✨ Nora Tian Jiangchuan, Managing Partner at Chuxin Capital, with Zhao Jiehui, CEO of Deepexi.
An episode I have loved recently is finally out! I want to record a few points that affected me deeply.
1️⃣ AI and enterprise. Kun said that AI is a “tool of production,” while data is a “means of production.” Companies need to use AI to bring their internal “dead” data (I mean its state here) back into use: gather data distributed or isolated across different areas—including complex, private, and unstructured material—and let it create value beyond the data itself. (Enterprise Agent = data + model + tools.
2️⃣ Managing upward, corporate governance, and decision-making. Nora said Deepexi has “the best corporate governance of any company I have invested in, without exception,” and that everything promised during fundraising was ultimately delivered. Kun said that the information he shares in executive meetings, company-wide livestreams, and even shareholder meetings has “no substantive differences.” That asset of trust and integrity, that degree of information transparency, the frequent shareholder and board meetings, and investors threatening to “break your legs” when you make a non-consensus decision... While editing the episode, I kept quietly thinking: Kun is incredible!!
3️⃣ Content strategy. “No adjectives” and “no shock-headline style.” Restraint also shows confidence. The rule is so detailed. “Do not borrow someone else's shine.” When introducing the company externally, “do not introduce the founder's background” and “do not introduce the shareholders' backgrounds”... This really moved me. The value lies in the product and service, not in the glow around the founder or investors.
...There is so much to say about this episode. Come listen! 👂
For example, not long ago I came across an article about the inner spirit of technology companies, and this episode shows it directly. There is also Kun's philosophy of building and establishing a B2B business... There is one thing I am most, most, most curious about! How did Kun make those two non-consensus decisions? Can someone dig into this for me? [laughing-crying R]
Oh, I also recently saw a post about Fireworks AI raising some number of millions of dollars. I realized that the two companies have two things in common: B2B and open source. The difference is that Fireworks AI helps companies “run” AI, while Deepexi helps companies “use AI well.”
A long time ago, I had a small thought: “In a lot of work, AI is something that makes a good thing better. Its presence or absence will not change an industry. More precisely, AI will not change certain organizations or institutions.” Organizations and institutions like the ones I described certainly exist. But with companies like Deepexi and Fireworks AI enabling other companies to apply AI and put it into practice, anything is possible!!
Maybe the only difference is how quickly or slowly they get there. 🧐