#22: Your GPT Will Make More Money Than Your Substack? Please Tell Us More…
发布时间 2023-11-10 18:00:29 来源
摘要
The gang discusses Open-AI's developer conference and whether the new ‘GPT Store’ is a serious platform developers and entrepreneurs should care about, or just another slick PR announcement. Also thoughts on the IPO market (why bankers keep saying it isn’t open, but will soon) and way too much discussion of The Morning (After) Show.
GPT-4正在为你翻译摘要中......
中英文字稿 ![](/static/new-window.png)
I was coming into this going, oh man, Sam gets to be called right on a whole bunch of vectors. And I think, oh, this is my favorite. Let's talk about this for the rest of the episode. But all the ways I've been right.
我从一开始就有这种想法,哇哦,山姆在许多领域都被称为正确。我以为这是我最喜欢的。让我们在这一集的剩余时间里谈论这个。但是,一切都是我正确的方式。
Hey guys, welcome to more or less this week coming to you from my iPhone because I can't figure out my home wifi. How's the wifi in the pool house? Apparently my desktop is just stopped logging in to Riverside. So we have some Riverside issues. We're going to have to figure out. I upgraded Chrome and it just will not log me in. So just to be clear, the lessons have less good audio this week and less good tech and the more. Let's get technology. The more we are not good at the tech. Have our mics ready. I will take that.
大家好,欢迎收看本周的《多多少少》(More or Less),我用我的iPhone向大家传递消息,因为我无法弄清楚我的家庭无线网络。游泳池小屋的无线网络怎么样?显然,我的台式电脑已经停止登录Riverside(可能是某个网站或应用)。所以我们有一些Riverside的问题,我们需要解决。我升级了Chrome浏览器,它就是无法使我登录。所以,明确一点,本周的课程音频质量较差,技术也不是很好。让我们更加注重技术。我们在技术方面还需要改进,但我们已准备好麦克风了。我接受这点。
But how are we guys doing? I there's been a lot of activity this week just happening in the world. A lot of cons are going. You guys were at Seinfeld and John Mayer to get that right. And you guys were at Seinfeld, right? You were. It was hilarious. We also that night you were at Seinfeld. We went to the Warren Miller film premiere for anyone who likes skiing. That was happening. I feel like Warren Miller is not what it once was.
但是我们这些人怎么样了?这周世界上发生了很多活动。很多会议正在进行中。你们去看了Seinfeld和John Mayer对吗?你们去了Seinfeld,对吧?你们去了。那真是太搞笑了。那天晚上我们还去了Warren Miller的滑雪电影首映会,适合喜欢滑雪的人。我觉得Warren Miller已经不如以前了。
What? Sam, what a hair tech. You know, Sam, that that was true for many years. But this years was actually quite good. It gave me all the all the feels. They did a 74 year retrospective on Warren. So it was like full of lots of. They did the big Nissan commercial again. I feel like there was like years. Basically just a big Nissan. What a Subaru or something. No, it was it was more retrospective this year. It was great.
什么?山姆,你好潮的发型啊。你知道,山姆,那个发型多年来一直很流行。但是今年的发型实际上非常棒。它带给了我很多感动。他们对沃伦进行了74年的回顾。因此,充满了很多东西。他们又播放了那个日产汽车的大型广告。我感觉好像这一年很特别。基本上就是一个大日产汽车。就像一辆斯巴鲁什么的。不,今年的表演更多的是回顾。真是很棒。
Who was Warren Miller for the people who don't know? Not all our listeners. If you don't know who Warren Miller is, we've just like stop listening. It doesn't matter. Yeah. Oh my God. Whatever. Listen, I did not know who Warren Miller was until I met Dave Warren. So Dave explained.
对那些不知道沃伦·米勒(Warren Miller)是谁的人来说,沃伦·米勒是谁呢?并不是所有的听众都知道。如果你不知道沃伦·米勒是谁,我们就像停止倾听。但这并不重要。是的,天啊。无论如何。听着,直到我遇见戴夫·沃伦之前,我并不知道沃伦·米勒是谁。所以戴夫给我解释了一下。
Warren Miller, you know, for 75 years or I guess he died a few years ago in 2018, but he made the very best ski movies kind of pioneered. He was like the skiing sports great marketeer. He was the storyteller. Well, for many years, they were the best because they were the only. Yeah. And then a young crop match stick and TGR or whatever came in blue. Warren Miller out of the water. Yeah. But and they had some corporate drama. Yeah. It's good. We're back at it. Warren's always always had the good humor. You know, I always love the humor of a good Warren Miller movie. Yes. Yes. They always had the good people doing stupid shit on skis was always their best segment. But it was nothing compared to Seinfeld. That was one of the most hilarious things I've. I think we had some different jokes. I heard some AI jokes on Twitter. I don't think we had the level of AI jokes that you had. Oh, yeah. He did a the whole opening bit was both phones. He did phones and then he did AI. That was so good. The phone was about how like you don't know something happens until your phone tells you eight years later that you were doing the thing. Yeah. I love that. I also love that he kind of comes out on stage and just says, I don't care what you do with your phones. Yeah. Everyone take out your phones and then I noticed actually everybody did and then they put them away. And so it was kind of a genius master move in comedy. I thought versus Chappelle who like made you put it in a weird fucking pouch that you can get out. Yeah. Yeah. Yeah. I thought it was a.
沃伦·米勒,你知道的,他制作了最好的滑雪电影,大概有75年了,我猜他在几年前去世了,2018年吧,但他开创了这个领域。他就像滑雪运动的市场推广者,他是一个讲故事的人。嗯,多年来,他们一直是最好的,因为他们是唯一的。是的,然后一群年轻的选手,比如Matchstick和TGR,他们突然就超过了沃伦·米勒。是的,他们也有一些公司纷争。嗯,现在又回来了。沃伦总是有很好的幽默感。你知道,我总是喜欢看沃伦·米勒电影中出现的幽默。是的,是的。他们总是找到一些搞笑的人在滑雪时做傻事,这常常是他们最好的片段。但和《宋飞传》相比,那就没什么了。那是我看过最搞笑的东西之一。我觉得我们有一些不同的笑话。我在Twitter上听过一些关于AI的笑话。我觉得我们没有达到那个水平的AI笑话。哦对,他整个开场的笑话是关于手机和AI的。太棒了。他关于手机的笑话是说,你不知道某件事发生了,直到你的手机在八年后告诉你你当时在做某件事。我喜欢那个。我也喜欢他上台后说,我不在乎你们怎么用手机。然后我注意到其实所有人都拿出手机,然后把它们收起来。这是一个善于喜剧的天才之举,我觉得与Chappelle不同,他让你把手机放在一个奇怪的袋子里拿不出来。是的,是的。我觉得那是一个聪明的喜剧策略。
Can we talk about this for a second? I'm curious in your couple relationship and if you're a marriage like are you if you go to a concert, let's say John Mayer, like, do you have your phone out most of the time recording the concert? Where are you like participating in the concert? Because Dave takes amazing videos, but then I were I worry for him while I'm trying to enjoy the concert that he's not getting the most out of the acoustics.
我们可以简短地谈一下这个吗?我对你们的伴侣关系很好奇,如果你们去看一场音乐会,比如约翰·梅耶,你们会不会大部分时间都拿着手机录制演唱会?你们是参与其中还是享受演唱会呢?因为戴夫拍摄的视频非常棒,但当我试图享受演唱会的时候,我担心他可能无法充分体验到声音效果。
You know, I was hoping she would say this because last night I was very restrained in my restraint last night. Yeah. Well, Brit was texting with, you know, some mutual friends of all of ours for like basically the entire concert. We were texting about how hot John Mayer is when he plays a guitar like that, by the way. For women, it's hot. Yeah, I don't, Sam, I don't think we're like hold up our phones record live experience. People, are we? No, but I'm not exactly known for being in the moment either. No, but you know what? We're going to talk about that later. That is not a public conversation.
你知道吗,我希望她会这样说,因为昨晚我在克制情绪方面非常克制。是的。嗯,布里特和我们大家的一些共同朋友在整个演唱会期间一直在发短信。顺便说一句,我们在讨论当约翰·梅耶弹奏那样的吉他时看起来有多帅,对女性来说这是很性感的。嗯,我不认为我们会拿出手机来录制现场体验,对吧?不是,但我也不是以身处当下而出名。不过你知道吗?我们以后会谈论这个。这不是公开讨论的话题。
But, but Dave, isn't this like all your AI hardware that's going to like make the screen vanish? Now I understand why you like it because. I don't know. I guess, look, I'm a photographer. So I kind of at concerts, I always try to do good photography, but I actually have shifted my MO on this. I try to get a couple of good ones and then be present for the rest of the concert. And then just beatbox the music too. What, well, we're on this. I assume I'm the only person who rushed to watch the season finale of the morning show. Not true. Exactly when it came out last night. No. I kind of pretend like she was trying, but she didn't try at all. They market that it comes out on Wednesdays. Apparently it's midnight Wednesday. Everybody knows. It's actually earlier. Anna was at a John Mayer concert. Okay. What? Okay. Well, this was just. That's dedication. This was missing. Well, that's like how the market of the Green Day tickets weren't available until today, but I bought them last night. I don't know what's going on.
但是,但是戴夫,这不是你的全部AI硬件吗,它会让屏幕消失?现在我明白你为什么喜欢它了,因为...我不知道。我猜,看,我是一名摄影师。所以在音乐会上,我总是努力拍好照片,但是我实际上已经改变了策略。我试着拍几张好照片,然后剩下的时间专心享受音乐。然后还可以用嘴模仿音乐节奏。嗯,我们继续吧。我猜我是唯一一个急着看《早间新闻秀》季终的人,不是吗?不对。确切地说应该是它昨晚一上线我就迫不及待看了。不过我觉得她只是假装尝试了一下,其实根本没努力。他们宣传说它会在每个周三发布,显然是周三凌晨。大家都知道。其实比那更早。安娜去了一场约翰·梅尔的音乐会。好吧。什么?好吧,这就是。真是敬业啊。不过这样就错过了。就像绿日演唱会的门票今天才开售一样,但我昨晚就买了。我不知道发生了什么。
Ooh, we're going to Green Day. Are you sure that as a reporter, you don't get early access? I'm positive. Apple doesn't give me any early access. Wait, business model innovation. How much more would you pay to watch the morning show one day ahead of the rest of the world? I probably wouldn't because it, I don't need to. Man. It don't need to be first or something. I just like, I just got to say. I don't know what's the few others of this show, but I cannot.
哦,我们要去看绿洲乐队的演出。作为一名记者,你确定没有提前进场的特权吗?我确信没有。苹果不会给我提前进场的机会。等等,商业模式的创新。你愿意为提前一天观看《早间新闻》节目付出多少额外费用?我可能不会,因为我不需要。朋友,我并不在乎是第一个或者什么的。我只是喜欢表达自己的观点。我不了解这个节目的一些细节,但我无法。
Every time it just sounded like the morning after pill to me. Like the title sounds like that. Not what the show's about. And look, I got into some hot water on the X because I was praising this and a lot of people were like, it's fucked. But there were some week episodes this season, but Billy Crudup, did I get that right? We saw their food. Like, I can't believe anybody was saying that. It's fantastic. They can act. It's fantastic. And the season finale, there's so many moments and it's not, they don't know anything to do with the news business, it's to do with relationships. Such good acting. Unbelievable acting. This is a past for me.
