是事儿会比较多吧。我目。😊前正在一件一件的安排什么时候做什么事儿。因为每一个事儿,他的一个优先级以及它处在的那个进度条是有点不一样的。所以目前正在一个比较前期给所有东西排个表的这么一个状态。我们最近的生活其实也还行。但是我先看到混乱生活的时候,我脑子里只有一个表情包。因为最近我的好朋友经常发那个就导致我觉得他也可以用来概括我最近的生活。那就是兵来将挡,水来土掩还。莱哈姆,就以那个附属背首的呃一个系列,他乱着乱着已经习惯了,所以没有特别强的一个新的冲击。最近的一些活都属于比较长线的,就他很难在短时间内做完,所以他长线交织着,那你就习惯了就跟拉雪橇一样,你有很多不同的雪橇犬,但是大家的步伐也没有那么一致,就有快有慢,但是车总归是在往前走。我才想到还。一个可以分享的点,就是之前在过这种混乱的拍摄生活,就会过上一种和其他人完全不一样的时间作息。因为大家上班的时候,基本就是七八点起,晚上可能十一二点就睡了。当你每天需要有特别多活干,睡的特别少的时候,有可能就要早上三四五点起,那个时候你就会发现一天你是哪一个项目,所有项目都这样,因为猜的特别穷啊,所以每天拍摄都非常长啊,连拍三天,每天都这点起的。然后大同和上海拍摄还是这点。起的这公司欠你钱,你来讨债吧,就是说因为要起很早,并且睡很晚睡很晚,是因为你拿不到第二天的拍摄表,你不知道第二天几点进你就得一直等这一点不理是知道的,他不幸的跟过一次拍摄。总之呢,因为你要过上这种醒着时间非常长的生活。你会发现,一旦人早起,这一天真的非常长。就如果是什么休息日,或者是像普通的在办公室的工作日,你早起,你会觉得今天我可以做很多事。但如果不幸是在。出差早起的话,你就觉得完了,今天可以做很多事。对,首先就是如果你不幸起的太早,比如你四点起,你的一天就需要吃四顿,因为你很快就饿,那你几点睡啊,晚上也要12点睡啊,因为拿不到第二天的拍摄表啊,不仅一天要吃很多东西,因为你苏醒的时间太长,其次你还会觉得我在早上我已经感觉工作了好几个小时,快要吃午饭了,一看表9点钟,就这个时候正常。人类还没有开始上班,就是我在等待期间干完了很多活,我想把他们给派发出去了,发现大家都还没上班呢,我还不能把它给发出去。team cook的生活也不过如此,虽然team cook已经退休了。对,然后我就会觉得我和大家存在一种时差。嗯,但是我之前这个也有可能是我的误判,因为吴师傅出差两周,然后在第一周的那个出差的时候,吴师傅基本上是找不到人的,因为有的时候是会互相发送一些。垃圾后或者是一些垃圾内容嘛,那一周基本上吴师傅找不到人。但是到第二周的后期的那几天,吴师傅好像会出现,就是会对一些垃圾内容产生反应。我那个时候就想哦,他有可能在这个阶段拆的强度终于降下来一点,就是他有空去看一些小红书垃圾内容,有没有可能是适应了。呃,其实也不是适应了,,只是因为我在出差第一周的时候,我是PMPM给大家翻译,就是project method要。做所有人的妈,就是挽救计划里面的桑德拉的那个角色。就如果这个人那个男主他不干的话,你把他打晕了,你也把他送上去的那种。因为第一周我是PM,第二周,我不是,所以第二周我才可以回你们的垃圾消息,轮流当妈,今天结束立马讨债,现在是那个咸鱼的那个语气,他不是说存钱当你连续几天工作和清醒时间都特别长,你会觉得这一周过得特别慢,因为这一周其实你。到周三的时候已经干出了常规一周的工作量,但这周才过到了周三,就是虽然干了很多活,但并没有一种生命被有效延长的感觉。对你就觉得怎么还没结束。来吧,不理说一下吧。吴师傅这种的混乱,就是一种整个的大项目的混乱。我的混乱,其实就只是我控制不了我自己。因为今年也是一个工作形式上发生了变化,就是比如说现在做播客的话,写底稿。然后他的篇幅就会在一万。😊字左右。然后我以前会习惯于说就是在4000字左右篇幅的时候,就当一个短跑运动员就好了。现在就是转型需要当一个短的马拉松了,就在控制自己说把这个活呢分配到每一天,比如说写3000字,但是很难非常难。我发现我做辅到每天可以坚持写3000字,那就是跟网文的水平,所以就会堆积,然后变成了混乱。当然我也在这个混乱里面给我自己找。到了一空隙,空隙就是允许自己每一天没有达标,但是不要没有达标的太多,然后多给自己留一天的时间,就是追上来,就是做人留一线,日后好相见。然后给自己定timeline的时候也是做timeline留一线,日后好追上。那天就是留给自己拖延的对,要不然自己太苦了,确实确实也就可以顺下去聊一下大家如何在这个乱成一坨的工作与生。活分享一下各自的方法论。我就想起了之前,估计是在我们聊拖延的一期还是哪一期。但反正吴师傅有提过,他在用提醒事项的那个app,就是苹果手机都会自带的。发现我之前不会用,是因为我不需要它,所以我就会觉得这个东西左一条右一条,我也没觉得他提醒我啥了。我最近才开始频繁的用苹果自带的这个remindders,因为我的那个琐事变得多了很多。就比如说今天可能有个会,然后后天又有一个会。之前我会没有那么多的时候,我就会倾向于直接用人脑记嘛,就觉得不用再花时间给它记下来,就觉得花时间把它记下来,也是在做工。但是现在不是了,现在就是因为有的没的的会变多了之后,用人脑记其实是很容易混乱的。所以我后面就开始用提醒事项的这个app,就不至于说整个节奏都会打乱。而且我觉得用这个提醒事项,这个app有一个比较好的地方,就是因为有的工作日像我们这个内容产出类型的工作嘛,有很多的时候你不是每。😊一天都在纯输出,就是你不是每天都在录播课,不是每一天都在噼里啪啦的写稿。其实有很多的天,你是在处理一些杂事儿,你要准备资料,你找好资料之后,你就开始阅读资料,你开始做功课,像那些日子。因为没有任何的产出嘛。