本集简介
双语字幕
仅展示文本字幕，不包含中文音频；想边听边看，请使用 Bayt 播客 App。
用声音。
用声音。
碰撞世界。
碰撞世界。
生动活泼。
生动活泼。
哈喽，大家好，我是丁教，欢迎来到今天的科技早知道。
哈喽，大家好，我是丁教，欢迎来到今天的科技早知道。
那今天我们想聊一下这个 opc 这个话题。
那今天我们想聊一下这个 opc 这个话题。
ai 时代，很多人都想自己做一家 opc，也就是 one person company，一人公司。
ai 时代，很多人都想自己做一家 opc，也就是 one person company，一人公司。
一个人再加几个 agent 可以做掉过去一整个团队的工作。
一个人再加几个 agent 可以做掉过去一整个团队的工作。
那今天我们的嘉宾是 gumroad 的创始人，他是小而美这本书的作者，然后这本书的英文名字叫做 the minimalist entrepreneurs。
那今天我们的嘉宾是 gumroad 的创始人，他是小而美这本书的作者，然后这本书的英文名字叫做 the minimalist entrepreneurs。
作者叫做 sahil lavingia。
作者叫做 sahil lavingia。
如果说今天大家讨论的这个 opc 啊，是 ai 火了之后才出现的概念的话，那我觉得 sahil 可能可以算得上是这个领域的 og 了。
如果说今天大家讨论的这个 opc 啊，是 ai 火了之后才出现的概念的话，那我觉得 sahil 可能可以算得上是这个领域的 og 了。
gumroad 一开始的时候确实是一家非小而美的公司，他们是完全走了硅谷那种标准的融资路径啊，拿投资、招团队，然后这个追求增长。
gumroad 一开始的时候确实是一家非小而美的公司，他们是完全走了硅谷那种标准的融资路径啊，拿投资、招团队，然后这个追求增长。
但是因为增长的速度啊，不如 vc 的预期，所以他们没有能够融到 b 轮。
但是因为增长的速度啊，不如 vc 的预期，所以他们没有能够融到 b 轮。
然后公司被迫大幅度裁员，从20多人团队然后缩到几个人，甚至是有一段时间其实就剩下了创始人扫修自己。
然后公司被迫大幅度裁员，从20多人团队然后缩到几个人，甚至是有一段时间其实就剩下了创始人扫修自己。
那也就是在这个过程当中，GAMER才变成了一家小并且高自动化、然后持续盈利的公司。
那也就是在这个过程当中，GAMER才变成了一家小并且高自动化、然后持续盈利的公司。
那所以对今天想要做OPC然后独立创业，或者是想用AI放大个人杠杆的这些创业者来说呢？
那所以对今天想要做OPC然后独立创业，或者是想用AI放大个人杠杆的这些创业者来说呢？
呃，他的经验其实还是蛮有参考价值的。
呃，他的经验其实还是蛮有参考价值的。
比如啊，在A镇的帮助下，GAMER作为一个年入2000万美元的公司，他们现在从技术到客服人员是怎么设置的？
比如啊，在A镇的帮助下，GAMER作为一个年入2000万美元的公司，他们现在从技术到客服人员是怎么设置的？
然后如果重来一遍的话，他们会更加重投入市场呢？
然后如果重来一遍的话，他们会更加重投入市场呢？
还是产品？
还是产品？
当然了，也有他在前些年参加马斯克的DOGE，就是政府效率部门之后，他又加入了IRS，就是美国税务局的一些他的思考。
当然了，也有他在前些年参加马斯克的DOGE，就是政府效率部门之后，他又加入了IRS，就是美国税务局的一些他的思考。
如果你想要做opc，嗯，这期节目或许会对你有一些启发。
如果你想要做opc，嗯，这期节目或许会对你有一些启发。
那这期是一个英文的访谈，按照惯例，我还是会一开始把内容用中文跟大家先总结一下。
那这期是一个英文的访谈，按照惯例，我还是会一开始把内容用中文跟大家先总结一下。
嗯，so yeah，i started building stuff when i was a kid，i guess you know，starting with photoshop and web design and had access to those tools and i saw。
嗯，so yeah，i started building stuff when i was a kid，i guess you know，starting with photoshop and web design and had access to those tools and i saw。
那散黑有自己的创业故事，其实还是蛮硅谷的。
那散黑有自己的创业故事，其实还是蛮硅谷的。
他很早就开始做自己的一些小东西，比如说是玩这个PHOTOSHOP，然后做网站。
他很早就开始做自己的一些小东西，比如说是玩这个PHOTOSHOP，然后做网站。
然后后来学习了这个IPHONE的开发，然后在APP STORE上发布了自己的应用等等。
然后后来学习了这个IPHONE的开发，然后在APP STORE上发布了自己的应用等等。
后来PINTEREST创始人看到他的作品之后就邀请他加入团队，他应该是PINTEREST的第二位员工。
后来PINTEREST创始人看到他的作品之后就邀请他加入团队，他应该是PINTEREST的第二位员工。
那个时候他还是在读大学，嗯，只读了一个学期，然后他就决定退学了，然后去加入了PINTEREST。
那个时候他还是在读大学，嗯，只读了一个学期，然后他就决定退学了，然后去加入了PINTEREST。
后来没有过多久啊，大概过了一年多之后吧，他就开始创办了自己的公司，叫做GOMERO。
后来没有过多久啊，大概过了一年多之后吧，他就开始创办了自己的公司，叫做GOMERO。
一开始的时候啊，他也是走的这种非常典型的硅谷的路径，融资、扩张、招人，追求高速度的这样的增长。
一开始的时候啊，他也是走的这种非常典型的硅谷的路径，融资、扩张、招人，追求高速度的这样的增长。
但是呢，GOMERO后来的发展啊，其实是并没有按照VC的预期，它是一个好生意，然后有收入，然后也有用户的增长，但并不是那种指数级爆发的增长。
但是呢，GOMERO后来的发展啊，其实是并没有按照VC的预期，它是一个好生意，然后有收入，然后也有用户的增长，但并不是那种指数级爆发的增长。
于是啊，GOMRO其实后来就裁员了，然后团队缩到很小。
于是啊，GOMRO其实后来就裁员了，然后团队缩到很小。
在又很长一段时间，他就搬到了其他的城市，然后几乎是一个人把公司维持了下来。
在又很长一段时间，他就搬到了其他的城市，然后几乎是一个人把公司维持了下来。
再后来又到了疫情期间，创作者经济爆发了，然后GOMRO其实又迎来了一波增长。
再后来又到了疫情期间，创作者经济爆发了，然后GOMRO其实又迎来了一波增长。
那他们现在仍然是一家年收入将近2000万级别的这样子的一个公司，但他的团队还是相对来说还是蛮小的。
那他们现在仍然是一家年收入将近2000万级别的这样子的一个公司，但他的团队还是相对来说还是蛮小的。
撒黑自己在节目里面说，哈，他们现在核心是三个工程师，然后加上兼职的一个CFO，然后加上兼职的法务，然后还有一点五个客服人员。
撒黑自己在节目里面说，哈，他们现在核心是三个工程师，然后加上兼职的一个CFO，然后加上兼职的法务，然后还有一点五个客服人员。
那很多的客服和运营流程啊，其实都用AI AGENT可以完全处理掉了。
那很多的客服和运营流程啊，其实都用AI AGENT可以完全处理掉了。
这段经历对于他来说还是蛮重要的，因为撒黑并不是说站在我要鼓吹大家全都是去做艺人公司的这个位上讲话。
这段经历对于他来说还是蛮重要的，因为撒黑并不是说站在我要鼓吹大家全都是去做艺人公司的这个位上讲话。
他是真实的经历过这个融资扩张的这个 vc 模式，然后也是经历过这种非常精简，然后以盈利优先，然后自主的这种小公司的模式。
他是真实的经历过这个融资扩张的这个 vc 模式，然后也是经历过这种非常精简，然后以盈利优先，然后自主的这种小公司的模式。
所以啊，当我们现在重新再谈 opc 的时候，他的判断我觉得还是会有一些超越周期的借鉴意义的。
所以啊，当我们现在重新再谈 opc 的时候，他的判断我觉得还是会有一些超越周期的借鉴意义的。
like people say one person company, i would say four people.
like people say one person company, i would say four people.