每次听起来都像是早晨避孕药。就像标题听起来的样子。不是节目的内容。看吧,因为我夸奖这个节目,我在X节目上陷入了麻烦,很多人都说这很糟糕。但这个季节有一些平庸的剧集,但比利·克拉普德,我说对了吗?我们看到了他们的表演。就像我简直不敢相信有人会这么说。他们可以演戏。太棒了。季节终结篇,有这么多的时刻,和新闻业务无关,而是关于人际关系。如此出色的演技。令人难以置信的表演。对我来说,这是一个过去时刻。
This scene is, well, the reason I could watch this without Sam is because I have COVID and he does not, hence my sultry voice and why I watched The Morning Show because Sam is not into this, but acting is just so good and just major props, major props. You just like John and I. John and I agree. John and him is fabulous. I just have to say. Sam, turns out you're less morning show. I understand. You know, Sam, I initially thought John and him may not be good in this role, but I think he really delivered. Just crushes it. And he's playing, I suppose we should talk about technology.
这个场景,恩,之所以我没有和山姆一起观看是因为我得了COVID,而他没有,所以我的嗓音变得妖娆,以及为什么我看了《早间新闻》,因为山姆对这个不感兴趣,但是演技真的太好了,无可挑剔。就像我和约翰一样,我们都认同约翰和他太棒了,我必须说。山姆,结果你不像早间新闻那样。我明白。你知道,山姆,我最初认为约翰和他在这个角色上可能不好,但我觉得他表现得非常出色。简直厉害。而且他在扮演角色的时候,我想我们应该谈谈科技。
I suppose we should. John Hand is playing the Elon Musk rocket maker of the episode. His company Hyperion is supposed to be space that. So although he seems more Bezos to me. No, he's definitely Elon. No, no. I mean, he's one with a little bit of Bezos. Yeah, he seems like a hybrid to me. Dead. I mean, this is not one minute. Show this season, walking by Jessica, watching it. And John Hand was like making an omelet than morning after they had sex, I guess. And like what like the Elon Musk does not make omelets for his significant others. Any number of them. That's why I'm saying Sam. This is it's a new character. It's a hybrid. Yeah, this is your point of oh my God. Well, just to the it was so good. Anyone who wants to discuss this in our back channel on the internet, I'm there for it. I'm looking for people. I've been tired for days.
我觉得我们应该这样做。约翰·汉德扮演的是这一集中的埃隆·马斯克火箭制造者。他的公司Hyperion被认为是太空公司。所以虽然他对我来说更像贝佐斯。不,他明显是埃隆。不,不是。我的意思是,他有点像贝佐斯。是啊,他对我来说是一个混合体。我是说......这不是一个分钟的节目,我看到杰西卡经过时,约翰·汉德就像在做一个煎蛋卷,可能是他们做完爱后的早晨。埃隆·马斯克可不会为他的亲密伴侣做煎蛋卷,无论是多少个。所以我在说,这是一个新的角色,是一个混合体。是啊,这是你说的意思。哦,我的天啊。只是......太好了。任何想要在互联网上讨论这个的人,我都在那里。最近我已经累了好几天了。
It's been about Moss and somehow we're to get it from that to the morning. Hey, we need a way to get the women to speak. I couldn't bridge. I couldn't find a bridge. So I left. But that's why I need you guys. You should drop into a bad channel. Complete non-segment is like intense political conversations like, Hey guys, John Hand morning after peptides. Let's see how that goes. I don't know if he's a pet type or not, but it's working for Jennifer. I mean, Jennifer Aniston who is hasn't aged a day since Rachel on front. You told me that she's a great asset to keep showing that. Yeah, she's been naked a few times. Yeah, I would agree. She is the whole thing. And like, I just you go, Corey fucking Ellison, right? That's the line. Everyone. Just hold that in your head as you watch this. I'm never going to watch this. I'm never going to watch this Nor will most of our listeners. So just give me the one lighter, which is at the end of the season. No, no, no, no, no, no, no, no. We are not doing a spoiler, Sam. I'm sorry. I'll go to the pool house after this and give it to you, but it's just, it's just so good.
大约已经过去了一段时间,我们要从那时到早晨。嘿,我们需要一种让女人们说话的方式。我无法架桥。我找不到桥梁。所以我离开了。但这就是为什么我需要你们这些家伙。你们应该来个不太好的频道。完全不切片,就像激烈的政治对话,比如,嘿,伙计们,约翰·汉德早上打肽链。让我们看看情况如何。我不知道他是不是聚合型的,但对詹妮弗来说是有效的。我的意思是,詹妮弗·安妮斯顿从未老去,自从在《六人行》中扮演瑞秋以来。你告诉我她是一个很棒的资产继续展示这一点。是的,她几次裸体出镜。是的,我同意。她是全部的一切。就像是,你去吧,科里,该死的埃里森,对吧?这是台词。每个人都把它记在脑子里,看这个的时候。我永远都不会去看。大多数听众也不会去看。所以给我一个关键词,在本季节目的最后。不,不,不,不,不,不,不,不。我们不会剧透,山姆。对不起,我会在之后去泳池屋给你,但它只是太棒了。
Okay. What happened in the world? The technology this week. What happened? We had AI, OpenAI Developer Day. More interesting than I thought. I experienced OpenAI Developer Day, which I appreciated, but several people on Twitter saying, Oh my God, Sam, lessons right. And I didn't know what that meant, but I think I pieced it together, which, and I always appreciate praise. I think the idea being that OpenAI has now swiped all these AI startups for sure and being an AI startup is a bad idea, which has been my line for a while. Is that basically right?
好的。世界上发生了什么?这周的科技动态有什么?我们有AI、OpenAI开发者日。比我想象中更有趣。我参加了OpenAI开发者日,我很喜欢,但是Twitter上有些人说,“天啊,Sam,你说得对。”我不知道那是什么意思,但我想我还是理解了,而且我总是很欣赏赞美。我想他们的意思是,OpenAI现在确实吸收了所有这些AI初创公司,而成为一个AI初创公司是一个坏主意,这一直是我的观点。这基本正确吗?
Yeah, I was, I was coming into this going, Oh man, Sam gets to be called right on a whole bunch of vectors. And I think, Oh, this is my favorite. Let's talk about this for the rest of the episode. But all the way, I've been right. But I think that you also have to give OpenAI serious credit for being in just absolute beast mode right now. This is one of, I think, the most impressive developer platform launches since maybe Facebook platform, if I do say so myself. And it's just unbelievable the amount of stuff they're putting out there. And while I think, Sam, your perspective has been broadly correct that, you know, be careful which business you're getting into and which startups you're funding. I do think that another aspect of our conversation, which is the creator economy and how these new tools are going to be used by my, you know, smaller entrepreneurs, micro entrepreneurs is actually quite an interesting conversation.
是的,我是这样想的,我原本以为萨姆将在很多方面被称为正确,我想,哦,这是我最喜欢的部分。让我们在剩下的节目里谈谈这个。但是到目前为止,我一直是正确的。但我认为我们也必须对OpenAI给予真正的认可,他们现在处于绝对强大的状态。这可能是自Facebook平台推出以来最令人印象深刻的开发者平台之一,如果我这么说的话。他们所发布的东西实在是令人难以置信。虽然我认为,萨姆,你的观点在很大程度上是正确的,你要小心投身于哪个行业,资助哪些初创公司。但我认为我们对话的另一个方面,即创作者经济以及这些新工具将如何被我那些更小规模的企业家、微型企业家使用,实际上是一个非常有趣的对话。
Can you, can you read from this with me? Because I actually haven't get as much attention to this. What like, what are the dig on locks for small businesses that are now?
你能,你能和我一起从这里读吗?因为我实际上对这个并没有给予太多关注。关于目前适用于小型企业的网络安全方面,有什么是被广泛讨论的呢?
Oh, can I go? Can I go? Yeah.
哦,我可以去吗?我可以去吗?是的。
So like just a couple hours ago, actually, I had like thousands of women on a zoom webinar. And I was for this, for self-made, which is a women's business accelerator, I help run. And I was explaining to them how what you're underplaying it.
就在几个小时前,实际上,我在一个Zoom网络研讨会上与成千上万的女性交流。我为女性创业加速器"Self-Made"工作,我是其中之一。我向她们解释了你正低估这件事的重要性。
Tell everyone what self-made is. Self-made started in the pandemic when millions of women were laid off or furloughed from their jobs. And I decided to help them just become entrepreneurs instead to make their own revenue and have flexible hours, et cetera. Most of them are small business entrepreneurs and not necessarily venture scale, some are venture scale. Real entrepreneurs.
告诉大家“自立自强”的含义。自立自强源于大流行时期,当千万女性失去工作或被暂时解雇时。我决定帮助她们转变为企业家,以获取独立的收入并拥有灵活的工作时间等等。她们大部分是小型企业家,并非必然具有风险投资规模,但其中也有一些是风险投资规模的创业者。真真正正的企业家。
Yeah. And I was like, ladies, if I could reach through the screen, I would have to shake them and be like, wake up. This is your moment. Like, this is the time, like just yesterday, you know, now there's like GPT apps that you can create and bots and, you know, all these different things.
是的。我就觉得,女士们,如果我能透过屏幕伸手进去,我一定要把她们摇醒,告诉她们醒醒吧。这是你们的机会。就像刚刚昨天一样,现在有了GPT应用、机器人和其他各种不同的东西。
And, you know, if you are first to the punch here, like, there's such an opportunity to start monetizing in so many new ways and to not have to know how to build an app or something like that and to have AI help you do things like, I mean, I thought what Sam demoed, it was really interesting, right?
而且,你知道的,如果你能在这方面抢占先机,那么有很多新的机会可以开始盈利,而且也不需要知道如何建立一个应用或者类似的东西,而是可以让人工智能帮助你做一些事情,比如,我觉得Sam演示的内容很有趣,对吧?
Like, he took some lectures from his Y Combinator days, right? And he uploaded them. And then he made like a Ask Sam bot, right? Which was like, ask him a- A VC advice. You guys are out of a job. VC advice bot or something, right?
他从他在Y Combinator的日子里参加了一些讲座,对吗?然后他把它们上传了。然后他就制作了一个像Ask Sam机器人,对吗?那个机器人可以提供给他关于风险投资的建议。你们将面临失业的风险投资建议机器人之类的东西,对吗?
And- That's exactly what the morning is. VC advice bot. Thank God. That was what I was thinking. I was just saying, it was, it did it in a matter of like a minute, right? And so it just blows your mind.
然而,这正是早晨所能带来的(感觉)。VC(风险投资)建议机器人。谢天谢地。这正是我在想的。我只是在说,它只用了不到一分钟的时间,你绝对会惊讶。
It was impressive. The power that can exist and how people can take advantage of it and monetize from it now.
这真是令人印象深刻。现如今,力量的存在以及人们如何利用并从中获益,这一切令人惊叹。
So what's the monetization plan as you buy these bots and they talk to them? You get like, after revenue share.