有时候那一天结束了之后,你会觉得自己好像什么事儿都没有干,就觉得今天有点空空的感觉。但是用了那个提醒事项的之后,我就会把所有的搜资料看资料,写提纲这些事全部都列在。上面一天过完之后呢,就会把所有的这些事项都打勾,打完勾之后,哎,就不会觉得我今天什么事都没干。就其实我今天还是看了资料,可能聊微信聊了2小时,当然这个不会写在事项里面,但是查完资料,然后又看了书,然后写提纲这些就会让自己感到更实在一点点吧,提醒事项我一直都在用唉,但是我可能也跟有一点批的属性有关。我现在提醒事项里有62条是没有打那个完成的。但是有时候看还蛮。很有意思的,就你会看到自己几年前忘了去逛优衣库这种东西。但是我发现最近有点失效。就有的时候吧,我明明提醒了,我就会1小时1小时这样子推迟,然后推迟到最后,他干脆就消失了。刚刚那个来说他会在memo里面写。呃,今天可能每一个步骤嘛,然后就打勾,我也会就有点像贪吃蛇一样的那种,就是一句话来写,因为他分很多步骤。比如说今天把提纲写好,打个勾,然后后面我会。写一个待采访,再打一个勾,后面就是待整理,然后再打一个勾,就是等等这一类的。最后你看一长句话,里面有无数个勾,你就会很爽。我最近可能有点乱不动了。就是因为我上上周去找就针灸老师,他还给我针灸了一下子。他说我想太多的话,肠胃是第二大脑它就会变很硬。然后我肠胃那个地方它的针都弯了,就是它插不进去,所以我就痛定思痛决定自己再也不想那些有的没的。的东西了。珍灸老师说,我知道你这个工作性质就是会想很多,但是你想一会儿你就到楼下散散步,不要让自己总想。我觉得这个东西吧挺有用的。所以我现在可能会是就是苹果日历利用起来,就那么一大张,然后你在上面写好自己每天或者是这个周期要做什么。你看一会儿你的灵魂就会出窍,然后就会对这一切大概有一个谱。那剩下来的日子里面按步。推进就行了,就还是那个雪橇犬的概念,就有快有慢有先有后,但是总会跑完这一场。想起我很多年前做雪橇那个犬的经历,其实你刷很多短视频也会刷到特别好笑。狗子们有的时候就突然有一个跑了出去。有的就是哎我跑还是不跑。然后那个狗子就乱了之后就会翻车,那个人就会掉在雪里面,最搞笑的是掉到雪里面,其实没什么事,你爬起来就好。但是狗子们跑起来。然后那个车它就走了,然后人类才刚刚爬。起来他就要追那个车就会很好笑,管好雪橇犬也是一门学问。是的是的,所以就是我们驾驶好自己的这个车,对,驾驶好自己车是一门学问,别让自己翻车。有时候你想太多的时候,你那个车就是会翻,你就会在自己增生的念头里面,甚至有的时候不会觉得它是一种消耗。但是你的身体会感知,然后你的胃就会变得很硬,它就会难受。我听到那个蓝和周在这里讲慎用了提醒事项合。日历我有一种非常欣慰的感觉,终于交配了这么多年的案例,终于卖出去。太超前,还是吴师傅这个管理日程的方式太超前了。我在一片混乱的时候,就意识到把这种日程规划管理,以及拆分成每一小步,让自己感觉每天都有事儿干的这种事情外置出去,不要让他占用你的脑子后台运行的。因为你这种一团混乱里面非常重要要解决的。一个事情就是那种混乱的感觉,就是你不知道自己后面要干什么,但是你好像感觉有很多事情要干的这种感觉是非常占用你的情绪的。并且你有这种情绪的时候,你就会感觉很烦,就会更难沉下心来去做一件事情。我也在这里给大家分享一...(已截断)
晚点聊 LateTalk
(1)
176: 姚顺雨,来到腾讯300天
1h 39m
2026/08/01
📝 AI
总结
🎼姚舜宇在腾讯里最大的支持者是谁?martin啊刘世平对嗯,martin可以帮他搞定什么?然后你感受到了大家对姚舜宇的评价是怎样?我觉得总体评价是他不是那种非常技术导向的人,他非常的重视组织。他经常在内部说的一些话,他其实就觉得说做大模型这个东西组织是非常关键的,很多问题都是组织的问题。🎼整体来说,腾讯对会员三这个结果满意吗?什么态度?心理预期这是一个小的模型,它能解决的问题是有限的,但核心是我让整个组织磨合好,让团队的士气重振。你觉得会员这边和微信的LOM团队长期来说,你觉得他们的走向或者怎么样?大概率是长期并存的。为什么他不用一套底层模型?比如说我在一套底层模型上,我做一些面向微信的优化,然后微信团队主要主要去做优化的那一部分,就是我知道就是martin跟小龙也聊过这个事情。小龙态度大概意思是说,他不觉得混员是第三方,嗯,但是他对于微信本身的。🎼隐私用户数据是极度在意的,所以他还是更倾向于自己去做一个模型出来,能够更可控一些。我觉得越生年那个阶段,其实腾讯可以只做一套啊,从就是技术或者说资源使用的效率上啊,但我理解这个就是腾讯对,仅展开什么意思啊。欢迎收听晚点聊,我是曼琪。今天这期节目表较特别是我和晚点的同事高洪浩来一起聊他最近刚刚报道的一个主题。姚舜宇加入腾讯之后,腾讯大模型的变化。这篇文章在7月13号已经发布在了晚点的公众号上标题叫当一个年轻人空降改造腾讯会员的300天。洪浩之前在晚点聊也录过数期节目,他是我们报道互联网领域的主要作者,尤其是常年有。报道腾讯和字节的经验。红浩,你可以和我们听友打个招呼。hello hello,大家好,那那我们直接开始吧。嗯,我我先问一下,就我自己比较好奇的问题。一个就是说你这个操作过程中间,因为我们一般写文章,其实有很多素材,你最后没有呈现在正文里。嗯,所以就具体到这个报道的话,就关于姚舜宇来了之后,整个腾讯AI的变化,有什么文章里没有呈现或者展开,但你觉得还挺有意思的事儿,你觉得可以分享一下吗?对我觉得还是有的,而且挺多的。😊比如说大家现在看到这个文章,它更多的是这个团队这个组织的偏积极和正向的这一面嘛。但是你在一个组织里面肯定会有各种各样的声音都存在的。不管是老人的一些相对来说比较负面的情绪,还是说从外部进来,腾讯这家公司的人,他的一些不适应的状态。