And the the reason is you want one engineer who's awake twelve hours a day, and then you need another engineer who's awake the other twelve hours of the day, and then you need the same for customer support.
And the the reason is you want one engineer who's awake twelve hours a day, and then you need another engineer who's awake the other twelve hours of the day, and then you need the same for customer support.
这些里面，我觉得有一个非常有意思的点，就是他并不认为一人公司就真的只能有一个人。
这些里面，我觉得有一个非常有意思的点，就是他并不认为一人公司就真的只能有一个人。
大家现在说的OPC啊，很容易想象就是呃，一个人加几个AI AGENT，然后就可以不需要团队了。
大家现在说的OPC啊，很容易想象就是呃，一个人加几个AI AGENT，然后就可以不需要团队了。
但桑秀说，呃，如果真想要给你的用户提供很好的体验的话，可能最后不是一个人，而是四个人。
但桑秀说，呃，如果真想要给你的用户提供很好的体验的话，可能最后不是一个人，而是四个人。
那四个人呢？
那四个人呢？
呃，产品客服各两个。
呃，产品客服各两个。
然后他们俩班倒，那为什么呢？
然后他们俩班倒，那为什么呢？
那比如说是GAMER，现在大量的客服已经是由AI来处理了。
那比如说是GAMER，现在大量的客服已经是由AI来处理了。
呃，他们说当然可能后面完全是可以用AI来代替，但是呢，如果用户发来问题，然后需要一个非常好的用户体验的时候，还是需要人在最后处理一些比较复杂的问题，特别是当AI可能解决不了的一些问题的时候啊，用户不应该等很长的时间。
呃，他们说当然可能后面完全是可以用AI来代替，但是呢，如果用户发来问题，然后需要一个非常好的用户体验的时候，还是需要人在最后处理一些比较复杂的问题，特别是当AI可能解决不了的一些问题的时候啊，用户不应该等很长的时间。
二是应该有一个非常厉害的人类，然后马上进来接手把问题处理掉。
二是应该有一个非常厉害的人类，然后马上进来接手把问题处理掉。
所以啊，他想象的这个极简的公司里面，呃，并不是完全没有人，而是有一个非常小的团队，然后负责产品、工程、研发等等，然后有人负责这个用户和支持。
所以啊，他想象的这个极简的公司里面，呃，并不是完全没有人，而是有一个非常小的团队，然后负责产品、工程、研发等等，然后有人负责这个用户和支持。
THE BUILDING STUFF IS NOT EVEN A DIFFERENTIATOR AT THIS POINT，SO IT'S EVEN MORE ABOUT WHO ARE YOU BUILDING FOR？
THE BUILDING STUFF IS NOT EVEN A DIFFERENTIATOR AT THIS POINT，SO IT'S EVEN MORE ABOUT WHO ARE YOU BUILDING FOR？
FOR AND THEN WHAT PROBLEM ARE YOU SOLVING FOR THEM？
FOR AND THEN WHAT PROBLEM ARE YOU SOLVING FOR THEM？
And then what how are you solving.
And then what how are you solving.
them 第二个我觉得他讲的一个比较重要的点就是 ai 可以帮你写代码，但是不能够帮你找到客户。
them 第二个我觉得他讲的一个比较重要的点就是 ai 可以帮你写代码，但是不能够帮你找到客户。
我问 sahil 现在其实 coding 的门槛越来越低，大家都可以白 code 那让 ai 写一个应用，那是不是小产品其实是反而更难推广，更难卖了？
我问 sahil 现在其实 coding 的门槛越来越低，大家都可以白 code 那让 ai 写一个应用，那是不是小产品其实是反而更难推广，更难卖了？
那以前可能写软件是一种门槛，现在每个人都可以让自己的AGENT去写一款健身追踪啊，比如说是日程管理啊，呃，各种小工具的。
那以前可能写软件是一种门槛，现在每个人都可以让自己的AGENT去写一款健身追踪啊，比如说是日程管理啊，呃，各种小工具的。
那用户为什么还需要买你的？
那用户为什么还需要买你的？
那桑秀他自己的回答也很直接，他说啊，你当然不可能只花15分钟然后让AGENT做一个AI的APP，然后就指望他能赚很多的钱，这个有点太天真了。
那桑秀他自己的回答也很直接，他说啊，你当然不可能只花15分钟然后让AGENT做一个AI的APP，然后就指望他能赚很多的钱，这个有点太天真了。
嗯，如果你真的是懂这个领域、懂这个赛道，然后花一个周末能够做出一个很完整、能够真正解决用户实际问题的这个软件啊。
嗯，如果你真的是懂这个领域、懂这个赛道，然后花一个周末能够做出一个很完整、能够真正解决用户实际问题的这个软件啊。
比如说是给承包商管理排期、发票、库存、合同的工具，稍微复杂一些的，它是仍然可以卖钱的。
比如说是给承包商管理排期、发票、库存、合同的工具，稍微复杂一些的，它是仍然可以卖钱的。
在过去，你可能需要花六个月的时间才能把这样的一款软件做出来，那现在可能一个周末就能做出一版。
在过去，你可能需要花六个月的时间才能把这样的一款软件做出来，那现在可能一个周末就能做出一版。
进行迭代。
进行迭代。
那所以AI真正改变的，其实不是产品自动帮你赚钱，而是它的试错成本变得更低了。
那所以AI真正改变的，其实不是产品自动帮你赚钱，而是它的试错成本变得更低了。
但是最核心的问题始终都没有变化：你到底为谁在做这样的一个产品？
但是最核心的问题始终都没有变化：你到底为谁在做这样的一个产品？
这样的一个产品是否真正地解决了用户的痛点？
这样的一个产品是否真正地解决了用户的痛点？
这也是SAHIU在回看《小而美》这本书的写作的时候说到的一个变化。
这也是SAHIU在回看《小而美》这本书的写作的时候说到的一个变化。
我问他，如果今天要重写这本书的话，他会做出一些什么改动？
我问他，如果今天要重写这本书的话，他会做出一些什么改动？
他回答就是说，他可能会少用COMMUNITY这一个词，因为这个词现在已经被说得有点太多了、太模糊了。
他回答就是说，他可能会少用COMMUNITY这一个词，因为这个词现在已经被说得有点太多了、太模糊了。
可能更直接的说法就是：谁是你的客户？
可能更直接的说法就是：谁是你的客户？
然后你想让谁的生活变得更加美好？
然后你想让谁的生活变得更加美好？
那AI时代呢？
那AI时代呢？
BUILD变得越来越便宜，那真正稀缺的可能就是这样子的一个判断。
BUILD变得越来越便宜，那真正稀缺的可能就是这样子的一个判断。
At the end of the day let's say you have a product that you've built and it's amazing and someone else has a product and it's not even as good as your product but they have humans selling it 第.
At the end of the day let's say you have a product that you've built and it's amazing and someone else has a product and it's not even as good as your product but they have humans selling it 第.