那么,在您购买这些机器人并与它们对话的过程中,有什么盈利计划呢?您会从中获得类似的收益分成方式吗?
I mean, basically, so you guys will remember when we had Scott on the podcast and we talked about agents, right? And agents are going to be everywhere. You're going to have AIs to do this and AIs to do that.
我的意思是,基本上,你们会记得我们在播客节目中邀请了斯科特,我们谈到了代理人对不对?而且代理人将无处不在。你将会有人工智能来做这个,人工智能来做那个。
And I think, again, it's a compelling demo experience version one, but to be able to create a GPT and train it on literally, like he opened his computer and dragged his lectures into it, right?
我认为,再一次说,这是一个引人入胜的演示体验第一版。但是,他是不是能够像打开电脑然后把讲座拖进去一样创建并对GPT进行训练呢?
Like, I instantly thought for some reason onboarding. Like, we have so many onboarding docs with information, like so many, what's information story? I've written about it in a million times. Our team has talked about it in a million ways. A basic thing.
我觉得这里应该是入职培训吧。我们有太多关于入职培训的文档,内容繁多,这里主要是讲些什么?我已经写了无数篇关于它的文章了,我们团队也已经以各种方式讨论过。实际上是一个基本的事情。
It's not going to replace having your onboarding session, but just like a little GPT. Very, very, very, very, very well.
它不会替代你的入职会议,只是像一个小型GPT(文本生成模型)一样。非常非常非常非常非常好。
I love that they've managed to brand GPT as their thing. That's very clever. I know. I thought that was a little funny, too. It's like, but that point is, so do they get access to all these? I actually just haven't followed this.
我喜欢他们成功将GPT作为他们的招牌品牌。这非常聪明。我知道。我也觉得这有点有趣。就是说,他们能够获取所有这些资源吗?实际上,我对此并没有做过深入了解。
So the idea is, I bet not all of our listeners have, you get to take all of your content, you chuck it into a bot that you instantiate, and then basically people can chat with that bot and they host it and you can charge for it. Is that basically it? Yes.
所以,这个想法是,我打赌并不是我们所有的听众都拥有这个要素。你将所有的内容放入到一个你创建的机器人中,然后人们可以与该机器人聊天,你可以托管它,并向他们收费。就是这样吗?是的。
Okay. Do they get access to the content? Who owns the content? That's up to the creator. There are privacy controls over the content. You can set your own policies.
好的。他们可以获得内容吗?谁是内容的所有者?这取决于创建者。内容有隐私控制。您可以设置自己的策略。
Is there any reason that the creator would have been incentivized to let open AI have access to the content? I guess it would be up to the creator, right?
有什么理由让创作者有动力让Open AI获得访问权限呢?我猜这取决于创作者,对吗?
Do they think that they're. But is there any data that's letting open AI have any access to it? I'm not sure. It's a great question.
他们是否认为他们是。但有没有任何数据让开放AI可以访问?我不确定。这是一个很好的问题。
There's also. The rev share is a little murky because it's a cut. I think that's a good question to look into understanding the subscription rev share a little bit.
还有另外一点。由于所得的份额不确定,利润分成有些模糊。我认为这是一个值得探究的好问题,可以更好地理解订阅收入分成。
Obviously, there's also. Is there a subscription over the all or do I charge my own. Is it a bot that I'm selling and they get a cut or is it like I get some adhesive? It's like the Apple App Store. It's like the App Store. Yeah. It's a bot that they're selling that both sides get a cut. Okay. Interesting.
显然,也有这个。所有的东西都有订阅吗,还是我自己付费。它是否像我卖掉一些粘性物品那样是一个机器人,并且它们会分成,还是说我会得到一些粘合物?就像苹果应用商店一样。就像应用商店一样。是的。它是一个他们正在销售的机器人,双方都会分成。好吧,很有趣。
And then there's also Sam. Enterprise customers can also deploy internal only GPTs. No, is it basically the story of instead of writing a book because no one's going to read my fucking book anymore, instead I just basically dump my book into GPT and look at the conversation at all? No, because it's push versus pull, right?
然后还有Sam。企业客户也可以部署仅用于内部的GPT。不,这基本上是这样一个故事,我不写书了,因为没人会读我的该死的书了,相反我只是将我的书简单地倒入GPT中,看着对话转变吗?不,因为这是推动与拉取的区别,对吧?
If you want to push your book onto people versus force people to ask questions that they don't know what to ask you about. You will definitely create a Sam GPT. I'm surprised you have that. Of course I will. I'm starting to have actually.
如果你想推销你的书,而不是强迫人们问他们不知道该问你什么。你一定会创建出一个Sam GPT。我很惊讶你已经有这个了。当然会的。我开始实际上有了。
Yeah. Frankly, guys, I think we should create a more or less GPT. We can just dump all of our transcripts from previous episodes.
是的,老实说,我认为我们应该创建一个更或多或少的GPT(生成式预训练模型)。我们只需把以前的剧集所有的对话文稿都倒进去就行了。
But this is the thing. This is my question. Obviously, I'm actually curious in terms of how this will play out because I totally get this idea that like. You take a book, for instance. There's a talk to an author type thing. You can read the book, which is the story I told you, or you can just be like, hey, here's a corpus of stuff I've generated and now ask you questions. I believe that's a thing. But it's not a hundred. It's the same thing with a knowledge base for how to use an app. If you said, okay, instead of your documentation, O'Reilly style, it's like I'd rather have that in a Q&A-able thing. I get that. But it is an interesting form factor in terms of how people actually buy and interact with these things. It basically is just content with a different index into it.
但是这是个问题。这是我的问题。显然,我实际上对这个将如何发展非常好奇,因为我完全理解这个想法,就像你拿一本书,有一种与作者对话的感觉。你可以阅读这本书,也就是我告诉你的故事,或者你可以简单地说,嘿,这是我生成的一些东西的资源库,现在请问你问题。我相信这是一个可以做到的事情。但并不是百分之百。这就像使用应用程序的知识基础一样。如果你说,好吧,不要你的文档,像 O'Reilly 风格那样,我宁愿有一个可以进行问答的东西。我理解这一点。但就实际购买和与这些东西互动的形式而言,它是一种有趣的形态。它基本上只是具有不同索引的内容。
Yeah. I was thinking about this because I started messing around with. I started actually trying to write a therapy bot with content that I'm just aware of. I started working through it. Really that's what I came to is that, okay, now basically every author on Earth can take their book content and put it into one of these bots and monetize it in another way. It's yet to be seen how incrementally valuable that is on top of the picture. It's just like, Jess, you run a subscription business. You're going to dump the entire information archive into one of these things that people can ask questions in every article you're written or. Sure, why not? Practically speaking, I'm actually curious, is this a very obvious use case for some of my. I would charge a lot of money for it. I'm going to make a written co-bot, too. But it's easy. Anyone who has content will just dump it in. The question is to me, what's the distribution and why are you. What's the value of it? Yeah, well, I think. These are all excellent questions.
是的,我在考虑这个问题,因为我开始瞎折腾。我开始尝试写一个我熟悉的内容的心理治疗机器人。我开始琢磨着。事实上,我发现,现在基本上地球上的每个作者都可以将他们的书籍内容放入其中一个这样的机器人,以另一种方式进行盈利。至于这是否具有增值效果,这还有待观察。就像Jess,你经营一个订阅业务。你要把整个信息档案库都放进这样的东西里,人们可以在每篇文章中提问或者……当然可以,为什么不呢?实际上,就我个人而言,我很好奇这是否是我使用的一个非常明显的用例。我会为此收取很高的费用。我也会制作一个书面合作机器人。但这很容易。任何拥有内容的人都可以直接放进去。对我来说,问题是分发和你做什么。它的价值是什么?是的,嗯,我认为这些都是很好的问题。
I think that the use case of chatting with a bot is going to exist, but I don't know how mainstream it's going to be. I have one big question now. It's like, what is the usage of borrowed? What is the. I mean, chat GPT, obviously we know. Information is reported a lot about their revenue surging and opening. I mean, people are paying for this stuff. Companies and small businesses are paying for it. But I haven't seen the chat bot interface go through the moon yet. But I do think it will gain more traction. Do I think it's how people are going to want to consume their news? No. But do I think for our $1,000 pro tier, doing some of the. It's a cool software product that some people get value out of. Absolutely. I mean, what's so impressive, and again, it was a good demo, is they made it seem really easy and really easy to refine with natural language. I think that's an interesting idea. Again, I should watch it, and I'm sure it is a cool demo. And I just think it's like, the fact that it's easy to make is not in and of itself making it a compelling product. It's cool. But I'm very curious what the actual. Are people going to pay more for the book that instead of it being read to them, or they're reading it, that they can just ask questions?
我认为与机器人聊天的用例将存在,但我不知道它会变得多么流行。现在我有一个重要的问题。就像,借用的用法是什么?我的意思是,聊天GPT显然我们都知道。有很多关于他们的收入激增和开放性的报道。我的意思是,人们正在为这些东西付费。公司和小企业正在为此付费。但我还没有看到聊天机器人界面被广泛使用。但我确实认为它将获得更多关注。我认为人们会想要通过这种方式获取新闻吗?不会。但是在我们1000美元的专业版中,我认为它是一个很棒的软件产品,一些人会从中获得价值。毫无疑问,令人印象深刻的是,它们让它看起来非常简单,并且非常容易用自然语言进行改进。我认为这是一个有趣的想法。再次强调,我应该去看看,我相信这是一个很酷的演示。我只是认为,事情变得容易并且易于制作,并不能使其成为一个引人注目的产品。虽然很酷,但我非常好奇实际情况是什么。人们会愿意为一本书支付更多费用,而不是自己阅读或者让它被读给他们,只是可以随便问问题吗?
Here's what I think. I mean, information subscribers are also subscribing to the information for knowledge. I mean, they're doing it to stay ahead, but we're a business publication. And people are querying us to be like, have you written about this company and what's their latest revenue? Is it a good phone? I mean, basically being like the information's.
这是我的想法。我是说,订阅信息的用户也是为了获取知识而订阅信息的。我是说,他们这样做是为了保持领先,但我们是一家商业出版物。人们向我们询问,比如,“你们是否写过关于这个公司的文章,他们最新的收入是多少?这是一部好手机吗?”我的意思是,基本上他们期望我们的信息是准确可靠的。
Yeah, the information search function already, we moved over to like using some GPT based stuff, and you're like, look, and this has always been the obvious idea, which is like, you don't really want the results to be articles, you just want it to be answers in your search function, I think is completely fine, right? And like, reasonable.
是的,信息搜索功能已经有了,我们转而使用了一些基于GPT的东西,你看,而且这一直是一个明显的想法,你并不真的想要结果是文章,而是只希望搜索功能中提供答案,我认为这完全没问题,对吧?而且挺合理的。
I am curious the depth of use on that, right? And the form of the algorithm actually working, right? Also, there's like the product experience. Like, Jess, I'm curious how this will work. It's like, I totally get why you have a search function on your website or in your service. Obviously going from search for a specific article result and then read the article is way shittier than just like the one box answer, right? If you have it in the archive. But my understanding is that is it an API based thing, or are they going to ask you to actually load like an app?