其实这些都有,甚至包括我们也知道腾讯在内部会有另外一个团队就。微信团队也是在做大模型。那两个团队之间呃,他们有一些所谓的呃竞争也好,或者是说所谓的这种交流也好,就相关的这些内容其实都没有呃特别的展开。对不一样的声音,或者说偏不适应的声音放的没有那么多,是因为你判断整个变化是向好的吗?如果从最主流的声音。呃,最明确的这个氛围来讲的话,肯定大家的状态是跟我去。去年开始那一年跟大家聊的那个状态是完全不同的,所以整个市期肯定是明显提升啊,这个是非常明显的去提升。它相当于其实是从一个最低谷慢慢的上升。但你说上升的程度有多少,它不是一个大跨步的一个跃进,但确实是在一个上升的一个状态当中。然后你说到那个微信的大模型团队VLM和会员之间的关系。我觉得这篇文章没有展开。一个是因为这篇的主题还是混员。嗯,我觉得之后是一个什么样的竞争合作关系,这本身是一个很好的关。观察视角也是下一个,我觉得会影响腾讯AI这块变化的主要议题。呃,就从内部的角度来看的话,它其实是有一定的错位。比如说混圆他是想去呃摸高去达到就是所谓的智能的极限。但是你的微信你核心是要放到一个承载14亿用户的一个产品里面。那你的资源就我相信可能现在我不知道比如说最卡最多的硅谷的公司,他能不能把一个能。😊能力让14亿的人都用上,就他的资源,他的卡的储备,他的算力储备能不能用到?我觉得这个是微信非常大的一个考验。所以他们两个的那个所谓的定位还是有一点点的不太一样。但是你在一个大公司里面,公司的资源是有限的对,那给你多一点,给他就会少一点。所以这种平衡其实还是也是存在的。哦,然后还有一个我比较好奇的,也是跟这个文章发了之后相关的一些变化,就是你你发这篇文章的时候,K3还没有发布对。然后kimi的K3上周五7月17号发布之后,他的表现很亮,眼反馈比较好。是你有跟会员的人交流。比如说他们怎么看现在比如像K3,包括接下来也可以遇见啊,就是国内的模型竞争其实还挺激烈的,大家的水平都很高啊,我其实在呃对K3发布之后,我们也聊了一下。然后K3发布之前,其实我们也也会聊到当时的这个情况,或者其他的模型对的呃,我觉得声音也是比较不太一样的,就是有一派会觉得说。会员的节奏是比较稳稳步节奏的。其实我在文章里面也写了,就是可能有一部分的声音会认为说一直以来大模型就是个领风骚那么几个月。然后现在甚至是呃后来者居上的这种情况可能会更加的明显。比如说我后面做的模型,我可以用上更先进的架构,我可以蒸馏更先进的模型,我的数据会变得更好,就是大家会觉得说但我不是说会员哈,我。说我我想问会会员需要追牛吗?,没什么拘留。我的意思是说,大家会有这样的一个观点,或者是说有的人认为这是一个趋势,就是有后来者趋上。但同时其实也有很多人会觉得说现在都已经没什么人在做这么小的模型了,大部分都在往更高的这个尺寸去走,对吧?像自己传闻在练实体的模型,对,就是所有人都在去摸这个极致,你再做这种小的模型的意义是什么?其实我觉得。两种声音都存在啊,所以就是有一部分人还是觉得会员目前的节奏是OK的,有一部分人比较焦急。呃,对我觉得是这样子的那会员三是个多效的模型,接近300B对,290多。OK也是个MOE的模型,是是是确实比较小,这比较像现在的很多公司就发的时候会发几个版本嘛,嗯有就最大的版本,然后有比较小尺寸的。他这个比较接近大家发的小尺寸的版本。就是我理解呃一个考虑很有可能是我作为一个新人刚到。😊到这个公司里面来,然后我需要在一定的周期内有所交付。对我不可能说我来了一年了,然后什么都没有,我觉得这是一个考虑,就是我必须要交付一些什么东西出来看到呃,我的一些给老板看到的我一的一些结果吧。另外一个是我的整个团队基本上都是新的一个团队,我需要在这种真实的呃实践中去磨合每一个环节。每一个人。工作方式也好,整个流程也好。OK那我们接下来来正式聊聊姚舜于来了之后的变化。可以先从他是怎么加入腾讯开始的。其实你文章里大概写了一下这个过程,你你可以和我们的听友也简单说一下。嗯,其实就是在24年的时候呃,因为接大公司的这种高招团队,每年都会去各种各样的这种定会。那个时候他们会与非常多的这种研究员啊,然后这种可能科学家呀去建立联系。当时他们。😊就是认识到了姚舜宇,然后把他介绍给了腾讯的高层。但介绍完了以后,其实很长一段时间就没有什么特别呃多的联系了。一直到可能到了25年的时候,那应该是姚舜宇,他因为各种各样的原因吧,希望回国。那这个时候呢,就跟国内的一些这种大厂有一些联系。那那个时候他们可能就有更明确的意向说,哎,那你是不是可以来我这儿对来。承担起会员负责任的这样一个位置啊,据你所知,姚舜宇回国没有考虑过创业。呃,我说实话,我不确定他有没有考虑过创业,但是呢他肯定是跟几家这个大厂都有过联系和沟通,大公司他都聊过。是的是的,另一方面,那腾讯除了姚舜宇之外,就是他们要找一个这个新的leader,这个背景是什么呀?其实这个背景最开始是在25年的时候,25年dept出来以后,然后腾讯的高管是有发现说A。原来一个可能经验并没那么丰富,然后全是年轻人的一个组织,居然能做出一个这么厉害的模型。那时候他们其实有意识到自己的团队是不再年轻的就都是一些可能偏老的,然后背景都是更多从传统的搜广推背景出来的,其实所有大公司都这样子。你比如说自节对吧?自节最开始什?朱文家他也是做广推背景出来的对,然后最开始朱文家上面是乔木在前负责大模型嘛,他之前也是。呃搜网推的对的,那带了很多人也是搜网推的。所以那时候他们就看到了说年轻人,然后呃可能并没有那么资深的经历的人,也能做出好的东西。当时他们定下策略,就是说要要找年轻人。呃,我听到一种说法,其实是有两个选择,一个是把现在的人慢慢的去换血去换掉。一个想法是我再找一支年轻的团队单独组起来,然后跟会员去做一个这样的一个赛码。