三点我觉得对很多创业者、产品人，或者是像我们这种的内容创作者都会有启发，就是最好的软件不一定是躺赢的，销售啊、故事啊，还有这个产品的人味非常的重要。
三点我觉得对很多创业者、产品人，或者是像我们这种的内容创作者都会有启发，就是最好的软件不一定是躺赢的，销售啊、故事啊，还有这个产品的人味非常的重要。
小秀很坦诚地聊到了当时这个他跟他的竞争对手PATRON的对比。
小秀很坦诚地聊到了当时这个他跟他的竞争对手PATRON的对比。
那刚中很长的一段时间都非常注重产品、注重设计，他自己啊也承认他自己更加喜欢去做产品，不是很喜欢做市场和推销。
那刚中很长的一段时间都非常注重产品、注重设计，他自己啊也承认他自己更加喜欢去做产品，不是很喜欢做市场和推销。
但PATRON在创作者市场里面其实后面做得更大，一个很重要的原因就是他们更擅长做市场营销，更擅长讲故事，然后更懂一些创作者的群体。
但PATRON在创作者市场里面其实后面做得更大，一个很重要的原因就是他们更擅长做市场营销，更擅长讲故事，然后更懂一些创作者的群体。
尤其是和这个YOUTUBE啊、音乐人啊、粉丝经济的那一套。
尤其是和这个YOUTUBE啊、音乐人啊、粉丝经济的那一套。
那这点在AI时代其实也蛮重要的，因为当工具越来越容易做的话，那软件本身也越来越容易被复制。
那这点在AI时代其实也蛮重要的，因为当工具越来越容易做的话，那软件本身也越来越容易被复制。
那真正能拉开差距的，其实不是诶，我也能做一样的功能，而是你是不是理解这一群呃，你的受众的这一群人，你是不是真的懂他们的痛点，然后是不是能够用他们懂得的语言，把产品真正的讲给他们听。
那真正能拉开差距的，其实不是诶，我也能做一样的功能，而是你是不是理解这一群呃，你的受众的这一群人，你是不是真的懂他们的痛点，然后是不是能够用他们懂得的语言，把产品真正的讲给他们听。
是不是能跟这一群人建立信任？
是不是能跟这一群人建立信任？
所以撒又说啊，就是如果是一个技术创始人今天要招一个人的话，他觉得应该是销售。
所以撒又说啊，就是如果是一个技术创始人今天要招一个人的话，他觉得应该是销售。
因为有的时候用户不是在理性的比较两个产品的功能，一个AI自动生成的邮件也许挺好的，但如果有一个真人在认真的听你讲，然后理解你的问题，跟你建立人和人之间的连接的话，仍然可能会赢下这个客户。
因为有的时候用户不是在理性的比较两个产品的功能，一个AI自动生成的邮件也许挺好的，但如果有一个真人在认真的听你讲，然后理解你的问题，跟你建立人和人之间的连接的话，仍然可能会赢下这个客户。
I think that actually more people want to be entrepreneurs they just don't want to run the business right so like.
I think that actually more people want to be entrepreneurs they just don't want to run the business right so like.
Let's say you have an amazing barista.
Let's say you have an amazing barista.
第四个讨论我觉得是非常有意思的，我自己觉得也是蛮开脑洞的，就是关于 ai 会不会接手 ceo 的这个工作，那 sahil 提到了一个还蛮有意思的想象，就是很多人创业其实他们不想经营公司，比如说是很厉害的一个咖啡师，他可能只是享受做咖啡这件事情而已。
第四个讨论我觉得是非常有意思的，我自己觉得也是蛮开脑洞的，就是关于 ai 会不会接手 ceo 的这个工作，那 sahil 提到了一个还蛮有意思的想象，就是很多人创业其实他们不想经营公司，比如说是很厉害的一个咖啡师，他可能只是享受做咖啡这件事情而已。
那比如说是财务、税务、供应链、法律、工资、库存等等这些很繁琐的事情，那。
那比如说是财务、税务、供应链、法律、工资、库存等等这些很繁琐的事情，那。
他其实是不想自己做的。
他其实是不想自己做的。
那未来有没有可能AI变成了CEO，大量处理这种公司运营的功能，而人类可以专注于自己做最擅长的、最有创造力、然后最享受的那一部分工作？
那未来有没有可能AI变成了CEO，大量处理这种公司运营的功能，而人类可以专注于自己做最擅长的、最有创造力、然后最享受的那一部分工作？
那过去的创业其实意味着你不仅只是需要做产品，然后包括变成老板之后要处理的很多鸡毛蒜皮的事情，你要变成管理者、HR、财务负责人、法务接口等等等等。
那过去的创业其实意味着你不仅只是需要做产品，然后包括变成老板之后要处理的很多鸡毛蒜皮的事情，你要变成管理者、HR、财务负责人、法务接口等等等等。
那未来可能会有这样的一种新的形态的出现，就是说你仍然拥有这家公司，但是AI会替你处理大量的这种机器运转的部分。
那未来可能会有这样的一种新的形态的出现，就是说你仍然拥有这家公司，但是AI会替你处理大量的这种机器运转的部分。
你更像是这一个事业里面最核心的创作者或者就是手艺人。
你更像是这一个事业里面最核心的创作者或者就是手艺人。
对，那虽然今天还不能完全实现，但是我觉得它提供了一个蛮有意思的方向吧，就是AI不会让每个人都变成CEO，而是让更多人不被CEO的这些杂物给困住。
对，那虽然今天还不能完全实现，但是我觉得它提供了一个蛮有意思的方向吧，就是AI不会让每个人都变成CEO，而是让更多人不被CEO的这些杂物给困住。
I still work 40 hours for the irs currently but four days a week and weekends and i i think about it.
I still work 40 hours for the irs currently but four days a week and weekends and i i think about it.
a lot 那第五个有意思的部分是撒黑欧最近进入美国政府系统之后的一些观察，他曾经短暂参加过这个 doge 就是政府效率部门，就是马斯克做的那个项目。
a lot 那第五个有意思的部分是撒黑欧最近进入美国政府系统之后的一些观察，他曾经短暂参加过这个 doge 就是政府效率部门，就是马斯克做的那个项目。
他后来又加入了 irs 然后也就是美国的税务局。
他后来又加入了 irs 然后也就是美国的税务局。
那继续做他的这个政府数字化服务的这个相关的工作。
那继续做他的这个政府数字化服务的这个相关的工作。
那这段经历挺特别的，因为他一边在运营 gumroad 然后一边。
那这段经历挺特别的，因为他一边在运营 gumroad 然后一边。
在做这个极其庞大、复杂、非常缓慢的服务所有人的公共系统的这个数字化。
在做这个极其庞大、复杂、非常缓慢的服务所有人的公共系统的这个数字化。
那他在节目里面其实也提到啊，那创业公司和政府最大的区别就是创业公司可以选择客户，但是呃，政府不行，政府必须服务所有人。
那他在节目里面其实也提到啊，那创业公司和政府最大的区别就是创业公司可以选择客户，但是呃，政府不行，政府必须服务所有人。
政府软件难做的地方就是它不仅仅只是要把这个界面做得好看，而是要想怎么样能够。
政府软件难做的地方就是它不仅仅只是要把这个界面做得好看，而是要想怎么样能够。
把原本要打电话、寄信、排队、填纸质表格的这些流程，变成普通人可以在线完成的服务。
把原本要打电话、寄信、排队、填纸质表格的这些流程，变成普通人可以在线完成的服务。
而且还得这个男女老少都能用得上手。
而且还得这个男女老少都能用得上手。
他举了一个例子，在美国过去的这个DMV，也就是美国的这个交管部门吧，过去他的这个线上的体验非常非常的差。
他举了一个例子，在美国过去的这个DMV，也就是美国的这个交管部门吧，过去他的这个线上的体验非常非常的差。
但是这几年很多州通过线上的这个自助服务，然后已经得到了明显的改善。
但是这几年很多州通过线上的这个自助服务，然后已经得到了明显的改善。
所以他希望IRS然后也能够完成类似的转变。
所以他希望IRS然后也能够完成类似的转变。
那这部分让我其实感觉到小而美的讨论，其实不仅仅只是聊创业公司的效率问题。
那这部分让我其实感觉到小而美的讨论，其实不仅仅只是聊创业公司的效率问题。
AI和软件当然可以让小团队更加有杠杆，但它们也可以让那些庞大的、低效的、影响很多人的公共的系统变得更加好用一些。
AI和软件当然可以让小团队更加有杠杆，但它们也可以让那些庞大的、低效的、影响很多人的公共的系统变得更加好用一些。
Like I would be thinking the minimalist entrepreneur，like it is minimalist，but it's the goal isn't like reduction to zero，either。
Like I would be thinking the minimalist entrepreneur，like it is minimalist，but it's the goal isn't like reduction to zero，either。
Like I enjoy，I I want to.
Like I enjoy，I I want to.