我很好奇它的使用深度,对吧?以及算法的工作形式,对吧?还有类似于产品体验的问题。比如,杰斯,我很好奇这个会如何运作。就好像,我完全明白为什么你的网站或服务上有一个搜索功能。显然,从搜索特定文章的结果然后阅读文章,比起只输出一个简短答案来说要糟糕得多,对吧?如果它已经存在于档案中的话。但我理解的是,这是基于API的还是他们会让你下载一个应用呢?
Okay, it's an API, right? So you just dump it all in and you can pull it back. Yeah. So you end up using it for one box is I do wonder though, like, that's interesting. How does that work with a subscription model? Wouldn't it just be like a paper call type thing, then? Well, you can decide who your this GPT is visible to, right? You could make it private and make it public.
好的,这是一个API,对吧?所以你只需将所有内容丢进去,然后就可以获取回来。是的。所以你最终只用它来处理一个盒子,我倒是有些疑问,那不是很有意思吗?在订阅模式下,这个如何运作呢?难道不只是一种按次付费的方式吗?嗯,你可以决定你的这个GPT对谁可见,对吧?你可以将其设为私密或公开。
So in the same way, like you could have your app, right? But it's saying, can you do this? Can you forget the bot thing where like you want to use the conversational interface? Like, is it set up in the current iteration where it's just like, you can dump your entire archive it and do the one box results on an API? So you just submit the question. I don't know how much you can dump.
所以同样的方式,就像你可以有你自己的应用程序一样,对吧?但是它在问,你能做到这一点吗?你能抛开那种想要使用对话式界面的机器人吗?就是说,当前的版本是否已经设置好,你可以简单地将整个存档上传并通过API获得唯一的结果?所以你只需要提交问题。我不知道你能够上传多少数据。
Yeah. I mean, basically they're going to use this to obviously drive their enterprise business when they make their money. So yeah, I mean, Sam dragged like a speech into it, right? So yeah, Sam, I think to answer Sam's question, I think the, in the, you know, we've been talking about the assistant or I'm sorry, the GPT's functionality, which actually, if you look at their blog, they have two separate blog posts and GPT's is kind of called out as its own thing, which is almost consumer grade, which I think reflects how we're talking about it. And then there's a separate post, which is all of the new APIs and, you know, new models and et cetera.
是的。我是说,基本上当他们赚钱时,他们将显然利用此技术推动他们的企业业务。所以是的,我是说,萨姆附带了一次发言,对吧?所以是的,我认为为了回答萨姆的问题,在我们一直谈论的助理或者我很抱歉,GPT的功能方面,实际上他们在博客上有两篇不同的博文,而GPT被单独列出,几乎可以称为消费级别,这反映了我们的讨论。然后还有一篇独立的博文,介绍了所有的新API和新模型等等。
And there's actually Sam to your question, a new API called the assistance API, which effectively is exactly what you just asked it. You know, it includes, you know, calling that function or that API call gives you access to code interpreter. You can use the retreat, the retrieval engine. So you can augment the assistant with knowledge from outside their models. You can search on proprietary domain data, product information, et cetera, all the kind of stuff you're saying. And then you can also invoke other functions. And so you kind of have like an assistant API that you can call and sign your app.
实际上,有一个名为辅助API的新API,它实际上就是你刚刚所问的。你知道,它包括调用那个函数或API调用来访问代码解释器。你可以使用检索引擎来增强助手的知识,从外部模型获取信息。你可以搜索专有领域数据、产品信息等等,就像你说的那样。你还可以调用其他函数。因此,你可以在你的应用程序中调用助手API。
Like the current way you do this, right, with open AI or really with any of these types of systems, right, is you say, okay, you can't possibly put your entire content library into like the buffer, right, of these chat. There's just no chance, like even if you expanded it, like not even close. But what you do is you dump it all into a graph database, right? And you do it in two steps. Step one is you say, okay, you basically do like a graph, you know, I'm sorry, a vector database, not a graph database. You put it in a vector database. And like you basically do, you do the embedding, you get it back, then you pass the query and you find like the 50 most relevant documents. And then you basically take the 50 most relevant documents and you chuck them into like a GPT style thing, right? And like say, hey, summarize this 50. So basically you do like a crude search to get like the kind of right area you're looking in and you do a narrow one.
你现在这样做的方式是对的,对吧,无论是使用Open AI还是其他类似的系统,你都会说,好的,你不可能把整个内容库放进这些对话的缓存中,根本不可能,即使你扩大了也差得太远。但你可以将所有内容存入图数据库中,这是两步操作。第一步,你可以将其放入一个向量数据库中,不是图数据库。你将其嵌入,然后传递查询并找到最相关的50个文档。然后你将这50个最相关的文档放入像GPT这样的模型中,让它概括这50个文档。所以基本上你进行了一个粗糙的搜索,以获取你感兴趣的正确领域,并进行了一个精细的搜索。
It's obviously better if you don't have to deal with infrastructure. You'll be like, here's everything I know. Like do all that shit for me. I am curious how it's actually working in the background, right? Like, is it just doing that? They just packaged it up in a way that like you as the developer don't have to think about basically doing those two steps, right? Or is that something? I think the answer is yes, Sam. And I think the other thing that they've done is they expanded the context window. So the context window is now 128 K. But that's not a.
很显然,如果不必处理基础设施问题,那会更好。你会像是说,这就是我所了解的一切,然后帮我完成所有那些破事吧。我好奇它在后台是如何运作的,对吧?就像,它只是在那样做吗?他们只是以一种让你作为开发者基本上不必考虑这两个步骤的方式将其打包起来,对吗?或者还有别的什么?我认为答案是肯定的,Sam。我认为他们所做的另一件事是扩大了上下文窗口。所以上下文窗口现在是128K。但这不是一个
To use your example, I mean, 128 K is like 300 pages of text, right? Yeah, but that's not that much. 300 pages of text is great for the for the refined search, right? So like I like take just like the corpus of the information articles, but probably what 10,000 articles just is sure at least. Yeah, whatever it is, right? You can take that. You can say, OK, in bed at all, do the vectorizing stuff, right? Like pull back for 300 pages, the most relevant, I don't know, 30 articles. So what to step back for a sec? Is this a kind of like app store Cambrian moment for this stuff? Like what do you think, Briff?
使用你的例子,我的意思是,128 K就像300页的文本,对吧?是的,但那也不算多。300页的文本对于深入搜索来说是很好的了,对吧?所以我可以只使用信息稿件的语料库,但可能至少有1万篇文章。是的,不管是什么,对吧?你可以拿出来。你可以说,好的,在这个调用里,把向量化的东西做一下,对吧?就像从300页中挑选出最相关的30篇文章。那么,给我们停下来思考一下,这对于这个东西来说是一种类似于应用商店寒武纪时刻吗?你觉得呢,布里夫?
I have two thoughts. One is that I think there continues to be not a meaningful venture play, but like a new job type, at least that will exist in this interim period, where like a lot of non-technical companies, I'm thinking of like all the recipe sites out there or something, right? Like they have technical teams, but like they need an AI interface person to figure out how to start using these things together. BuzzFeed did that and it didn't work, but OK, let's go back to that. I'm just saying like I think this like idea of a prompt engineer, like not even like beyond prompt engineer, like an AI like vertical org and companies going to be important, especially because this is rapidly accelerating so quickly. Like you need someone at the information right now to maybe it's Sam to like tinker with all the stuff and get it working and understand how to use it. And I think like every organization should be asking themselves this question right now.
我有两个想法。第一个想法是,我认为现在存在的并不是一个有意义的创业项目,而更像是一种新的职业类型,至少在这个过渡期间它将存在。比如很多非技术公司,我想到的是像所有的食谱网站之类的。他们有技术团队,但是他们需要一个人工智能界面专家来解决如何将这些东西应用在一起的问题。BuzzFeed尝试过这样做,但没有成功,但是好吧,让我们回到这个问题上来。我只是想说,我认为这种"快速工程师"的想法很重要,不仅仅是"快速工程师",像一个人工智能的垂直机构和公司也很重要,尤其是因为这个领域发展得非常迅速。像现在这样,在信息领域,你需要有人,也许是Sam,来不断调试所有这些东西,并且让其正常运行,并且理解如何使用它。我认为每个组织现在都应该问自己这个问题。
And then the second thing is like I keep thinking about like how does this continue to create an existential threat to Google? Let's pretend that like Bard, you know, doesn't work out as they want and like KPT and open AI continue to excel and exceed and maybe like, you know, 10 years from now the world looks very different. Like how does this impact the private internet and public internet in different ways? If like our first search is actually through an open AI text box or voice chat versus a Google search box, like what are the downstream effects? The former Googler is a good time for a flash poll. Are you more or less Google? Because this is a big question. This is one of the most interesting questions in tech right now, I think.
然后第二个问题是,我一直在思考,这如何继续对谷歌构成一种存在威胁呢?假设像巴德这样的公司无法按照他们的意愿发展,而KPT和OpenAI继续取得卓越成绩,甚至在未来10年,世界看起来截然不同。那么这将以何种方式影响私人互联网和公共互联网呢?如果我们的首次搜索实际上是通过OpenAI的文本框或语音聊天而不是谷歌搜索框,这会产生什么连锁效应呢?作为一名曾经为谷歌工作的人,现在是进行一次小范围调查的好时机。你对谷歌的依赖是增加了还是减少了?这是当前科技领域最有趣的问题之一,我认为。
It's a really hard question. I'm more barred than I am GPT actually. I had a board meeting earlier this week for two days and actually was using Bard throughout the board meeting to like call people out for stats they were talking about or like think quicker than other people in the meeting. And it was like an accelerator for me. Wait a minute. When you go to board meetings, you actually talk, you don't look at Instagram. It's a two day board meeting. And so I had to do a lot of things. You're not a very good venture capitalist. You're not having to see the 12 people. You're actually checking facts. They're texting with 12 other venture capitalists while they're in their board meetings. Exactly. Bard will just has been giving me better results. Maybe it's learning me more. I don't know. But anyway, but then I saw all this new open AI stuff yesterday and I was like, oh shoot, like this is awesome. And so now I'm like leaning more towards open AI and and they're. They're where they're fan bread.
这是一个非常难的问题。实际上,我对此比GPT更有把握。本周早些时候,我参加了一场为期两天的董事会会议,实际上在整个会议中,我一直在使用Bard来反驳他人的数据或者更快地思考问题。对我来说,它就像是一个加速器。等一下,当你参加董事会会议时,你实际上是在交谈,而不是看Instagram。这是一个为期两天的董事会会议。所以我必须做很多事情。你不是一个很好的风险投资家。你不需要见到12个人。你实际上在核实事实。他们在开着董事会会议时,会和其他12个风险投资家发短信。确切地说,Bard给了我更好的结果。也许它正在更多地了解我。我不知道。但是昨天我看到了所有这些新的OpenAI东西,我想,哇,这真棒。所以现在我更倾向于支持OpenAI和他们。他们是值得拥有的东西。
Well, they have the best tech right now. To go back to Jess's question about the Cambrian, like are we at a Cambrian moment? I think the other thing that's happening here is that the GPT-4 Turbo has a is 3x cheaper than GPT-4 per API call. It sounds like which to me, the Cambrian moment question actually rests on the cost question. Right? Yeah. I'm seeing this. I'm seeing just companies that we're seeing come through the door and pitch us on various different, you know, whatever they're building. Everybody's looking at using these APIs in one way or the other. And at the end of the day, the types of use cases you can build for are highly constrained by the prices. And I think nobody's really talking about that brass tax thing, which is that like this thing can only move, you know, as fast as the price goes down. And it's not actually moving down that quickly. And so.
