藉由提供更好的 AI 工具，
幫助建立社會榜樣，
創造更好的未來。 We build better AI tools,
to set a different kind of example,
and help create a future worth looking forward to.

「熵」在物理裡是失序的量度，一個孤立的系統只會越來越亂。社會也一樣，只是亂起來的樣子比較具體：人不敢相信陌生人，每件事都得自己重新查證；資源明明存在，卻沒有人拿得到；出了意外只能自己扛，因為底下沒有網子接住。

這些事情的反面有個名字，叫社會資本。人與人之間的互信、真的伸手就拿得到的資源、意外發生時接得住人的安全網。它不會自己長出來，只會被人一點一點放進去，或者一點一點抽走。

所以我們用一個很簡單的標準看一個人做的事：他做完之後，這個社會的互信是變多還是變少？拿得到的資源是變多還是變少？下一個遇上麻煩的人，處境是變好還是變壞？

答案是「變多」的，就是我們說的負熵創造者。這跟一般講的成功不是同一件事。有人賺了很多錢，但每一塊都是從別人的信任裡挖出來的；也有人一輩子沒發財，卻讓身邊的人更敢於相信別人。我們想當後者，也想讓後者變得比較容易當。

AI 工具是我們挑的槓桿。互信、資源、安全網，這三件事過去都貴到只有少數人負擔得起。如果工具夠好、夠便宜，它們才有機會變成預設值。

In physics, entropy measures disorder: an isolated system only ever gets messier. Societies can drift the same way, though the disorder takes a more familiar form. People stop trusting strangers and have to verify everything for themselves. Resources exist, yet remain out of reach. When something goes wrong, you have to carry it alone because there is no safety net beneath you.

The opposite of all that has a name: social capital. Trust between people, resources that are genuinely within reach, and a safety net that catches someone when things go wrong. None of it grows on its own. People build it a little at a time, or slowly draw it away.

So we judge what a person does by a simple test. When they are done, is there more trust in the world, or less? Are more resources within reach, or fewer? Is the next person who runs into trouble better off, or worse?

When the answer is “more,” that is the true mark of a creator of negentropy. It has little to do with conventional success. You can build a fortune by exploiting others' trust, or you might never become wealthy, yet leave behind a world where people trust more deeply. We choose the second path—and hope to make it easier for others to walk it too.

AI is the lever we have chosen to pull. For too long, trust, access, and security have been expensive privileges. Make the tools powerful enough and cheap enough, and what was once a privilege can finally become the default.