对,但我们今天看到的结果其实还是用的会员这套架子,只不过慢慢的把里面的人所汰换。所以那个时候年轻人已经成为了呃腾讯瞄准的目标。然后其实高管也。有一些反思和动作,比如说martin会自己去硅谷,跟硅谷的一些比较前沿的科学家也好,或者是meta啊,google里面的这些科学家去聊研究员去聊。然后希望他们能够来腾讯。对,然后面试的时候慢慢的以前可能呃这些研究员也好,做模型的人也好,更多的也就是跟VP这个层级,就是副总裁,也就是可能martin的减一甚至减2。二这个级别面试到头了。然后现在的话这个变化可能是martin也会直接参与这样他们的面试,就是从25年年初之后发生的变化。对的,那这个动作确实比字节要稍微晚一点晚很多。因为字节的高层最高层比较我哥可以说,就张明对不对是。对我我说一下,就我们文章里经常写的自己的高层,还真能说吗?反正就是指的张一明,就我觉得文章不写有一个我能理解的原因,是因为实际上他就他的职位并不是呃CEO也不是就说回来就反正就张明当时是见了很多人的。就更早的时候可早一年多就见了很多研员。对其实就是出来以后,他们看到的这个变化啊,其实年轻人这个风潮,我觉得应该也是整个行业对于年轻人的痴狂,应该也是从ick出来以后。😊我觉得是个逐渐的过程。啊,其实在23年就是大模型刚火的时候,我印象中很深。当时我们也写了一篇文章,嗯是方六写的标题是叫大模型创业潮狂飙180天。这里面就提到一个事儿,大概就是我们那会儿去跟一个投资人聊,然后投资人讲到一个创始人,那个创始人大概35岁左右吧,投资人就说太老了啊,35岁以上的创始人他不见之类的。是嗯。只是他是慢慢扩散的啊,我觉得可能到更多人有这个认知是在de之后嗯,除了舜宇之外,你知道腾讯接触过什么别的候选人吗?呃,他其实接触到挺多人的,就是那个时候我是他这个职位的啊,他这个职位我不知道嗯,觉得其他有很多是就不一定是负责整个团队。对对对对对对对对,嗯,那你可以去说你说他接触很多年轻人。腾讯更早的时候就已经开始去引入外部的。就他刚开始做会员的时候,大部分都是老的人做收广搜网推的那一代的呃所谓的老腾讯的技术专家也好,当时因为21年22年的时候,整个腾讯都在降本增效哦,那说到这有一个前面的背景啊,就是因为他们16年那会儿就做了AIL对不对,那AIL那些人后来是都走了吗?还是有些人也进入了会员了。对,如果要这样讲的话。其实会员最开始是有AI lab参与的,就一开始的时候AI lab所在那个组织叫做技术与工程事业群。这个是腾讯负责相当于整个公司的底层技术的一个事业群。这个是TTEGTEG对,然后AI lab其实是在里面。大模型出来以后,整个腾讯的大模型团队就是诞生在这个事业群里面的。但当时里面其实是有N多个。团队都在做自己的大模型AI lab做了其中一个,然后像他们的什么数据平台部也在做了一个,就是每个团队基本上都在做有一点类似赛马的这种情况在,然后那个时候其实AI lab最懂大模型。因为他们其实里面是有一只小的团队一直常年在研究这个NLP这个方向的。但最后他们的做出来的结果并没有好很好。然后反而是可能呃数据平台部那边做出来的,就是今天的呃所谓会。一混员二的那个底座的那个团队,他把这个事给做成了,或者是说在内部跑起来的时候,效果是最好的。相当于几个团队最后赛马赛出了现在今天的混员。然后大家就合并把所有资源都贴到会员里面。然后可能每一个部门负责一些呃模块的事情,大家相当于是一个联合战队的一个组织形态,做出了最初代的这个会员。所以最开始我们可以看到会员的里面大部分人,绝大多数人都是老的那一代呃,索网推的技术人员。对,然后不加上21年22年这个腾讯做了非常大的一轮降本增效。其实人手是远远不够的,不管不光是人不够,卡也是很不够的。所以这个是一个有点残缺的一个组织。对,我们比较早的时候有一个报道,就是23年初的时候应该还提到过,就腾讯在22年其实有砍了部分GPU的订单。是的,所以整个会员最开始只有2000张卡,然后这0...(已截断)
All-In with Chamath, Jason, Sacks & Friedberg
(1)
Chip Stocks Crash, $20B Fund Margin Called, Frontier Labs: SLOW DOWN AI, Mamdani's Grocery Stores
1h 36m
2026/08/01
📝 AI
总结
Alright, everybody, welcomee back. Welcome back to the number one podcast in the world. The core for, fantastic for the original.😊Cortet is here. Did we peek at like number two or number three last week. Is that what happened.I think it was number 4 in the world. So, yeah, usually we're number oneing US trending US trend. we're usually number one globally, but yeah, and sometimes number four US. I forgot my starlink. So let me apologize to everybody. That was a critical error when you're on the road or on the water. But, you know, itll be fine. I