嗯,他们目前拥有最先进的技术。关于杰斯提出的关于寒武纪的问题,我们是否正处于寒武纪时刻?我认为这里正在发生的另一件事是GPT-4 Turbo每次API调用比GPT-4便宜3倍。在我看来,这个寒武纪时刻的问题实际上取决于成本问题。对吗?是的。我看到这一点。我们看到各种公司来找我们推销各种产品,无论他们在构建什么。每个人都在考虑以某种方式使用这些APIs。最终,你可以构建的用例类型受到价格的高度限制。而且我认为没有人真正谈论到这个实际问题,也就是这个东西只能以价格下降的速度前进。而实际上,价格下降的速度并不是那么快。
It's a bit cheaper. I mean, it's also just like you start with the highest values. It's funny. I'm looking at like the blog post. It's like, I don't know. I'm such a less on this. Like it these seem like Fart apps to me, right? Like the laundry buddy. And you have to start like Fart apps. That's where everything starts. No, you don't have to start Fart apps, right? Well, I mean, Sam, Facebook platform started there. Yeah, but Facebook platform is not as big as it did. iPhone started there. The iPhone was all lighters. Was all lighters. This makes me.
这个便宜一点点。我是说,它也就像你从最高价值开始一样。真有趣。我在看博客文章,就像,我也不知道。我在这方面没有太多经验。这些对我来说就像放屁应用一样,对吧?就像洗衣伙伴这样的东西。而且你得像放屁应用一样开始。不,你不一定非得从放屁应用开始对吧?嗯,我是说,Facebook 平台是从那里开始的。是啊,但它现在已经没有那么大了。iPhone也是从那里开始的。iPhone当时只有麦克风。这让我……
I talked about this a little bit last week, but like my current framework on the first big revenue piles for these things are the enterprise world. And so what I think OpenAI does so well, because they've got really good product engineers working with, you know, who also obviously really good researchers. And they just like build these products that people understand that we're talking about, that everyone's talking about. And then it's like some CIO can go to his board and be like, and we're doing this big enterprise partnership or something.
我上周已经稍微谈到了这个问题,但目前我对这些东西的第一大收入来源的框架是以企业世界为基础的。所以我认为OpenAI非常擅长做的是,因为他们有非常优秀的产品工程师,他们与非常优秀的研究人员一起工作。他们构建的产品让人们能够理解我们所谈论的内容,也是每个人都在讨论的内容。然后,某个首席信息官可以向他的董事会报告,说我们正在进行一项重大的企业合作伙伴关系之类的事情。
So there's almost like a marketing element to it. Because it. And when I think they regained with this announcement a little bit, although like I just like, they regained a lot of buzz, right? Like Dave, you were echoing this. Like everyone thinks they're in beast mode, right? And that just means that like, that's like what technical people say about technical teams they admire, right? And a lot of that is marketing. And so then I think the question is like, where it translates. I think it's momentum too. It's not just marketing. It's momentum. It's like it presses the amount of work. Yeah, it's obviously technology behind.
所以这几乎有一个营销元素在其中。因为它。当我想到他们通过这个公告稍微恢复了一些,虽然就像我刚刚提到的,他们恢复了很多话题,对吧?像Dave说的那样。像每个人都认为他们处于“野兽模式”一样,对吧?这就是技术人员对他们钦佩的技术团队所说的。其中很大一部分是营销。然后问题就是,这在哪里有所体现。我认为这也是动力。不仅仅是营销。这是动力。它加大了工作量。是的,显然背后还有技术。
Like they're pushing on tech. I mean, what I always wonder though is, is this a field where we're just one innovation away from something being obsolete, right? Like, is it gonna. You know, to sort of LLMs progress like very linearly and scale and the big guys who have all the money and can get it from the sovereign wealth funds of the world will scale? And that will happen. Yeah, I mean, I think that's been Sam's argument, right?
就像他们在推动科技一样。我的意思是,我一直想知道的是,这是一个我们是否只需要一个创新就能让某样东西变得过时的领域,对吧?就像,它会不会像这样按照非常线性的方式和规模来前进,那些有钱并能从世界上的主权基金中获得资金的大公司将会扩大规模?这将会发生。是的,我想这就是Sam的论点,对吧?
Yeah, what did I just say? At some point, there'll be another paper or another breakthrough and that can come from anywhere. I don't know. I think that I'm an under on that. Like, I think the reality is that we've had a big breakthrough. The LLMs stuff is cool and it is useful. And to be clear, the hey, how the fuck do I play this board game? Turning that from a YouTube video into an Q&A thing until I get the right answer. I want that. Like the onboarding, absolutely. Like, I definitely believe that things that are currently effectively help centers are gonna get way better because of this, right? That's obvious. And I think that the table stakes that'll happen, a lot of people will provide that. Is this a new paradigm? I just don't like less convinced. And I think I'm as extremely skeptical in the, this is some sort of, this is asymptoting. We know what this looks like and it's gonna be useful and it's good. I don't think there's something like major next break through that.
是的,我刚才说了什么?在某个时候,会有另一个论文或者另一个突破,它可能来自任何地方。我不知道。我认为在这件事上,我是持保留态度的。就像我认为现实情况是,我们已经有了一次重大突破。LLMs的东西很酷,也很有用。而且要明确一点,嘿,我怎么玩这个棋盘游戏?把它从YouTube视频变成问答形式,直到我得到正确答案。我想要那样的。就像入门导航一样,绝对需要。我确信由于这次突破,当前有效的帮助中心的服务将会变得更好,这是显而易见的。我认为很多人会提供这种服务。这是一个新的模式吗?对此,我不太确信。我对这种“渐近”的理念非常怀疑。我们知道这看起来是什么样子,它将会是有用的,这是好事。但我不认为会有什么重大的下一次突破。
Did you guys follow XAI? So XAI being Elon Musk's originally anti-woke AI because Elon, who by the way, started OpenAI when it was a nonprofit. Yes, you should be confused. I still don't understand how this was a nonprofit. He actually did start it, but I think he marketed himself as having started it. Does anyone understand how this became a nonprofit into a corporate? We've documented this extensively. They needed money. It still doesn't make sense to me. Had to raise money and they went to the node, Kosla, who put it.
你们有关注XAI吗?XAI是埃隆·马斯克最初反对觉醒运动的人工智能项目,马斯克当时创办了OpenAI,而且当时它还是一个非盈利组织。是的,你可能会感到困惑。我仍然不明白它怎么会成为非盈利组织。事实上,马斯克确实创办了它,但我觉得他宣传自己只是参与创办了它。有人明白它是如何从非盈利组织转变为公司的吗?我们对此进行了详细记录。他们需要资金。这对我来说仍然不合理。他们不得不筹集资金,然后找到了投资者Kosla。
Well, to be clear, they turned it from a nonprofit into a nonprofit. They didn't like to be clear the not, not for profit still exists. It's the shellboard. Plenty of people sit on it. Sam likes to say he serves at the pleasure of the nonprofit. I don't know what that means.
好吧,明确起见,他们将其从一个非盈利机构变成了一个非盈利机构。他们不喜欢明确表示,非盈利机构仍然存在。这只是一个名存实亡的名义机构,很多人只是占着它的名额。Sam喜欢说他是为非盈利机构忠诚服务的。我不知道这是什么意思。
But anyway, we saw the first blog post that asserts some things about the capabilities of XAI, the Musk backed LLM. He's spending a lot of time on this. He's trained on, you guessed it, X is formerly known as Twitter's kind of Corbus. Yeah, which is valuable. It's going to be a wild AI. Yeah, it's going to be crazy. Are the bots embedded? Imagine a health center. It's like, imagine a health center you can talk to whose entire nakorbus of knowledge is Twitter. It's going to be a ball. 50% accurate information.
但无论如何,我们看到了第一篇关于XAI(由马斯克支持的LLM)能力的博客文章。他在这个项目上投入了很多时间。你猜对了,他是在用X,也就是之前被称为Twitter的一种Corbus进行训练。是的,这很有价值。这将会是一种疯狂的人工智能。是的,它将会很疯狂。这些机器人是嵌入式的吗?想象一个健康中心。就好像是,想象一个可以与你交谈、整个知识基于Twitter的健康中心。这将会很有趣。有50%的准确信息。
So it's going to be like the most insane AI ever. And it was hilarious. One of the things I saw somebody built already with this new stuff, which I thought was kind of interesting, is people are taking streaming frames of sports video into the GPT for vision API, and they're then text-to-speeching it. And they're effectively creating sports commentators that are actually quite good. That's one thing that I saw. And then there's this other thing. I'm trying to find it right now. I can't find it. But apparently Google, I think on YouTube with sports and the NFL, they're basically have these little rings that show up underneath the athletes as they move into position before the play starts. And the AI is actually predicting who's going to rush the quarterback better than the announcers right now. And that's kind of interesting.
所以它将成为史上最不可思议的人工智能。而且这真是太好笑了。其中一件我看到有人用这种新技术已经构建起来的有趣事物是,他们将体育视频的流媒体帧输入到GPT视觉API中,然后将其转换为文本朗读出来。他们有效地创造了相当出色的体育解说员。这是我看到的一件事。然后还有另外一件事。我正在努力找到它,但我现在找不到。但显然谷歌,在YouTube上关于体育和美式橄榄球方面,实际上有这些小圆圈显示在运动员的身下,当他们在比赛开始前进入位置时。而人工智能实际上能够相对于解说员更准确地预测谁会更好地冲向四分卫。这挺有意思的。
So here's an AI application that someone should for sure go make a ton of money on. I 100% believe that someone can make a shitload of money building an AI to do so. And then you're like, I'm not going to say that. I'm just going to say that. I'm just going to say that. You know what I mean? That is a marketplace. People do all the sports betting play-by-play games, whatever. I am sure that someone could use AI. And that's an obvious win. That's a very short-term win. Yeah. This isn't going to last forever.
所以,这里有一个人肯定可以赚很多钱的人工智能应用。我百分之百相信有人可以构建一款人工智能应用来做到这一点,赚一大笔钱。然后你说,我不会说那个词。我只会这么说。你知道我的意思吧?那是一个市场。人们玩各种各样的体育赌注游戏。我敢肯定有人可以利用人工智能。这是一个明显的胜利。这是一个短期的胜利。是的。这不会永远持续下去。
How does this change the NFL? I was sitting here watching this and I'm like, okay, are the play collars on the D-line on both sides of the team going to now have to out AI each other to call plays. Because right now you could just watch the YouTube stream and see the AI already is telling you information that people didn't have before. And that's a pretty interesting question. Is that you're asking whether we should be worried that poor sports commentators all 12 of them are going to lose their jobs? I'm not talking about commentator Sam. I'm talking about play collars. I think there are some sports that are safe. There's even fewer. Okay, soccer is a sport that doesn't have plays like football. And I don't think AI can be as useful in soccer and sports betting. So maybe- Oh, come on. You don't think the AI is going to like this person's about to score. Where they're going to pass the ball next? Yeah, you want to bet? But they're like before the game. I'm the one who's a skeptical AI. For sure AI can do that. Yeah, but it's more variable. Soccer players don't have microphones in their ears, right? The way that football gets played.