I have a company that I actually enjoy working on.
I have a company that I actually enjoy working on.
I don't want to go to zero because then I have to do something else.
I don't want to go to zero because then I have to do something else.
总结一下，这些并不是一个简单的AI时代艺人公司OPC要怎么做的一期节目。
总结一下，这些并不是一个简单的AI时代艺人公司OPC要怎么做的一期节目。
嗯，我觉得我可能有三个TAKEAWAY吧：就是AI让创作者的这个门槛、执行门槛降低了，但是没有降低判断门槛；还是要知道你的客户是谁，然后谁会为你买单。
嗯，我觉得我可能有三个TAKEAWAY吧：就是AI让创作者的这个门槛、执行门槛降低了，但是没有降低判断门槛；还是要知道你的客户是谁，然后谁会为你买单。
那这个其实是没有变的。
那这个其实是没有变的。
那第二个就是AI可以让公司变小，但是人还是非常的重要，特别是销售啊、故事啊。
那第二个就是AI可以让公司变小，但是人还是非常的重要，特别是销售啊、故事啊。
信任呢，用户理解这些东西暂时还是没有办法能够用AI直接去替代。
信任呢，用户理解这些东西暂时还是没有办法能够用AI直接去替代。
甚至是可能如果有人来去做的话，会让你的产品更加这个优于竞争对手的产品。
甚至是可能如果有人来去做的话，会让你的产品更加这个优于竞争对手的产品。
然后第三就是小而美，不是越小越好，也不是最好，就是公司里没人，它反而像是一种选择权。
然后第三就是小而美，不是越小越好，也不是最好，就是公司里没人，它反而像是一种选择权。
嗯，创始人可以决定自己想要多大的公司，然后想要承担多少的管理。
嗯，创始人可以决定自己想要多大的公司，然后想要承担多少的管理。
然后想要把时间花在什么事情上面？
然后想要把时间花在什么事情上面？
我觉得撒黑又不是在鼓吹大家都去创业，也不是说AI会让创业变得更加的轻松。
我觉得撒黑又不是在鼓吹大家都去创业，也不是说AI会让创业变得更加的轻松。
它更像是说，嗯，当工具变得足够强的时候。
它更像是说，嗯，当工具变得足够强的时候。
大家终于有机会去重新选择一种属于自己的工作方式，不是每个人都需要加入这个大公司，然后不是每个公司都需要融资扩张。
大家终于有机会去重新选择一种属于自己的工作方式，不是每个人都需要加入这个大公司，然后不是每个公司都需要融资扩张。
你可以从一个真实的客户开始，从一个能收费的小产品开始，可能这才是小而美在AI时代变得更加重要的原因。
你可以从一个真实的客户开始，从一个能收费的小产品开始，可能这才是小而美在AI时代变得更加重要的原因。
如果你正在思考AI时代的创业，然后个人杠杆啊、自由职业啊，或者只想知道这个小而美这一套东西在今天还有没有意义。
如果你正在思考AI时代的创业，然后个人杠杆啊、自由职业啊，或者只想知道这个小而美这一套东西在今天还有没有意义。
我觉得这一期可能会对你有一些启发。
我觉得这一期可能会对你有一些启发。
那欢迎大家继续收听节目的这个完整版。
那欢迎大家继续收听节目的这个完整版。
那嘉宾是GUMROCK创始人、小而美的作者Sahil Love India。
那嘉宾是GUMROCK创始人、小而美的作者Sahil Love India。
节目开始前，想和大家分享一个消息。
节目开始前，想和大家分享一个消息。
研究显示，到2050年，全球将有一半人会近视，1/10会高度近视。
研究显示，到2050年，全球将有一半人会近视，1/10会高度近视。
现在几乎所有国家都在谈论近视率的上升。
现在几乎所有国家都在谈论近视率的上升。
我们的青少年播客KNOCK KNOCK世界本周就聊到这个话题。
我们的青少年播客KNOCK KNOCK世界本周就聊到这个话题。
那节目除了从科学层面解释近视，还解读了是什么因素让近视变成了一个全世界都重视的公共医疗问题。
那节目除了从科学层面解释近视，还解读了是什么因素让近视变成了一个全世界都重视的公共医疗问题。
感兴趣的朋友，可以点击单集简介中的链接收听。
感兴趣的朋友，可以点击单集简介中的链接收听。
Yeah, yeah, back to your previous experience.
Yeah, yeah, back to your previous experience.
So you worked at Pinterest for a year or so, and then you went straight to your own business.
So you worked at Pinterest for a year or so, and then you went straight to your own business.
So you were 20, and then you raised a ton of VC funding.
So you were 20, and then you raised a ton of VC funding.
Yeah, I raised a million, one point one million dollar seed round February or May or something of the next year.
Yeah, I raised a million, one point one million dollar seed round February or May or something of the next year.
I raised a Series A which was like a seven nine dollar.
I raised a Series A which was like a seven nine dollar.
I think round from planet perkins led that round.
I think round from planet perkins led that round.
That was like the first time i was like, oh, this is really like serious before it was kind of like i'm doing this thing and it's kind of weird but it feels like a reality tv show.
That was like the first time i was like, oh, this is really like serious before it was kind of like i'm doing this thing and it's kind of weird but it feels like a reality tv show.
But then i was like, oh wow, i have a board and a series a and like seven million dollars and a team and an office.
But then i was like, oh wow, i have a board and a series a and like seven million dollars and a team and an office.
Okay, now I'm like locked in like four years or five years or whatnot, and just working on all the time, you know?
Okay, now I'm like locked in like four years or five years or whatnot, and just working on all the time, you know?
Live work and you know kind of experience.
Live work and you know kind of experience.
We still do that today.
We still do that today.
We just do it with a very small team of about five back then.
We just do it with a very small team of about five back then.
We grew to about 25 at peak.
We grew to about 25 at peak.
We were about to run out of money.
We were about to run out of money.
We got a bridge from Kline Perkins $2 million.
We got a bridge from Kline Perkins $2 million.
And so at this point, we'd raised about 10 total.
And so at this point, we'd raised about 10 total.
And we basically had about a year to like, get the profitability.
And we basically had about a year to like, get the profitability.
We failed, we failed to raise a series b around the funding.
We failed, we failed to raise a series b around the funding.
So we basically downsized, we laid everyone off, we went down to three employees and then just down to me.
So we basically downsized, we laid everyone off, we went down to three employees and then just down to me.
And then i basically just ran the business myself.
And then i basically just ran the business myself.
We got rid of the office.
We got rid of the office.
Uh, for about uh five years from like 2015 to Obed basically happening and just kind of break even Skeleton Crew profitable a couple contractors helping out here and there, and then COVID happened and that was our moment of like whoa we're like growing like super fast that was really working like again like all the time for like two years three years.
Uh, for about uh five years from like 2015 to Obed basically happening and just kind of break even Skeleton Crew profitable a couple contractors helping out here and there, and then COVID happened and that was our moment of like whoa we're like growing like super fast that was really working like again like all the time for like two years three years.
It wasn't planned for it was fully remote and then that ended in like 2023 2024, right?
It wasn't planned for it was fully remote and then that ended in like 2023 2024, right?
Interest rates went up.
Interest rates went up.
We increased our pricing.
We increased our pricing.
Now we're like a very nicely profitable business.
Now we're like a very nicely profitable business.
We have great finance function, legal function, all all you need to run a business, you know, i can basically run it.
We have great finance function, legal function, all all you need to run a business, you know, i can basically run it.
A day a week, and then I work at the IRS.
A day a week, and then I work at the IRS.
I still work 40 hours for the IRS currently, but four days a week now.
I still work 40 hours for the IRS currently, but four days a week now.
Like, if you go back to 2015, right?
Like, if you go back to 2015, right?
When I laid everyone off, we got to profitable.
When I laid everyone off, we got to profitable.
We had three people after that.
We had three people after that.
We had like two or three contractors to do support and bug fixes.
We had like two or three contractors to do support and bug fixes.
If that happened today, I would probably not hire those humans.
If that happened today, I would probably not hire those humans.