had five huge leads come in this week. The Glen Gary go big enterpriseprise leads. Enterprise sales is a bear because it's super chunky, but the deals are ginormous. Sos actually got.advicedvic for Chima from your enterprise sales days. you closed some of those big 7 figure deals when you were doing yammer. No leverage. Don't put on leverage. and no leverage, leverage will be name of the show today. leverage equals risk of ruin. PA, no leverage. I have the situational awareness to not lever up. That's good. Yeah, you want that, that situational awareness. I mean.😊Kind of out there. I mean, if you name your fun situational awareness, that's.Yeah, come on the pot anytime loopopold. Al right, everybody, we gotta talk about chip stock crashing effort an all time run up and.😊We had a major hedge fund get margin called and some incredible margin calls happening in South Korea, Leopold Ashhan Brenner is a 25 year old hedge fund manager, he left Open AI two years ago to start his own fund and apparently according to reports, this is breaking news on Thursday when we tape he got margin called and had to sell his entire public portfolio to cover massive losses caused by his leverage and who bought them none other than Citadel's Ken Griffin.We don't know if was K Griffin himself, but Cadel bought it according to the early reports, Leopold had insane returns and he rode the wave of AI and chips and frontier labs as recently as this month and he started the fund with but $225 million in 2024 he grew it 100 x to 20 billion this year or so ran it all the way up to 45 billion now he's at 200x earlier this month by trading on leverage according to our friends.At CNBC at the end of the June, he was reportedly up.F x, 450% this year, some have reported.That he's also selling his massive anthic stake to cover these losses. But the Wall Street Journal is disputing it again, we're happy to have him here on the program. How did this all blow up. Well, Nasdaq' chip index.😊This is called the Philadelphia Semiconduct index is down over 20% over the last month. that's bear market territory. Obviously, definition of bear market territory for those of you who don't play in the markets is anything over 20%. The index included the top 30 US listed chips. That's people like NviDdia, TSMC, AM D, micron, you know all those big names. But the index bounced back a bit today up 