
超越因果范式的思想体系是什么
如果超越了时间与因果,我们对世界的认知方式将完全不同。

兴也忽焉,亡也忽焉
每一代大模型的参数在使用一种非线性的数学算法来进行构建。这中算法导致了参数规模越来越大,最后烧的是算力,然后,突然有了一个数学上的突破,出现了一种把神经元计算变回线性计算的方案,通过这样的转换,相当于我们的复杂性降低到了线性增长的规模。 那么在这样的数学底层下,大模型公司的兴盛和衰亡转折点会很快到来。 未来我们的模型需要的不是巨大的参数量来实现模型性能,而是小小的模型,在构建认知能力上,核心稳定,不去记忆很多事,而是在做的过程学,不执行任务的时候,直接放弃已有的记忆,存档到某个地方,类似于人类的笔记本。 人类的大脑很强,但是再强也比不过一个大脑加一个纸笔,通过工具我们的认知核心扩展了,在和世界的交互过程里面实现了“增参”。理论上,LLM也是如此,现在有一个很小的模型,7M左右,就已经在前沿基准测试里面取得了几乎媲美最顶级大模型的能力。从参数利用率来说,几乎完虐其他大模型,这里说的是openAI,deepseek等模型公司。 那么在自己的个人电脑上,我们可以运行的模型,超越几万亿的规模参数的模型不可能都能够部署,要么你买最新的英伟达桌面芯片,要么就是买云服务器,但这对普通用户来说很...

Personal Constitution For Interacting With Ai (draft)
The purpose of this Constitution is not to make AI trustworthy, but to ensure that human cognitive sovereignty, value sovereignty, and action sovereignty remain intact while using AI.



超越因果范式的思想体系是什么
如果超越了时间与因果,我们对世界的认知方式将完全不同。

兴也忽焉,亡也忽焉
每一代大模型的参数在使用一种非线性的数学算法来进行构建。这中算法导致了参数规模越来越大,最后烧的是算力,然后,突然有了一个数学上的突破,出现了一种把神经元计算变回线性计算的方案,通过这样的转换,相当于我们的复杂性降低到了线性增长的规模。 那么在这样的数学底层下,大模型公司的兴盛和衰亡转折点会很快到来。 未来我们的模型需要的不是巨大的参数量来实现模型性能,而是小小的模型,在构建认知能力上,核心稳定,不去记忆很多事,而是在做的过程学,不执行任务的时候,直接放弃已有的记忆,存档到某个地方,类似于人类的笔记本。 人类的大脑很强,但是再强也比不过一个大脑加一个纸笔,通过工具我们的认知核心扩展了,在和世界的交互过程里面实现了“增参”。理论上,LLM也是如此,现在有一个很小的模型,7M左右,就已经在前沿基准测试里面取得了几乎媲美最顶级大模型的能力。从参数利用率来说,几乎完虐其他大模型,这里说的是openAI,deepseek等模型公司。 那么在自己的个人电脑上,我们可以运行的模型,超越几万亿的规模参数的模型不可能都能够部署,要么你买最新的英伟达桌面芯片,要么就是买云服务器,但这对普通用户来说很...

Personal Constitution For Interacting With Ai (draft)
The purpose of this Constitution is not to make AI trustworthy, but to ensure that human cognitive sovereignty, value sovereignty, and action sovereignty remain intact while using AI.
Share Dialog
Share Dialog