Right?
Right?
I would.
I would.
Spend more time being like, how can we just have an an ai agent?
Spend more time being like, how can we just have an an ai agent?
Automate these processes that i that someone has to do.
Automate these processes that i that someone has to do.
Whether it's customer support, well, and you can't do it 100%.
Whether it's customer support, well, and you can't do it 100%.
But i think back then you could automate basically 20% of the work and so you had a human do 80% of the work.
But i think back then you could automate basically 20% of the work and so you had a human do 80% of the work.
Uh, it required like four to five hours a day, not eight hours a day, but each of these jobs was like enough where I would not do be able to do it sustainably.
Uh, it required like four to five hours a day, not eight hours a day, but each of these jobs was like enough where I would not do be able to do it sustainably.
I would just get burned out, um, and I would have a job that I would not be able to get away from.
I would just get burned out, um, and I would have a job that I would not be able to get away from.
And but today now you know AI agents would allow for basically probably something like.
And but today now you know AI agents would allow for basically probably something like.
An hour a day or 15 minutes a day and i still have a team, right?
An hour a day or 15 minutes a day and i still have a team, right?
So you can kind of but they're like a lot of the work they do is observing the ai and doing additional product development work and things like that.
So you can kind of but they're like a lot of the work they do is observing the ai and doing additional product development work and things like that.
So for first time founder who want to start a tech company as for now.
So for first time founder who want to start a tech company as for now.
So do you think any job function or hard to replace these days or just not existent?
So do you think any job function or hard to replace these days or just not existent?
As long as you know how to code, you know how to use agent, how you know how to deploy, you know architectural works, and then you can just basically build everything by yourself.
As long as you know how to code, you know how to use agent, how you know how to deploy, you know architectural works, and then you can just basically build everything by yourself.
I mean, I think there's nothing that you can't have an agent do.
I mean, I think there's nothing that you can't have an agent do.
It's more that there are just certain things that are still time-consuming even if you have an agent do the work.
It's more that there are just certain things that are still time-consuming even if you have an agent do the work.
And so if you just have a limited amount of time, you're going to need to hire other humans right?
And so if you just have a limited amount of time, you're going to need to hire other humans right?
So for example.
So for example.
Even if it a lot of like the monthly closed financial book function bookkeeping functions at the company with AI.
Even if it a lot of like the monthly closed financial book function bookkeeping functions at the company with AI.
We still have to pay taxes to a bunch of different countries and states every month, every quarter, every year, sales tax.
We still have to pay taxes to a bunch of different countries and states every month, every quarter, every year, sales tax.
And if those government websites are annoying to work with, it doesn't matter how good your ai tool is that like you're still gonna have to have like a human like you know type in their a name and like check their phone and you know whatever and and they'll have a capture that you know a robot won't be able to solve.
And if those government websites are annoying to work with, it doesn't matter how good your ai tool is that like you're still gonna have to have like a human like you know type in their a name and like check their phone and you know whatever and and they'll have a capture that you know a robot won't be able to solve.
Uh, or you'll have to call a human or who knows what right?
Uh, or you'll have to call a human or who knows what right?
And anything else that you think that maybe short term cannot be replaced by agent fully.
And anything else that you think that maybe short term cannot be replaced by agent fully.
I mean if, if you're doing something purely with its software, I would say the first person you should probably hire would be a salesperson because presumably you're building the product.
I mean if, if you're doing something purely with its software, I would say the first person you should probably hire would be a salesperson because presumably you're building the product.
You're probably doing sales in the beginning, but eventually you're gonna want, even with an infinite amounts of agents, you're still going to be wanting to spend all your time shipping product.
You're probably doing sales in the beginning, but eventually you're gonna want, even with an infinite amounts of agents, you're still going to be wanting to spend all your time shipping product.
And bugs and thinking at least that's my from my bias or my perspective.
And bugs and thinking at least that's my from my bias or my perspective.
So i would probably want to bring in a gum ready.
So i would probably want to bring in a gum ready.
We brought in this guy ryan delk.
We brought in this guy ryan delk.
But similarly, i would bring in somebody who just would focus purely, you know, some super young.
But similarly, i would bring in somebody who just would focus purely, you know, some super young.
Kid who can just.
Kid who can just.
Email creators talk to them all day long, get them to use the product.
Email creators talk to them all day long, get them to use the product.
So sometimes the users still want some like human touch.
So sometimes the users still want some like human touch.
They don't want to be like 'oh I'm talking to a agent again'.
They don't want to be like 'oh I'm talking to a agent again'.
I want some like a.
I want some like a.
human at the end of the day.
human at the end of the day.
Let's say you have a product that you've built and it's amazing.
Let's say you have a product that you've built and it's amazing.
And someone else has a product, and it's not even as good as your product, but they have humans selling it, and you're business, and you get two emails: one is like AI written, it's very good, it solves your problem; but what is human?
And someone else has a product, and it's not even as good as your product, but they have humans selling it, and you're business, and you get two emails: one is like AI written, it's very good, it solves your problem; but what is human?
I do think that like that person that may win the deal, it's almost irrational in a way, right?
I do think that like that person that may win the deal, it's almost irrational in a way, right?
Like an engineer, you may hard time justify, like this doesn't make sense.
Like an engineer, you may hard time justify, like this doesn't make sense.
This is a hard lesson.
This is a hard lesson.
I've learned, i don't even think i've learned it to be honest.
I've learned, i don't even think i've learned it to be honest.
Like this, the best software does not win, right?
Like this, the best software does not win, right?
Like so i focused on building the best software.
Like so i focused on building the best software.
But patreon, you know, is a much larger company.
But patreon, you know, is a much larger company.
I don't think they're soft that good, but they nailed the sales marketing aspect.
I don't think they're soft that good, but they nailed the sales marketing aspect.
I think and now the good because they've been able to hire a lot of great people and they've been able to prove the product and they have ai and all this stuff.
I think and now the good because they've been able to hire a lot of great people and they've been able to prove the product and they have ai and all this stuff.
So now it's like a pretty good product.
So now it's like a pretty good product.
But like when we were competing, i remember being like, why are we losing?
But like when we were competing, i remember being like, why are we losing?
What on?
What on?
Fundamentally different products as well, but... But I do think, like you know, just like leadership wise and stuff, we are very focused on product engineering design.
Fundamentally different products as well, but... But I do think, like you know, just like leadership wise and stuff, we are very focused on product engineering design.
They are very focused on basically marketing and making money.
They are very focused on basically marketing and making money.
creative campaigns like customer outreach and stuff.
creative campaigns like customer outreach and stuff.
Yeah, I think they were way more connected in the YouTube.
Yeah, I think they were way more connected in the YouTube.
They knew that audience and they went after that audience.
They knew that audience and they went after that audience.
Go after that, we didn't even think of that audience as a potential target.
Go after that, we didn't even think of that audience as a potential target.
I think that business model was more aligned with what you know, ultimately worked in the market.
I think that business model was more aligned with what you know, ultimately worked in the market.
Like a sort of more patronage based model of supporting creators and selling digital products.
Like a sort of more patronage based model of supporting creators and selling digital products.
Like selling digital products is also kind of like a millennial thing to do in a way, right?
Like selling digital products is also kind of like a millennial thing to do in a way, right?
Like it's like a.
Like it's like a.
It's a timely thing and people are not as much selling selling digital products as much anymore.
It's a timely thing and people are not as much selling selling digital products as much anymore.
I think last year a lot of a company there are hiring this role like storytellers i think for like vc fun like some tech companies they want this kind of storyteller.
I think last year a lot of a company there are hiring this role like storytellers i think for like vc fun like some tech companies they want this kind of storyteller.
I don't know what do you think about that?
I don't know what do you think about that?
Is that like kind of related to marketing and like creative stuff?
Is that like kind of related to marketing and like creative stuff?
It's just another way of saying content creator, right?
It's just another way of saying content creator, right?
It's funny, there's only so many roles in software businesses, right?
It's funny, there's only so many roles in software businesses, right?
And you could even say, we see for software businesses, media businesses are also software businesses.