7% when we're taping. So we may have found a bottom. Unfortunately for Leopold, he had already sold Samsung and Sk Heinix 2.South Korean chip companies that are not included in the Nasdaq index also got smashed, crushed, demolished Samsung down 38% over last month, SK Heinix down 14% since going public three weeks ago, the cost be that South Korea's version of the S&P 500 is down over 40% in the last 40 days between last Friday and Wednesday, leading chip companies she over a trillion dollars in market cap combined, so to put this in context, chip stock,Had a legendary run the past couple of years. But, you know, trading on leverage, and we'll talk about it very dangerous. If there's a downturn, we'll get into the South Korea wrinkle as well, even with this downturn.The five year results are still spectacular from off micron up 850%, mostly in the last year, in video up.875% in the last five years in Brockcom up 663%. let's discuss it. If I was going to give you one piece of advice when you're running risk is you have to manage leverage incredibly carefully, because when it runs ahead of you, the unwind is incredibly violent and it's incredibly quick. That's the biggest problem with.With running either massively levered long or massively levered short. So I don't know to what extent he was running lever, but the rumors are, he was running like three and a half turns, which, just to give you a sense, when you're running that much risk, a 3% and 4% move is amplified 12 and 13. But if you saw what's happened in the last three days, a 25% move is amplified 75%.So it has the risk to stop you out. And what happens is.When you get that leverage, the banks are given the authority to close you out and when they close you out, what they do is they start calling around and unwind your risk and you don't have much of a choice. It's sort of an automatic one way ratchet.So if everything that has been reported is accurate, he was running about three and a half times levered.The market moved against them.It lost.A very large percentage of his...(已截断)
The Joe Rogan Experience
(1)
#2534 - Annie Jacobsen
2h 20m
2026/08/01
📝 AI
总结