这会改变美国橄榄球联盟(NFL)的情况吗?我坐在这里看着,心里在想,好吧,现在球队两边的D线上的比赛领导者(play callers)是否需要通过使用人工智能来决定战术。因为现在你可以通过观看YouTube直播,看到人工智能已经提供了以前人们不知道的信息。这是一个非常有趣的问题。你是在问我们是否担心那些可怜的体育评论员(12个)会失业吗?我不是在说体育评论员Sam。我在说的是比赛领导者。我认为有一些体育运动是安全的,但数量更少。好吧,足球是一项没有像美式足球这样的战术的运动。我不认为人工智能在足球和体育博彩中能起到很大的作用。也许-哎呀,别这样说。你不认为人工智能能预测某个人即将得分,或者他们下一步会传球给谁吗?你想打赌吗?但是他们是在比赛之前预测。我才是一个对人工智能表示怀疑的人。当然,人工智能可以做到这一点。但是足球运动员没有像橄榄球那样带着耳麦,这是橄榄球比赛中常见的情况。
I think I'm asking sort of an NFL grade question. Like how does this change the actual rules of the game? I love the idea of AI coaches for sports teams. That would be amazing. Like if you just like we just hired this AI as our coach, we're going to pay $12 million a year. And then all the players are just that is the purest encapsulation of the future I could possibly imagine. Dave was Dion Sanders for Halloween, who's the coach of the Colorado Fos. And now he can really be like AIDN's and AI coach. I love this. That's the name of the app.
我觉得我问的是一个类似于美国职业橄榄球联赛水平的问题。比如,这怎么改变了现实比赛规则?我喜欢有AI教练来指导运动队的想法。那将是令人惊叹的。比如,如果我们刚刚雇佣了一个AI作为我们的教练,年薪1200万美元。然后所有的球员都只是未来最纯粹的体现。戴夫在万圣节当了科罗拉多福斯队的主教练唐·桑德斯,现在他可以真正成为AIDN的AI教练。我喜欢这个。这就是该应用的名字。
I honestly think this might be the place where football mirrors America best. And I think you're totally right. You're just going to have like the people because the whole thing that everyone says, like, oh, people do all the creative work and like AI like does all like the stupid work. No, the opposite. like AI is the manager and people are the doers. And if that happens first in the NFL. So like, you know, you have like the quarterback rebelling against the AI instructions and like, like the whole dynamic of that's going to be hilarious.
我真心认为这可能是足球最好地反映美国的地方。而且我认为你完全正确。你只需要喜欢这些人,因为所有人都说,噢,人类做所有的创造性工作,而AI才做一些愚蠢的工作。不,恰恰相反。AI是经理,而人类是执行者。如果这首先在美式橄榄球中发生。就像,你知道的,你会看到四分卫反对AI的指令,整个动态都会很滑稽。
So go back to which sports are the safer sports then if this is the future of sports. I don't think this kills these sports. But I think this is just like enough words are left unchanged. Like swimming is probably a safer sport. Right? Like it's like, because the AI would just like swim fast. Like you can't really AI swimming. The coaching basically anything where you're trying to beat a clock, you're kind of you're probably okay. It's like, here's a great idea person who I'm coaching as your AI coach is like, why don't you try harder? It's like in the NFL, it'll be like, Hey, you center, why don't you stand on your head? And there's like, do you really want me to stand on my heads? like, yes. Don't worry. Don't have like a real time or real time coaching. But okay.
那么,如果这是体育运动的未来,那我们就回到哪些更安全的运动上去吧。我不认为这会导致这些运动消亡,但我觉得只是足够保留原有的一些要素就可以了。比如游泳可能是一项更安全的运动,对吧?就像是,因为AI只是会游得快而已。你无法真的有AI来游泳。在所有试图打破时间纪录的运动中,你可能是安全的。就像,我作为你的教练,一个很棒的主意是,你的AI教练会说:“为什么不更加努力呢?”就像在NFL(美式足球联盟)里,教练会说:“嘿,你,中锋,为什么不用头站起来?”然后会有这样的反应:“你真的希望我用头站起来吗?”就像是,“是的。”不用担心,不再需要实时或即时教练。但没问题。
So are there any other very exciting promising use cases before we move to non AI? I think sports betting is really interesting. Why wouldn't why would not AI just completely dominate sports betting from now on? I'm sure it will. I think it's AI betting and coaching. The problem is that will people still want to be in the sports betting industry when there's just all machines like it's just well street. I mean, my one of my favorite old sayings is that the singularity already happened on the wall. 100% and it's just 100%. We already are just doing what the machines tell us to do. Yeah.
在我们转向非人工智能领域之前,还有其他非常令人兴奋的有前景的应用吗?我觉得体育博彩非常有趣。为什么不直接让人工智能彻底主导体育博彩呢?我相信它会的。我认为是人工智能在博彩和教练方面的运用。问题是,当整个体育博彩行业都变成了机器,就像是华尔街一样,人们是否仍然愿意从事这个行业呢?我的一个最喜欢的老话就是,智能已经在墙上发生了。我们已经完全听从机器的指示。是的。
So how is this any different? Right. I just I think what it might be is actually the video angle, right, which is like historically like sports like Wall Street, the data streams you're using are generally not like they're like not multimedia. And so if the real thing that AI can do in this moment is take a bunch of like unstructured data from sports and like turn it into enough of a feed, then all of a sudden sports betting is just Wall Street. And like just like day traders can't possibly win against like algorithmic high frequency traders, like you just would expect that AI is going to completely dominate sports betting to the point that like no one will do it for fun.
那么这有什么不同呢?对。我只是觉得可能是因为视频角度不同,历史上体育比赛就像华尔街一样,你使用的数据流通常不是多媒体的。所以,如果AI此时真正的作用是将一堆来自体育比赛的非结构化数据转化为足够的信息源,那么体育博彩就和华尔街一样了。就像日内交易者不可能击败算法高频交易者一样,你会期望AI完全主导体育博彩,以至于没有人再出于娱乐而进行体育博彩。
Right. Because I mean that future seems clearly already here. It's just interesting. It's like I will be like by being too good, it will just destroy the entire sports betting market. Unless you're a swimmer and maybe many others. But does that mean that poker is next? I mean, basically anything where you got this.
没错。因为我的意思是未来似乎已经清晰地到来了。这很有趣。就好像我太优秀了,它将摧毁整个体育博彩市场一样。除非你是一名游泳运动员,或许还有其他很多选项。但这是否意味着接下来就轮到扑克了呢?我是说,基本上任何你有所掌握的游戏都适用。
I thought like people are doing this on screen on like the poker TV channel. I don't know. Yeah, because they show you. Well, I don't think people talk about chess, right? Like computers can be as a chess pretty much no problem with rare exception. People still obviously play chess. In fact, chess is having a renaissance despite that, even though it's kind of futile because it's like we can't like computers just be you anyway. Most of us aren't playing computer. People play computers all the time in chess. Like that's like how they play computers to practice and then they go play other humans. But here's the interesting with sports betting, right? It's like you're not going to figure out like who's using an algorithm versus not, right? And so these game, any game or it's going to be untraceable or unclear where algorithms or like whether AI is playing, whether you're playing as AI or humans, like are going to get really uninteresting for humans because they're just going to lose all the time, right?
我认为人们在电视上像扑克电视频道这样做。我不知道。是的,因为他们会展示给你看。不过我觉得人们不会谈论国际象棋,对吧?电脑几乎可以轻松地击败人类,几乎没有例外。不过人们显然还是继续下棋的。事实上,尽管如此,在国际象棋方面正在出现一种复兴,尽管这有点无谓,因为我们实际上无法战胜电脑。大部分人都不是在与电脑对弈。人们在国际象棋中经常与电脑对弈,以此进行练习,然后再与其他人对弈。但是在体育博彩方面有一个有趣的问题,对吧?你无法弄清楚谁在使用算法,谁没有,对吗?所以这些游戏,无论是哪种,都将难以追踪,或者不清楚算法或者人工智能是否在参与,对于人类来说,这些游戏将变得无聊,因为他们总会输,对吧?
So it's going to be interesting to figure out how to like take a sports marketplace and be like, okay, we're going to be a sports betting, but no AI is allowed and how do you enforce that, right? Like will it end up being that people like get in a room and like, you know, they have it like tech blocking and they're like, this is still fun to do sports betting. We just can't do it online anymore because we can't trust that AI isn't just going to die. And by the way, similar vein casinos are kind of flocked, right? I think that if you're playing Blackjack in Vegas and I've got my like humane thing happening and something in my ear telling me whether I should flip the next card or not. That's kind of, but Britt, I actually think just to push you on that, I think that like casinos already have this problem and they've already kind of like actually solved it by being in person and they just throw everyone out who does that. I think it's actually more online betting as fucked, right?
所以,要弄清楚如何将一个体育市场转变成类似的赛事投注市场,并且规定不允许使用人工智能,这将非常有趣,但如何执行这一规定呢?比如,人们可能会聚在一个房间里,用技术屏蔽器来进行体育博彩。我们仍然可以很享受这种方式,只是不能再在线上进行,因为我们无法保证人工智能不会出问题。顺便说一句,类似情况也发生在赌场,对吧?如果我在拉斯维加斯玩二十一点,然后我有一个耳机里告诉我下一张牌是否应该翻开,这就有点问题了。但是,布里特,我想再追问一下,我认为赌场已经面临这个问题,而且他们通过亲自在场的方式已经解决了这个问题,他们只是把做这种事情的人都赶出去了。我认为问题实际上更多地出现在在线博彩上,对吗?
Basically, you're going to need to create spaces that are human only for things like gambling. What I'm saying is I just think people are going to, I think there are smaller technologies that now exist where it might be harder to catch the people in the casinos. And so I can imagine there are, there's some period of time where they're, they're figuring out how to like play the casinos differently. Sure. So there's going to be bringing down the house 2.0. Yeah. Like there's a, there's a movie here. I think there's a movie plot line. These are some really interesting predictions. I like that. We did our AI versions. We keep coming back and you just can't avoid AI as much as we want to.
基本上,你需要创建仅限人类的空间,比如赌场。我的意思是,我认为人们可能会使用现有的较小技术来在赌场中难以发现。所以我可以想象,他们将会有一段时间来研究如何以不同的方式在赌场中玩。当然了,之后就会有《击败庄家2.0》之类的故事出现。是的,这是一部电影的构思。这些是一些非常有趣的预测。我喜欢这个。我们一直在讨论人工智能的版本,无论我们多想避免,它总是无法回避。
But I want to make sure we also talk about Britt, you were going to like another huge theme right now. We're seeing a lot of deals. We're seeing a lot of M&A. We're seeing a lot of hopefulness, wishful thinking about more deals, more IPOs. Information has a story today about all the ways the bankers screwed up the last IPOs. But what are you hearing, Britt?