And you could even say, we see for software businesses, media businesses are also software businesses.
Like there's the people who build the product.
Like there's the people who build the product.
Software engineers, designers, then there's the people who sell the product, and then there's people who like market the product, and in my opinion, the people who market the product we just keep changing their name in order to hire them.
Software engineers, designers, then there's the people who sell the product, and then there's people who like market the product, and in my opinion, the people who market the product we just keep changing their name in order to hire them.
But they used to call it marketers, communication, those people, PR, influencers, content creators, social media.
But they used to call it marketers, communication, those people, PR, influencers, content creators, social media.
You're just hiring people to generate impressions, and it really is like, doesn't even really matter what they do.
You're just hiring people to generate impressions, and it really is like, doesn't even really matter what they do.
But the truth is, there are two, at least two kinds of people.
But the truth is, there are two, at least two kinds of people.
There's the people who want to talk to the customer, and then the people who want to build a product, and they almost never overlap, right?
There's the people who want to talk to the customer, and then the people who want to build a product, and they almost never overlap, right?
Like founders happen to overlap a little bit, but there's almost always gonna be a strong preference one way or the other.
Like founders happen to overlap a little bit, but there's almost always gonna be a strong preference one way or the other.
I would say like Steve Jobs, my guess is he enjoyed building product, he didn't enjoy talking to customers.
I would say like Steve Jobs, my guess is he enjoyed building product, he didn't enjoy talking to customers.
You know, then you have someone like larry ellison who probably like talking to customers and didn't really care that much about the product.
You know, then you have someone like larry ellison who probably like talking to customers and didn't really care that much about the product.
You can use that, you can even use the products and kind of tell like, okay, this got this works, but it was because they just were really good at sales.
You can use that, you can even use the products and kind of tell like, okay, this got this works, but it was because they just were really good at sales.
So like microsoft, right?
So like microsoft, right?
It's a classic probably example of this.
It's a classic probably example of this.
I care way more about design and product, right?
I care way more about design and product, right?
So, so anyway, i, you know, i think you're always gonna have like gumroad could shrink even more in size, but i think.
So, so anyway, i, you know, i think you're always gonna have like gumroad could shrink even more in size, but i think.
I think maximally we could shrink to effectively four people.
I think maximally we could shrink to effectively four people.
I would say that's like the like people say one person company.
I would say that's like the like people say one person company.
I would say four people and the reason is you want one engineer who's awake, you know, 12 hours a day and then you need another engineer who's awake the other 12 hours in the day and then you need the same for customer support.
I would say four people and the reason is you want one engineer who's awake, you know, 12 hours a day and then you need another engineer who's awake the other 12 hours in the day and then you need the same for customer support.
So if you want, like, to offer it the best possible customer experience, which is like effectively almost like an effectively an instant reply from an AI.
So if you want, like, to offer it the best possible customer experience, which is like effectively almost like an effectively an instant reply from an AI.
If that doesn't solve the issue, a 15-minute really, really diligent, you know, 10 times the compute sort of response.
If that doesn't solve the issue, a 15-minute really, really diligent, you know, 10 times the compute sort of response.
If that doesn't solve the problem, 15 minutes in an hour, you get a human.
If that doesn't solve the problem, 15 minutes in an hour, you get a human.
Who's getting paid?
Who's getting paid?
A significant amount of money, who effectively has nothing else to do, cause everything's been automated, who can just work on your issue, solve the issue, solve the bug.
A significant amount of money, who effectively has nothing else to do, cause everything's been automated, who can just work on your issue, solve the issue, solve the bug.
This bug would normally take two weeks to solve, but they can literally work on it.
This bug would normally take two weeks to solve, but they can literally work on it.
And if they can't finish it, they they structure it so they can hand it off to the person who then works on it.
And if they can't finish it, they they structure it so they can hand it off to the person who then works on it.
That to me is like.
That to me is like.
Yes, i could choose to fire everybody again and just run gumroad myself, right?
Yes, i could choose to fire everybody again and just run gumroad myself, right?
And i would just have to come into the office which i have.
And i would just have to come into the office which i have.
It's five minute walk from my house.
It's five minute walk from my house.
It has water and food and people clean it weekly.
It has water and food and people clean it weekly.
I could come in as internet and come in.
I could come in as internet and come in.
I can work for like 15 minutes a day, do minimal support you but but is that.
I can work for like 15 minutes a day, do minimal support you but but is that.
That like the bright energy right to like minimize.
That like the bright energy right to like minimize.
Like i would be thinking the middle of the entrepreneur.
Like i would be thinking the middle of the entrepreneur.
I want to have a company that i actually enjoy working on.
I want to have a company that i actually enjoy working on.
I don't want to go to zero because then i have to do something else.
I don't want to go to zero because then i have to do something else.
that i think team ferris is like the four four hours.
that i think team ferris is like the four four hours.
Work week and he's goal is not work.
Work week and he's goal is not work.
I think the core takeaway for me is you can choose, you can decide whether you want to pick this way or not.
I think the core takeaway for me is you can choose, you can decide whether you want to pick this way or not.
I think for a lot of my chinese friend who work for like biden's or red nose or like feeling they're kind of they have to keep running.
I think for a lot of my chinese friend who work for like biden's or red nose or like feeling they're kind of they have to keep running.
If i stop everything will be ruined.
If i stop everything will be ruined.
They were like so afraid to like not responding to calls or emails and 10 p.
They were like so afraid to like not responding to calls or emails and 10 p.
m.
m.
They always go back to to bed.
They always go back to to bed.
Go back to their apartment at 12, they look horrible and they feel like they're not healthy.
Go back to their apartment at 12, they look horrible and they feel like they're not healthy.
I just feel like they don't have a choice.
I just feel like they don't have a choice.
Yeah, I mean the truth is, like it feels like you don't have a choice because if you get off the train, it'll continue right?
Yeah, I mean the truth is, like it feels like you don't have a choice because if you get off the train, it'll continue right?
And even if you cash out with $5 million, if you waited one more year, you'd have $10 million, right?
And even if you cash out with $5 million, if you waited one more year, you'd have $10 million, right?
It's sort of a human psychological condition.
It's sort of a human psychological condition.
That like is unresolvable.
That like is unresolvable.
It's it's like it's it is the nature of the industry.
It's it's like it's it is the nature of the industry.
Like sorry you picked an industry in which it does have these crazy multipliers and you know like i saw yesterday like cognition raised a billion dollars and a 26 billion dollar valuation.
Like sorry you picked an industry in which it does have these crazy multipliers and you know like i saw yesterday like cognition raised a billion dollars and a 26 billion dollar valuation.
I was like wow that's crazy a billion dollars raised that's insane but then anthropic just raised 65 billion.
I was like wow that's crazy a billion dollars raised that's insane but then anthropic just raised 65 billion.
you know, it is like incomprehensibly large.
you know, it is like incomprehensibly large.
It is twice as large as the largest IPO in history, and it is still private.
It is twice as large as the largest IPO in history, and it is still private.
It is like hard to comprehend, right?
It is like hard to comprehend, right?
Yeah, so I was just trying to get to the to the point that you said that a lot of people got laid off, and some of them might choose to become entrepreneur, but entrepreneur life seems like very scary for a lot of people.
Yeah, so I was just trying to get to the to the point that you said that a lot of people got laid off, and some of them might choose to become entrepreneur, but entrepreneur life seems like very scary for a lot of people.
There's not a lot of place to learn how to be entrepreneurs, but do you think things will be different by now?
There's not a lot of place to learn how to be entrepreneurs, but do you think things will be different by now?
I could imagine that the AI.
I could imagine that the AI.
TAKES OVER THE ROLE OF THE CEO.
TAKES OVER THE ROLE OF THE CEO.
I THINK THAT ACTUALLY MORE PEOPLE WANT TO BE ENTREPRENEURS.
I THINK THAT ACTUALLY MORE PEOPLE WANT TO BE ENTREPRENEURS.
They just don't want to run the business, right?
They just don't want to run the business, right?
So like, let's say you have an amazing barista, they may want their own coffee shop, right?