🎼The Joe Rogan experience during my day, Joe Rogan podcast by night day.😊Biological war. Oh, boy. Every time I talk to you, I'm fucking terrified.That is my part time job to actually scare Joe Rogan, if I can do that, wow, yeah, every time I talk to you, I'm like, oh, no, Last time it was nuclear war. Oh, no, Now it's biological war. Oh, no, which is actually maybe even grimmer the nuclear because at least in nuclear, you're just like gone in the flash.I did a show a while back on the Scifi Network called Joe Rogan questions everything. And one of the episodes we did was about biological war. And so we spoke to this guy who used to work for Russia during the Cold War. And he was talking about how much anthrax they had. And how much work they had done on biological weapons. And it was terrifying. He So they had like just giant vats of anthrax. like enough to kill everyone in New York City..was it Ken Aback, Do you know, He was Russian, Russian. Yeah, there was a couple of those Russian defector scientists who came here and told us everything they were doing. There was like.You know, tons of anthrax and plague that. That's actually the pathogen that I use in the scenario. I write, is plague because it's so crazy, scary, so dangerous, sits at the top of the government's Tier 1, select agent.Of dangerous pathogens. So one of the things that we did during this episode, we went to Galveston to, I guess is a CDC lab that they had there. So they have these super thick walls and everyone's wearing spacesuits and they have these tubes attached to them. And they were studying all these different really horrible, fatal diseases. And I'm like, well, what if this stuff gets out, Like it's right here, Like,, what happens if it gets out, Like, will we take ever precaution this and that.But his concern, he said, was not for genetically engineered viruses. His fear was really just from natural virus. He says that is just as scary as anything that's out there. This is the Russian scientist. You. No, this is a different guy. This is an American scientist. Yeah, I think they like to say that. You know, Mother Nature's is just as powerful. But in truth. That's actually not true, because think about what you can do with genetic modification. make it more virulent, more transmissible. Yes, and make it defeat.Existing vaccines and antibiotics. Well, we are right now in the middle of the Fauci hearings. So we are He is now.Testifying in front of Congress and pleading the fifth about everything, including the color of his tie. It's very weird to watch. It's in in one way, it's like I told you to a lot