但是我希望我们也能谈一谈布里特,你现在将要谈论的是另一个重要主题。我们正在看到很多交易。我们正在看到很多并购。我们正在看到很多希望,对更多交易、更多上市表示期盼的愿望。今天《Information》杂志有一篇关于银行家们如何搞砸上一个IPO的报道。但你听到了什么消息,布里特?
Well, I am hearing a couple of things. One is bankers are still currently advising to not go out for an IPO, but to kind of get your ducks in a row in that the window will potentially reopen spring 2024. I personally- I'm not at all very interested in telling you more. And I'm not so dubious about this. I think it's, I think all of 2024 is going to be pretty difficult. But I do think the M&A markets are going to continue to explode.
好吧,我听到了几件事。首先,银行家们目前仍然建议不要进行首次公开发行(IPO),而是要在2024年春季时井井有条地准备好,因为窗口可能会重新开放。就我个人而言,我对告诉你更多信息不太感兴趣。对于这个问题,我也不太怀疑。我认为整个2024年都会相当困难。但我确实认为并购市场将继续激烈活跃。
What else are the bankers going to say? Like what do you expect them to say? They can say- They're kind of out of a job. The markets are- They can't say go now. I was like, okay, we agree you can't go now. But you should be really ready because we promise that there's going to be a moment coming where we can make money.
银行家还能说什么呢?你期待他们说什么?他们可能会说-他们有点失业了。市场现在不行了-他们不能说现在走。我当时说,好吧,我们同意你们现在不能走。但是你们应该做好准备,因为我们保证会有一个时刻到来,我们可以赚钱。
Yeah. I'm just- I'm reporting from the crowd, from the sidelines. What's happening? I think the only thing that you can- The only thing that you can take signal from that is that even the bankers realize you can't go public now. Yes. And when would your prediction be Sam of when you could go public?
是的。我只是——我只是在人群中、在场外报道。发生了什么?我认为你唯一能够——你唯一能够从中得出信号的是,即使是银行家也意识到现在不适合上市。是的。Sam,根据你的预测,你们何时能够上市?
Never. Never. Like this is- This is ending. It's not that it won't happen. Of course, someone's no public is not going to zero, but the era of lots of IPOs for tech is just done. It turns out that these are pretty shitty companies in general. There are some good companies out there though. Of course there are, but there aren't that many.
从未。从未。就像这样——这就是结束。并不是说它不会发生。当然,某些公司不会完全公开上市,但科技公司大规模公开上市的时代已经过去了。结果证明,总体而言,这些公司都相当糟糕。当然,还有一些好公司存在。当然,有,但并不多。
But will those companies succeed in their IPO? Do you think? Not unless they're incredible because everyone's so burned on these tech IPOs that they're not going to get the benefit of the doubt. The bar for being a good company is way higher than it ever was before because I can just just buy more Microsoft and ride the AI wave that way.
但是那些公司的首次公开募股(IPO)会成功吗?你认为呢?除非它们非常出色,因为每个人对这些科技公司的首次公开募股感到非常失望,他们不太可能得到怀疑的优势。现在一个公司能成为好公司的门槛比以往任何时候都更高,因为我可以直接购买更多的微软股票来乘着人工智能的浪潮。
Yeah. I mean, Sam, you're riding a risk off market, right? And the question is when will we move back to risk on? But I think the thing I understand is even in a risk on market where people are willing to buy more equities, et cetera, in this moment, there's very little reason to buy sub-scale tech companies because the benefits of AI, et cetera, are just going to accrue to the biggest companies. If you want to go risk on, just buy more long-call options on meta, buy more long-call options on Microsoft. There's no upper bound to how they're being in there, clearly going to get all the value from it.
是的。我的意思是,Sam,你正在逃避风险的市场中进行投资,对吗?问题是我们何时转向追求风险回报呢?但我认为我理解的一件事是,即使在一个人们愿意购买更多股票等的追求风险回报的市场中,在此刻购买一些规模较小的科技公司是没有多少理由的,因为人工智能等的好处只会流向最大的公司。如果你想要追求风险回报,只需购买更多关于Meta和微软的认购期权,它们明显将从中获得全部价值,而这没有上限。
I think we have to get over as an industry this idea that there's serious public market demand for $10 billion tech stories. I just don't know. There are a few outliers though. I was tracking, my favorite one of late is Duolingo that I've been digging deeper into, which opened the year at 96. It's now at 167. And it has a $7 billion market cap. It's not the biggest, but just has good numbers. It's a good business. It's showing good. It's fine. I'm just saying, there are success stories that exist out there. It's fine.
我认为作为一个行业,我们必须放弃这种观点,即市场对于价值100亿美元的科技公司有着巨大需求。我不太确定。尽管如此,还是有一些例外的。我最近一直在关注的一家公司是Duolingo,我对它的了解越来越深入,年初它的股价为96,现在涨到了167。它的市值为70亿美元。或许不是最大的,但数字还是不错的。它是一家不错的企业,表现也很好。我只是想说,成功案例还是存在的,没问题。
In 2022, they lost $50 million on revenue of 300. And the fact that they were a little pull off, the growth of the stock is impressive based on some of the numbers. Again, it's fine. Duolingo is not going to go from one to infinity. In fact, it was hilarious. I was talking about a company that was probably threatened by AI. It's like, why on earth would you care about learning other language if I can just auto-translate things on the fly? I'm not saying that people won't do OK. I'm just saying, the story where Duolingo is 10 times bigger than it is is like.
在2022年,他们的收入为300美元,亏损了5000万美元。尽管事实上他们稍微有些落后,但根据一些数字来看,股票的增长仍然令人印象深刻。不过,这没什么大不了的。Duolingo不会一下子从无到有。事实上,这很有意思。我曾经谈论过一家可能受到人工智能威胁的公司。如果我可以随时自动翻译东西,那你为什么还要关心学习其他语言呢?我并不是说人们不能过得好。我只是说,Duolingo十倍于现在的故事就有些夸张了。
Talking about that, I'm just saying that it is possible to be a successful stock right now in the US dollar sub $10 billion. It's all about the expectations though, and what these companies. Duolingo is down. It's basically where it was in July of 2021. When it went. It's basically flat-red IPO. Which is a lot better than a lot of stocks up there right now.
说到这个,我只是想说,现在在美元低于100亿美元的股票中有可能获得成功。然而,这完全取决于市场对这些公司的期望。多邻国(Duolingo)的股价下滑。基本上回到了2021年7月的水平。这是一次平淡的首次公开募股。相比目前市场上的许多股票,这还算不错了。
Totally. But that can't be the bar. No one's buying a public market equity with the expectation that it's going to go down and then get back to where it was. Yeah. You need growth stories. You need consistent good. I'm not arguing. Yeah, yeah. There's going to be another. Does this mean the mega caps become mega mega caps? Yes, definitely.
当然。但这不能成为标准。没有人会购买一只公开市场股权,希望它会下跌然后重新回到原点。是的,你需要成长的故事。你需要持续的好消息。我不是在争论。是的,是的,还会有另外一种情况。这是否意味着大规模企业会变成超大规模企业?是的,绝对是这样。
Yeah. And also, all these stocks are doing a little bit better because the market thinks the Fed is less likely to raise rates. Yeah. And look, the market moves altogether. It's not like these things are totally disconnected. There are always, if you're looking for a marginal hole, places to look.
是的。而且,所有这些股票表现稍好,因为市场认为美联储不太可能加息。是的。而且看,整个市场都在一起动。不是说这些事情完全没有关联。总是有一些地方可以寻找到边际机会。
But look, the story used to be very simple, which is companies could only be so big. It used to be they couldn't have a hundred billion out of company. Now you can't have a trillion dollar company. And there were two reasons for that. One is people just couldn't wrap their heads around anything possibly being that big. And two, the story was pretty simple, which is you get really big, you get less efficient, you look for the next thing where growth is going to come.
但是看,过去这个故事很简单,即企业只能做得那么大。过去,企业不能超过一千亿美元。现在,你也不能拥有万亿美元的企业。这其中有两个原因。一是人们无法理解某事物可能变得如此庞大。第二,这个故事非常简单,即你变得非常大后,效率降低,然后寻找下一个增长点。
The big companies in the era we're in have demonstrated the ability to be massively profitable and continue to grow unbelievably quickly and leverage new technologies well. And I just think that that makes the public market demand certainly for money losing, but even for upscale story, so much less compelling. Because if you want to be risk-on, just buy more of the obvious shit. Why dick around with trying to make 5% more on some weird bet would have risks.
在我们所处的时代,大公司展示了赚钱能力强大、快速增长并善于利用新技术的能力。我认为这使得公开市场对于亏损的需求变得不太吸引人,甚至连高端故事也是如此。因为如果你想冒更大的风险,只需购买更明显的东西即可。何必费力去追求一些带有风险的怪异赌注,只为多赚5%呢?
Again, it doesn't mean you can't build great companies. I actually am totally with you, Britt, that the opportunities for building, I've said there's a bunch like AI, just like cloud before, it creates incredible opportunities for small businesses to do fabulously. If you want to start a company with 10 people and make a ton of money doing something specific, huge opportunities. But I just think this kind of game of how do we build 10, $20 billion public companies that are quote unquote tech stories that are fast growth and people are going to want to buy as public market investors. I just think we're done with that for the foreseeable future.
再说一遍,这并不意味着你不能建立伟大的公司。实际上,我完全支持你,布里特,构建公司的机会很多,就像之前的云计算一样,人工智能就是其中之一,它为小企业创造了难以置信的机遇。如果你想要以10个员工创办一家公司,并通过某种特定的方式赚取大量资金,那么机会是巨大的。但是,我只是认为在可预见的未来,我们已经结束了构建10个、20亿美元的所谓科技故事,其增长速度很快,并且公共市场投资者会希望购买这些的局面。
OK, before we wrap, we've got a lot of territory. Anything else? People just catching people's eyes this week on the mind, in the inbox. I just think we're getting a lot of deal rumors and the information.
好的,在我们结束之前,我们还有很多话题需要讨论。还有其他事情吗?这周有些人的举动引起了大家的注意,这在我们的心里,也在我们的收件箱里。我觉得我们现在听到了很多与交易有关的传闻和信息。
I tell you. All about small businesses. You know, that's my thing now if you hadn't noticed. Yes, all about them. Any new story? How do we fund these small businesses, Sam?
我告诉你,关于小企业的一切。你知道吗,这是我现在致力于的事。是的,关于它们的一切。有任何新闻吗?我们如何为这些小企业提供资金,山姆?
I think that's one of the big questions right now is I've watched this. You want to make some money and we send it to their bank. Well, that's true. But how do you build venture? How does I think there's this is probably a whole other pod, but I think we do actually have to go into this for like an entire pod, which is how do you build a venture capital portfolio that, you know, operates in this world, right? If go ahead. I have feelings and views on this. I want to hear Dave's. You wrote a slide that day. I think that I think we should do an entire pod on this. I just think it's like a very, very long topic. I have a long walk with a venture capitalist tomorrow. I was I was invited to go for a walk with a venture capitalist in my neighborhood tomorrow.