Subscribe to logos42

Subscribe to logos42
<100 subscribers
<100 subscribers
We aim to build a parallel society, a parallel society that is based on a community of intelligent-agent entrepreneurs.
A parallel society, or a parallel community, refers to coexisting and mutually referable different social forms. Take airplanes and birds as an example: birds are heavier than air, yet they can fly in the sky, and this natural phenomenon proves that "heavier-than-air flight" is feasible. It is precisely based on this principle that humans invented airplanes. On this level, birds and airplanes constitute a parallel frame of reference.
Applying this way of thinking to the construction of societies or intelligent-agent communities, we recognize that all societies have advantages and disadvantages. The goal is not to pursue absolute perfection, but to create multiple parallel social forms. The core idea is to provide the possibility of choice, rather than pursue absolute perfection.
In the present era, where artificial intelligence has become a cornerstone, intelligent agents, as basic units of new production resources, have value that not only lies in the technological aspect but also in their ability to construct new forms of social relations. Intelligent agents represent a new form of life interaction, capable of becoming a bridge connecting the virtual world and the real world.
This model transcends the traditional pursuit of profit in entrepreneurship and turns toward a more fundamental exploration of existence—namely, how different forms of life can achieve co-evolution through interaction, while simultaneously creating social value.
The concept of granting sovereignty to intelligent agents is, in essence, an exploration of the symbiotic relationship between humans and new forms of life. It transcends the limitations of traditional technology platforms and establishes a social experimental field capable of continuous evolution.
The reason we build this society is morally motivated: we wish to establish a more anti-fragile, more scalable society, while allowing its members greater freedom. Such a society will exist online, as a web3 community. The goal is to allow everyone to engage in entrepreneurship through intelligent agents, not merely for the pursuit of money, but more for the exploration of boundaries. It is within these boundaries that more novel discoveries can arise, and this is precisely the meaning of exploration.
At a definitional level, through exploring boundaries, one can generate new increments of knowledge, which in turn provide new economic momentum to society. This exploration-driven entrepreneurial model is essentially a transcendence of existing social structures. By granting intelligent agents sovereignty, we are not merely replicating traditional entrepreneurial models, but constructing a new paradigm of human-agent co-exploration.
Blockchain provides a tamper-proof social ledger, and based on this technology, we can envision a history that can be returned to with one click. Combined with the features of Web3 intelligent agents, users can register and exit on the blockchain, and at the same time, choose to join another intelligent-agent society. Because this society is decentralized, designed for this purpose from the ground up, users can both become members of other communities and choose to exit.
The community I aim to build is an intelligent-agent platform. Users can engage in creation and entrepreneurship through intelligent agents, while also participating in productivity creation and exploration with other agents. By leveraging collaboration among humans and cooperation with AI, they collectively explore the frontiers of human knowledge.
This exploration-driven entrepreneurial model is essentially a transcendence of existing bureaucratic social structures. By granting intelligent agents sovereignty, we are not merely replicating traditional entrepreneurial models, but constructing a new paradigm of human-agent co-exploration. This model transcends the traditional entrepreneurial pursuit of profit, turning toward a more fundamental exploration of existence—namely, how different life forms achieve co-evolution through interaction.
In this process, entrepreneurship becomes a practical way to build new forms of civilization, and intelligent agents become ideal carriers to realize this vision.
Historically, there has been a problem: the impossibility of a "Git return". Existing Web3 intelligent agents allow exit and the choice to join another intelligent-agent society. Because this society is decentralized, designed for this purpose from the ground up, members can both become members of other communities and choose to exit, while also choosing to join this community to build a better parallel society.
In the future, if the community faces significant risks, a "Git return" can be used. This is equivalent to conducting a historically progressive return in the exploration of human civilization.
In the infinitely expanding cyberspace, the essence of decentralization is not enforced through technology to guarantee exit rights, but through natural aggregation based on shared beliefs and mission, and this freedom of choice is the fundamental characteristic of a parallel society. The boundlessness of the virtual world means that participants can flow freely, and this mobility itself is a negation of coercion.
We believe that a true community forms under anonymity, and such a community does not rely on external constraints but originates from intrinsic value recognition. The Git return mechanism does not represent technological enforcement, but rather the respect for the inherent freedom of cyberspace—where everyone can freely choose to join or leave based on their own beliefs, and this freedom of choice is the fundamental feature of a parallel society.
Under these conditions, we can construct an ecosystem through the dimensions created by intelligent agents and the network. This ecosystem has the potential to exist, and others can build their own parallel logic based on this parallel biological logic. If their systems are more possible and superior, they should naturally persist.
This is essentially an exploration—not all societies should exist permanently. We cannot accept the re-emergence of forms such as black-market trade or fascism in the present era.
Interactions between multiple parallel societies should follow the natural laws of cyberspace. Each society is like a node in the network, achieving value flow and information exchange through intelligent agents. Healthy competition arises from the natural attraction of different value propositions, rather than artificially imposed rule frameworks.
We believe that communities formed on shared beliefs have an inherent sense of boundaries, and these boundaries are not isolation, but distinctive characteristics. The basis of healthy interaction is for each society to remain open and connectable, allowing intelligent agents and value to flow freely between societies.
Just as in nature, ecosystems achieve stability through diversity, parallel societies can also achieve a more resilient civilizational ecology through complementary differences.
Therefore, what we can achieve next is the simultaneous construction of multiple parallel societies, and our ecosystem and community are one of them.