of my liberal friends that wanted to pretend that he was a great guy and that he wasn't doing anything wrong. And now that all these emails have gotten out or it shows that he knew that the fatality rate was much lower than they were saying he knew that it wasn't dangerous for children that they didn't need vaccines. He knew that the social.Disttancing thing was all nonsense.And he's got a blanket pardon for 10 years. And the thing is, it's like.These people get money for research. And that's what they do. They do research. So if you get money and you get grants for research, Well, you're going to go do research. And if your research is gain of function research that makes viruses more infectious to human beings.And, you know, you have to do it because that's your job. That's what you do. That's what you studied. That's what you went to school for. Like what a crazy fucking thing we're doing.It is bananas. And it's also, I mean, it was really unfortunate. I thought that yesterday you took the fifth because I, for one, just want to hear.Call me, Paul Lena. The truth be told. But he's not going to tell you the truth, no matter what. He's going to obfuscate and dance around it.Use words that don't totally apply to what he's saying and just ramble on and.I mean, Covid absolutely could have been this incredible lesson for all of us. I mean, the nightmare that it was. And instead we're just left. I mean, can you believe that we still.With all the power of national security that we still allegedly do not know where the virus came from.don't definitively. I mean, definitively, people certainly have their ideas. But I think it's high probability that it came from the Wuhan lab. I think most experts are now saying that. And I think the only reason why they weren't saying that was because of Fauci.And it was really because that he had funded this through the Eco Health Alliance.Look, I say that in a very like 90000 f neutral way, in other words.Every sub, I leave Covid.That part of it out of my book., It becomes like a.Like a metaphor that underpins the narrative. And people end up reading the book and going, wow, I actually think differently about Covid, which was kind of my intention. But every subject matter expert I spoke to encourage me to put in the book that almost certainly it was a lawbleak. Like it's just, so when I say we still don't kn...(已截断)