我认为这是现在一个重要的问题之一,我一直在观察这个。你想赚些钱,我们把它汇到他们的银行账户上。嗯,没错。但是如何建立风险投资呢?我认为这可能是一个完全不同的话题,但是我认为我们实际上必须完全探讨一下,比如在这个世界中如何建立一个风险投资组合,对吗?请继续说下去。我有自己的感觉和观点,我想听听Dave的意见。你那天写了一张幻灯片,我认为我们应该全程解释一下。我只是觉得这是一个非常非常长的话题。我明天要和一个风险投资家一起散步,我被邀请明天在我附近和一个风险投资家一起散步。
Wait, wait, wait, wait, wait. Adventure capitalist is going to go and walk. I know. I'm finding you for. I want an adventure capital. So the walk today in our neighborhood. Did you know that I sometimes hear the venture capital happening over there. And I'm like not venture capital in here.
等一下,等一下,等一下,等一下,等一下。冒险资本家要去散步了。我知道。我找到你了。我想要一份冒险资本。所以今天在我们的附近散步。你知道吗,我有时候听到那边发生冒险资本的声音。而我却没有参与其中。
So here's here's the question that an LP asked us. Britain, I heard a meeting. I think I guess it was last week asked us why is it that every single, why is it that all VCs and LPs use one contract in Silicon Valley? Well, they don't like me invented a whole different one for investing in careers. That's true, but that's very rare. And for the most part, the ecosystem uses one contract and there are all of these different forms of risk, right? And everyone goes around pitching all these, you know, limited partners, the same contract. And the each limited partner wants different types of risk. There's all different kinds of institutions.
所以这是一个LP问我们的问题。英国有个会议,我听说是上周,他们问我们为什么每个VC和LP在硅谷都使用同一个合同?嗯,他们不想为投资职业发明一个完全不同的合同。这是真的,但这种情况非常罕见。大部分时候,生态系统都使用一个合同,而风险类型却多种多样。每个有限合伙人都想要不同类型的风险,而且存在各种不同的机构。
And by the way, at the venture capital level, like there's all of these new tools that the AI revolution is putting out into the ether that's that Sam, I think you're very right, are going to empower the small business. They're going to empower the small team of people to make a lot of cash flow, to make really interesting businesses. But that's not what the contract is set up for. Okay. So I think that there's like, I think we have to take a hard look at the ecosystem. And actually we need to be a lot more coordinated about this and work to create more types of contracts that we're all operating on and change the way the conversation is playing out.
顺便说一下,在风险投资层面上,就像AI革命推出的新工具一样,我认为你说得很对,将会给小企业带来力量。它们将帮助小团队赚取大量现金流,创办非常有趣的企业。但是,合同并没有为此设定。所以我认为,我们需要对这个生态系统进行深入研究。实际上,我们需要更加协调一致地行动,并努力创建更多种类的合同来约定我们的活动,并改变当前的对话方式。
This is fun. Let's do this for the next podcast. I think it's a great topic. But just to answer your LPs question, there's a very simple reason, which is scalability, right? Like the last 20 years, at least 10 of venture capital, the whole point is that it's a backwater small industry trying to act like a real financial product. And the way you get standard, its scalability is standardization, right? It makes it easier for there to be a robust market around things, but the easier diligence things, it makes it easier to pile capital in.
这很有趣。我们下一期的播客就做这个主题吧,我觉得它很棒。但是仅仅回答你的合伙人提出的问题,原因就非常简单,那就是可扩展性,对吧?就像过去20年中至少有10年的创业投资一样,整个重点都是要将这个小行业变成真正的金融产品。而要实现标准化就是为了获得可扩展性,对吧?这样可以更容易地建立一个健壮的市场,更轻松地进行尽职调查,更容易向其中注入资本。
But like the reality is we're coming back to the roots and the roots are, this is a niche industry to provide high risk capital to weird situations. And the reality is small businesses, things that can't be public company, 20 billion dollar public companies ever are very poorly served by everyone right now because banks won't lend to them. And there's no capital for them and venture capitalists is not set up for it. And we've done a lot of work. I've done a lot of work. Like, how do you invest in creators? Totally alternative form of finance with totally different structures and rules. How do you think about funding small businesses? What are the exit opportunities? Like, I think this is the jam of what to talk about. So I'm excited about our next episode.
但现实是,我们正在回归根源,而根源就是在一个小众行业为奇特情况提供高风险资本。而现实是,小企业,那些无法成为公共公司的事物,那些永远不会有200亿美元的公共公司,由于银行不愿向他们借贷,它们在现阶段遭受了非常糟糕的服务。而且他们没有资金,风险投资家也没有为他们做好准备。我们已经做了很多工作。我已经做了很多工作。比如,如何投资创作者?这是一种完全不同的金融形式,具有完全不同的结构和规则。你如何思考对小企业的资助?有什么退出机会?我认为这是我们应该讨论的要点。所以我对我们下一期的节目感到兴奋。
I will just say, why are you guys doing that? The information just broke an interesting story that, um, snap laid off 20 product managers, Evan Spiegel. We were just to go back to and that, um, um, the vice president of engineering is leaving. But I remember maybe he liked our episode. I think he liked our founder episode, which is just like the founder flex, you know, he wants to work on product. He wants to take out their jobs. And it's like fire 20. Yeah. I'm like pretty. Well, now I have no idea. I have no idea how prime managers they have, but I'm almost certain that will now work faster and better. Oh my God.
我只想说,你们在干什么呢?这条信息刚刚爆出了一个有趣的新闻,Snap公司解雇了20名产品经理,Evan Spiegel。我们本来打算回去的,而且,嗯,嗯,工程副总裁也要离职了。但我记得他可能喜欢我们的创始人那一集。我觉得他喜欢我们关于创始人的那一集,就像是创始人的炫耀,你知道吗,他想要做产品工作,想要踢走别人的工作。而且他们炒了20人。是啊,我觉得挺不错的。现在我完全不知道他们有多少产品经理,但我几乎可以确定现在他们会工作更快更好。哦天哪。
Okay. Well, R.I. P.M. And great. Well, thank you to everyone for listening. As always, we love doing this for you guys. We love the feedback. We love the shares. We love the ratings. Please give us lots. Subscribe. Um, we have to do these things like contests, right? Like is this what the YouTubers do? Like subscribe and then we'll like take you to a wave pool. We didn't give away. Yeah. Yeah. Yeah. Give away and review review. The ratings are important. Okay. Subscribe and review. Subscribe and we'll quest. We'll give you COVID on the way to the wave pool. Yeah. It would be delightful, but you still get to do your pod. Um, and guys, I wish I could talk more about this way. Pull, but I can't. It's like NDA. It's very exciting. And, um, with that, thank you to everyone. We will be back here next week, apparently talking about these venture capital topics or whatever. I'm all for this. These are my favorite topic we've wondered. I think our listeners are aware. We've got a bunch of new contracts we've developed to. And I have this huge mailing list of like every VC and LP that read this, this slide deck I put out that we can then blast the septemberton. You know, my goal, guys, I want us to do the more or less greatest hits, but we need to also hit the other high notes. So we're going to, we're going to do a mixed bag. We'll go to greatest hits. We'll, we'll get some new things in, but that this is how we evolve. But thank you to everyone.
好的。嗯,R.I.P.M.也很好。非常感谢大家的聆听。一如既往,我们喜欢为大家做这个节目。我们喜欢反馈意见。我们喜欢分享。我们喜欢评级。请给我们很多评级。订阅我们。嗯,我们要像一些竞赛那样做这些事情,对吧?就像这是YouTuber们做的吗?订阅然后我们会带你去一个浪池。我们有奖品派发。对,对,对。派发和评价。评级很重要。好的。订阅和评价。订阅,我们会找你COVID-19在去浪池的路上。是的,这将很愉快,但你还是可以继续做你的播客。嗯,伙计们,我希望我能多谈谈这个浪池,但我不能。就像保密协议一样。这非常令人兴奋。谢谢大家。下周我们会回到这里,显然会谈论一下这些风险投资的话题或其他事情。我完全赞同这个。这是我最喜欢的话题之一。我们已经签了一批新合同。我还有一张巨大的邮件列表,里面有每一个风投和有限合伙人,他们都读过我发表的这个幻灯片,我们接下来可以给他们发邮件。伙计们,我希望我们做一些更多或更少的最佳曲目,但我们也需要涵盖其他高潮部分。我们将混合一些。我们会演奏经典曲目,也会尝试新东西,这就是我们进步的方式。但还是要感谢大家。
Nice to see you, Morin. I also, I forgot to say Dave, a pop of color today. I always tease you just for you, Jess. I always tease you. You've got a hue of information, Rad. Did you guys get the black tie gala? Did you get the invite? Oh, yeah. We'll be in there. We'll be in the new. You see the new logo? You saved the new, the information black tie gala. Oh, there was another black tie gala that we went to, which we're not, we were sworn to secrecy about. We cannot talk about. I just want to show you guys got the invite. We did. And I hope the same thing happens at your black tie gala as what happened at this black tie gala because I don't think so. Can we, can we judge San Francisco by the, the, how well San Francisco is coming back by the number of black tie gala's per week? We have to, you know, I really debated making this one black tie and I'm just going. I think it's good. I think it's good. It's a good era. OK. Zero difference. OK. Fire.
很高兴见到你,莫琳。还有,我忘了说,戴夫,今天加点色彩。我总是逗你玩,因为你是傑斯。我总是逗你玩。你有点信息,辣德。你们参加黑色领带盛会了吗?你们收到邀请了吗?哦,是的,我们会在那里。我们会出现在那个新的场合里。你看到新的标志了吗?你保存了那个新的,信息黑色领带盛会。哦,还有另一个黑色领带盛会,我们参加了,但我们被要求保密,不能谈论。我只是想展示给你们,你们得到了邀请。我们收到了。我希望你们的黑色领带盛会能像这个一样成功,因为我觉得不太可能。我们能通过每周的黑色领带盛会数量来评判旧金山的复苏情况吗?你知道,我真的在考虑让这次活动成为黑色领带,我想这样做挺好的。我觉得很好。这是个好时代。好的,没有差别。好的,真赞。
Well friends. Back to your zoo. Bye.
好了,朋友们。回到你们的动物园去吧。再见。
If you enjoyed this show, please leave us a virtual high five by rating it and reviewing it on Apple podcast, Spotify, YouTube, or wherever you get your podcast. Find more information about each episode in the show notes and follow us on social media by searching for at more or less at Dave Moran at lesson at J lesson. And as for me, I'm at Britt. See you guys next time.
如果你喜欢这个节目,请给我们一个虚拟高五,通过在Apple播客、Spotify、YouTube或你获取播客的任何地方给予评分和评论。在节目说明中查看更多有关每集的信息,并在社交媒体上关注我们,搜索@atmoreorless、@DaveMoran和@Jlesson。至于我,我是Britt。下次见!