So like, let's say you have an amazing barista, they may want their own coffee shop, right?
They may want their own design, their own name, their own location, their own hours.
They may want their own design, their own name, their own location, their own hours.
They want all the money that they would make from that, but they don't want to deal with the distribution stuff.
They want all the money that they would make from that, but they don't want to deal with the distribution stuff.
They don't want to deal... Maybe they'll even do the marketing stuff, but they don't want to deal with.
They don't want to deal... Maybe they'll even do the marketing stuff, but they don't want to deal with.
you know sourcing.
you know sourcing.
They don't want to deal with supply chain management.
They don't want to deal with supply chain management.
They don't want to deal with payroll.
They don't want to deal with payroll.
They don't want to do with taxes.
They don't want to do with taxes.
They don't want to deal with finance.
They don't want to deal with finance.
They don't want to deal with legal.
They don't want to deal with legal.
You could imagine that they use an agent that agent does all is effectively that seal of the company.
You could imagine that they use an agent that agent does all is effectively that seal of the company.
They run the business.
They run the business.
AND THEY EFFECTIVELY ALMOST EMPLOY THE HUMAN, THE HUMAN IS STILL THE OWNER OF THE BUSINESS, RIGHT?
AND THEY EFFECTIVELY ALMOST EMPLOY THE HUMAN, THE HUMAN IS STILL THE OWNER OF THE BUSINESS, RIGHT?
LIKE LEGALLY, THAT'S THE CASE.
LIKE LEGALLY, THAT'S THE CASE.
But it's like almost like i just am the employee.
But it's like almost like i just am the employee.
I am the software engineer of gumroad.
I am the software engineer of gumroad.
I just work on gumroad.
I just work on gumroad.
Ai runs the company.
Ai runs the company.
Ai does the finance, the legal, this executive function, the hiring, the firing, all of that stuff.
Ai does the finance, the legal, this executive function, the hiring, the firing, all of that stuff.
So, you know, i could see a world, you could sort of have an app on your phone that kind of like runs your business.
So, you know, i could see a world, you could sort of have an app on your phone that kind of like runs your business.
and you get to.
and you get to.
Basically unlock do the stuff.
Basically unlock do the stuff.
展开剩余字幕（还有 134 条）
BLOCK EVEN AS THINGS GET MORE FUTURISTIC, YOU COULD IMAGINE A WORLD IN WHICH.
BLOCK EVEN AS THINGS GET MORE FUTURISTIC, YOU COULD IMAGINE A WORLD IN WHICH.
LIKE EVEN.
LIKE EVEN.
Other business partners become automated.
Other business partners become automated.
So, for example, you could imagine that like right now, you you would still have to then like unlock, take all the deliveries.
So, for example, you could imagine that like right now, you you would still have to then like unlock, take all the deliveries.
You know, do all that stuff.
You know, do all that stuff.
But eventually, you could imagine a service, if there's enough demand by this new kind of market, you could imagine a service that comes in.
But eventually, you could imagine a service, if there's enough demand by this new kind of market, you could imagine a service that comes in.
They unlock, they have a key code that unlocks the door.
They unlock, they have a key code that unlocks the door.
They've they or you know your inventory for you.
They've they or you know your inventory for you.
And then you know they take a photo and then ai like confirms it and updates your inventory, right?
And then you know they take a photo and then ai like confirms it and updates your inventory, right?
Or you do some sink on your side when you come into the office and you just take a photo with your app and you just sinks your inventory and it just confirms it just makes sure like it is this here is this year.
Or you do some sink on your side when you come into the office and you just take a photo with your app and you just sinks your inventory and it just confirms it just makes sure like it is this here is this year.
Yep.
Yep.
Yep.
Yep.
Good.
Good.
Good.
Good.
Good.
Good.
Awesome.
Awesome.
And you just do your like, you know, your pre-work checklist and then you start your job and then you have your post-work checklist and then ai is just kind of running the business.
And you just do your like, you know, your pre-work checklist and then you start your job and then you have your post-work checklist and then ai is just kind of running the business.
you know, interesting idea.
you know, interesting idea.
Yeah.
Yeah.
Yeah.
Yeah.
You can just do the work that's most interesting and then ai can take care of all the rest.
You can just do the work that's most interesting and then ai can take care of all the rest.
Yeah.
Yeah.
Like let's say you want to make 20% more money, right?
Like let's say you want to make 20% more money, right?
You can just do it.
You can just do it.
You can just tell your ai.
You can just tell your ai.
Hey, i want to make 20% more money.
Hey, i want to make 20% more money.
It's like cool.
It's like cool.
You want to be open more hours?
You want to be open more hours?
Do you want to do a sale?
Do you want to do a sale?
Do you want to do a brand?
Do you want to do a brand?
Like like these are the ways that at coffee shops make more money when they want like there's some degree of control you have.
Like like these are the ways that at coffee shops make more money when they want like there's some degree of control you have.
Over your income, right?
Over your income, right?
It's not infinite like you don't you can't make but like there's some degree of of control here like an uber you can choose to make more money.
It's not infinite like you don't you can't make but like there's some degree of of control here like an uber you can choose to make more money.
Uh, you can spend money instagram ads, right?
Uh, you can spend money instagram ads, right?
Like there's all sorts of stuff you could choose to do.
Like there's all sorts of stuff you could choose to do.
You could record a tiktok like you could, you could, you could do a loyal at a loyalty program.
You could record a tiktok like you could, you could, you could do a loyal at a loyalty program.
You could do a all sorts of stuff and so yeah, i could i could totally see a world in which, you know, there's an ai that kind of motivates you.
You could do a all sorts of stuff and so yeah, i could i could totally see a world in which, you know, there's an ai that kind of motivates you.
I think we'll experiment at gumroad with these ideas within the context of gumroad.
I think we'll experiment at gumroad with these ideas within the context of gumroad.
At helping creators sort of like automate sale part.
At helping creators sort of like automate sale part.
So anything changed that in this ai agent era, then what you wrote in the books?
So anything changed that in this ai agent era, then what you wrote in the books?
Do you think anything is different these days that you will change if you were write a book?
Do you think anything is different these days that you will change if you were write a book?
Um, i mean, i would say, i mean, philosophically that it's it's similar.
Um, i mean, i would say, i mean, philosophically that it's it's similar.
I would say the biggest thing is that i would focus more on building.
I would say the biggest thing is that i would focus more on building.
Earlier, so I think that when I wrote the book, it was still like high cost to building stuff, so there was a lot of emphasis on like what you wanted to do before you built things, not that much, but like the manual valuable process.
Earlier, so I think that when I wrote the book, it was still like high cost to building stuff, so there was a lot of emphasis on like what you wanted to do before you built things, not that much, but like the manual valuable process.
For example, I would still say it is valid, but you would do that in the context of AI, right?
For example, I would still say it is valid, but you would do that in the context of AI, right?
When you are trying to do the manual valuable process to like eventually automate some.
When you are trying to do the manual valuable process to like eventually automate some.
Problem, now you just basically have your ai do it.
Problem, now you just basically have your ai do it.
So, for example, like this, if we built the thing that automates creator sales, we work on it until we found that there was a creator that was starting to make money.
So, for example, like this, if we built the thing that automates creator sales, we work on it until we found that there was a creator that was starting to make money.
Completely automa- automatically, in an autonomous way, you know basically.
Completely automa- automatically, in an autonomous way, you know basically.
Generating, you know, landing pages, products, a combination of products, landing pages, and emails sent out, cause that's like the way that they reach out to their audience typically, proactively.
Generating, you know, landing pages, products, a combination of products, landing pages, and emails sent out, cause that's like the way that they reach out to their audience typically, proactively.
And we just mess around with that until, you know, in a ma-, very manual way, like just humans like trying stuff, shipping features, messing around with prompts, until it's like, oh, this work.
And we just mess around with that until, you know, in a ma-, very manual way, like just humans like trying stuff, shipping features, messing around with prompts, until it's like, oh, this work.