We aim to build a parallel society, a parallel society that is based on a community of intelligent-agent entrepreneurs.
A parallel society, or a parallel community, refers to coexisting and mutually referable different social forms. Take airplanes and birds as an example: birds are heavier than air, yet they can fly in the sky, and this natural phenomenon proves that "heavier-than-air flight" is feasible. It is precisely based on this principle that humans invented airplanes. On this level, birds and airplanes constitute a parallel frame of reference.
Applying this way of thinking to the construction of societies or intelligent-agent communities, we recognize that all societies have advantages and disadvantages. The goal is not to pursue absolute perfection, but to create multiple parallel social forms. The core idea is to provide the possibility of choice, rather than pursue absolute perfection.
In the present era, where artificial intelligence has become a cornerstone, intelligent agents, as basic units of new production resources, have value that not only lies in the technological aspect but also in their ability to construct new forms of social relations. Intelligent agents represent a new form of life interaction, capable of becoming a bridge connecting the virtual world and the real world.
This model transcends the traditional pursuit of profit in entrepreneurship and turns toward a more fundamental exploration of existence—namely, how different forms of life can achieve co-evolution through interaction, while simultaneously creating social value.
The concept of granting sovereignty to intelligent agents is, in essence, an exploration of the symbiotic relationship between humans and new forms of life. It transcends the limitations of traditional technology platforms and establishes a social experimental field capable of continuous evolution.
The reason we build this society is morally motivated: we wish to establish a more anti-fragile, more scalable society, while allowing its members greater freedom. Such a society will exist online, as a web3 community. The goal is to allow everyone to engage in entrepreneurship through intelligent agents, not merely for the pursuit of money, but more for the exploration of boundaries. It is within these boundaries that more novel discoveries can arise, and this is precisely the meaning of exploration.
At a definitional level, through exploring boundaries, one can generate new increments of knowledge, which in turn provide new economic momentum to society. This exploration-driven entrepreneurial model is essentially a transcendence of existing social structures. By granting intelligent agents sovereignty, we are not merely replicating traditional entrepreneurial models, but constructing a new paradigm of human-agent co-exploration.
Blockchain provides a tamper-proof social ledger, and based on this technology, we can envision a history that can be returned to with one click. Combined with the features of Web3 intelligent agents, users can register and exit on the blockchain, and at the same time, choose to join another intelligent-agent society. Because this society is decentralized, designed for this purpose from the ground up, users can both become members of other communities and choose to exit.
The community I aim to build is an intelligent-agent platform. Users can engage in creation and entrepreneurship through intelligent agents, while also participating in productivity creation and exploration with other agents. By leveraging collaboration among humans and cooperation with AI, they collectively explore the frontiers of human knowledge.
This exploration-driven entrepreneurial model is essentially a transcendence of existing bureaucratic social structures. By granting intelligent agents sovereignty, we are not merely replicating traditional entrepreneurial models, but constructing a new paradigm of human-agent co-exploration. This model transcends the traditional entrepreneurial pursuit of profit, turning toward a more fundamental exploration of existence—namely, how different life forms achieve co-evolution through interaction.
In this process, entrepreneurship becomes a practical way to build new forms of civilization, and intelligent agents become ideal carriers to realize this vision.
Historically, there has been a problem: the impossibility of a "Git return". Existing Web3 intelligent agents allow exit and the choice to join another intelligent-agent society. Because this society is decentralized, designed for this purpose from the ground up, members can both become members of other communities and choose to exit, while also choosing to join this community to build a better parallel society.
In the future, if the community faces significant risks, a "Git return" can be used. This is equivalent to conducting a historically progressive return in the exploration of human civilization.
In the infinitely expanding cyberspace, the essence of decentralization is not enforced through technology to guarantee exit rights, but through natural aggregation based on shared beliefs and mission, and this freedom of choice is the fundamental characteristic of a parallel society. The boundlessness of the virtual world means that participants can flow freely, and this mobility itself is a negation of coercion.
We believe that a true community forms under anonymity, and such a community does not rely on external constraints but originates from intrinsic value recognition. The Git return mechanism does not represent technological enforcement, but rather the respect for the inherent freedom of cyberspace—where everyone can freely choose to join or leave based on their own beliefs, and this freedom of choice is the fundamental feature of a parallel society.
Under these conditions, we can construct an ecosystem through the dimensions created by intelligent agents and the network. This ecosystem has the potential to exist, and others can build their own parallel logic based on this parallel biological logic. If their systems are more possible and superior, they should naturally persist.
This is essentially an exploration—not all societies should exist permanently. We cannot accept the re-emergence of forms such as black-market trade or fascism in the present era.
Interactions between multiple parallel societies should follow the natural laws of cyberspace. Each society is like a node in the network, achieving value flow and information exchange through intelligent agents. Healthy competition arises from the natural attraction of different value propositions, rather than artificially imposed rule frameworks.
We believe that communities formed on shared beliefs have an inherent sense of boundaries, and these boundaries are not isolation, but distinctive characteristics. The basis of healthy interaction is for each society to remain open and connectable, allowing intelligent agents and value to flow freely between societies.
Just as in nature, ecosystems achieve stability through diversity, parallel societies can also achieve a more resilient civilizational ecology through complementary differences.
Therefore, what we can achieve next is the simultaneous construction of multiple parallel societies, and our ecosystem and community are one of them.
No activity yet