This email in this way to this subset of the audience framed in this way, which is, you know, and you just kind of have to build that mental map, and then you can turn it into a product feature that's fully automated, right?
This email in this way to this subset of the audience framed in this way, which is, you know, and you just kind of have to build that mental map, and then you can turn it into a product feature that's fully automated, right?
Or they just have a toggle, and they're done.
Or they just have a toggle, and they're done.
That's the goal.
That's the goal.
They shouldn't have to do anything else, cause at at once once there's more than that.
They shouldn't have to do anything else, cause at at once once there's more than that.
The conversion rate, the usage is like 0, right?
The conversion rate, the usage is like 0, right?
Like you, it has to be like... Oh yeah, I'll just turn that on.
Like you, it has to be like... Oh yeah, I'll just turn that on.
And eventually, it's so good that it's like default on, and they have to turn it off.
And eventually, it's so good that it's like default on, and they have to turn it off.
That would be, you know, that's the kind of the ultimate kind of goal.
That would be, you know, that's the kind of the ultimate kind of goal.
Um, but otherwise, like building community and those parts stay the same.
Um, but otherwise, like building community and those parts stay the same.
I, to be honest, I think if I were to rewrite it, I would not use the word 'community' because I just think the word 'community' is like loaded.
I, to be honest, I think if I were to rewrite it, I would not use the word 'community' because I just think the word 'community' is like loaded.
People look like.
People look like.
Think too much about it, like if you... If you want to be a musician, who is your audience?
Think too much about it, like if you... If you want to be a musician, who is your audience?
Who is your listenership?
Who is your listenership?
But this, as soon as you are doing a business, you need an audience.
But this, as soon as you are doing a business, you need an audience.
So that's really what it's about: who is whose life you want to make better.
So that's really what it's about: who is whose life you want to make better.
Because once, once you know who that person is, then you can actually do stuff.
Because once, once you know who that person is, then you can actually do stuff.
Like, who's giving you money for your software?
Like, who's giving you money for your software?
You know.
You know.
Uh, and that it really is.
Uh, and that it really is.
Who are you building for?
Who are you building for?
And then what problem are you solving for them?
And then what problem are you solving for them?
And then what, how are you solving them?
And then what, how are you solving them?
And he's not the right problem.
And he's not the right problem.
Is that the right solution to the problem?
Is that the right solution to the problem?
And the cool thing is now because building can happen so quickly, you should be able to iterate on that fast.
And the cool thing is now because building can happen so quickly, you should be able to iterate on that fast.
Yeah, you can test constantly.
Yeah, you can test constantly.
Yeah.
Yeah.
instead of like, oh okay, i'm gonna spend six months or a year trying this one thing, this one use case.
instead of like, oh okay, i'm gonna spend six months or a year trying this one thing, this one use case.
You know, it's like two weeks, two weeks, two weeks, two weeks to, you know, like you should be able to go much faster.
You know, it's like two weeks, two weeks, two weeks, two weeks to, you know, like you should be able to go much faster.
You don't have revenue in two weeks, new customer, new problem, new solution, every two weeks until you have revenue.
You don't have revenue in two weeks, new customer, new problem, new solution, every two weeks until you have revenue.
Once you have revenue, then you have a 10% chance.
Once you have revenue, then you have a 10% chance.
This becomes a real business.
This becomes a real business.
Yeah, one last question, just my curiosity.
Yeah, one last question, just my curiosity.
So how is it different from working in silicon valley and in for government for rs like.
So how is it different from working in silicon valley and in for government for rs like.
I mean, very different.
I mean, very different.
I would say there's two big differences, or let's say three.
I would say there's two big differences, or let's say three.
One is incentives, like so in startups, you're... It's kind of selfish in a way, right?
One is incentives, like so in startups, you're... It's kind of selfish in a way, right?
Like you're going to get rich.
Like you're going to get rich.
It's a very pure in a way incentive that works very well.
It's a very pure in a way incentive that works very well.
You don't have that in government.
You don't have that in government.
In government, the incentive.
In government, the incentive.
It's hard to explain even honestly like it doesn't really make sense.
It's hard to explain even honestly like it doesn't really make sense.
You don't want to think about it too hard honestly like it just feels good.
You don't want to think about it too hard honestly like it just feels good.
It's status.
It's status.
I don't know like impact more positive impact than i would have private sector.
I don't know like impact more positive impact than i would have private sector.
It is i would say it's more fun personally because it is more challenging for that.
It is i would say it's more fun personally because it is more challenging for that.
Maybe that's why i'm doing it.
Maybe that's why i'm doing it.
But then the other big one is is time duration.
But then the other big one is is time duration.
So.
So.
Startups, you do have a long time duration, but I think with government, it's much more like everyone is thinking about it, and it's just... It's like everything you build is going to exist for 50 years, whether you like it or not, or 100 years.
Startups, you do have a long time duration, but I think with government, it's much more like everyone is thinking about it, and it's just... It's like everything you build is going to exist for 50 years, whether you like it or not, or 100 years.
I mean, you could see with the stamps, right?
I mean, you could see with the stamps, right?
Like things will exist for a long time, so you when you make these decisions, you really.
Like things will exist for a long time, so you when you make these decisions, you really.
Want to make the right decisions, I think that's in startups.
Want to make the right decisions, I think that's in startups.
It's more like move fast, break things, iterate.
It's more like move fast, break things, iterate.
It's a little harder to do that in the government.
It's a little harder to do that in the government.
And then the last thing I would say is that startups get to pick their customer, and so you just get to focus on like rich people who have a problem and you want to solve it, and they give you money and you're off to the races.
And then the last thing I would say is that startups get to pick their customer, and so you just get to focus on like rich people who have a problem and you want to solve it, and they give you money and you're off to the races.
With government, it's like you have to serve 100% of people and it's not about making money.
With government, it's like you have to serve 100% of people and it's not about making money.
It's just about service.
It's just about service.
It's just about delivering the good.
It's just about delivering the good.
Typically, they need to do it.
Typically, they need to do it.
They don't want to do it.
They don't want to do it.
And so it's just about and it's like 100% product market fit, right?
And so it's just about and it's like 100% product market fit, right?
Like everything i build.
Like everything i build.
Gets 100% usage every single person who uses it or want needs to use it.
Gets 100% usage every single person who uses it or want needs to use it.
We'll use it because they need to the customer satisfaction is like 100% because it's like, oh, i was able to automate.
We'll use it because they need to the customer satisfaction is like 100% because it's like, oh, i was able to automate.
This was automated or like this was self-served.
This was automated or like this was self-served.
It's a, it's a, it's a great way of and i would like to believe that by doing this.
It's a, it's a, it's a great way of and i would like to believe that by doing this.
AND THE REAL IMPACT IS, IF I CAN DO THIS SUCCESSFULLY.
AND THE REAL IMPACT IS, IF I CAN DO THIS SUCCESSFULLY.
Then people will have a better image of the government because they're they're experience with the government has improved many many people's experience with the federal government is taxes typically.
Then people will have a better image of the government because they're they're experience with the government has improved many many people's experience with the federal government is taxes typically.
Thank you so much.
Thank you so much.
Yeah, hopefully we can can chat again.
Yeah, hopefully we can can chat again.
Yeah.
Yeah.
thank you for your time.
thank you for your time.
That's good.
That's good.
Thanks so much.
Thanks so much.
Have a good rest of your day.
Have a good rest of your day.
Thank you.
Thank you.
今天的科技早知道就到这里了。
今天的科技早知道就到这里了。
听完之后，如果你有任何的想法，欢迎在评论区和我们。
听完之后，如果你有任何的想法，欢迎在评论区和我们。
交流；如果有任何想听的话题，也请告诉我们。
交流；如果有任何想听的话题，也请告诉我们。
如果你喜欢我们的节目，请记得分享给更多的朋友。
如果你喜欢我们的节目，请记得分享给更多的朋友。
那我们下期再见。
那我们下期再见。
下期再见。
下期再见。
关于 Bayt 播客
Bayt 提供中文+原文双语音频和字幕，帮助你打破语言障碍，轻松听懂全球优质播客。