4 Comments
User's avatar
Here's what I'm thinking's avatar

Thank you. Very informative.

Expand full comment
Nate Toth's avatar

Investment bubbles don't mean the end of an industry. Dot-com bubble burst didn't kill the internet, obviously. Investment bubbles are just part of the growth cycle of an industry. Also, AI is going to require constant innovation. Your article suggests that innovation isn't as important in Chinese companies as it is in American companies. That's 100% true. You could almost say that innovation is what give American tech companies value. Therefore, it's likely that DeepSeek will be a short lived thing. In addition, without easy/legal access to GPUs, continued growth and innovation will be impossible.

Expand full comment
Here's what I'm thinking's avatar

AI is not .com so probably different type of bubble. Innovation is limited by scaling walls, even with removing the regulation and environmental constraints of the current administration. Problematically the main ways to get an any version of an investment return, and that is not counting decades of labor, uncounted hours, is surveillance, things like porn and "companion" sites and basically all of the fakery that we see for people to plagiarize their way through life working less and bullshitting more.

Expand full comment
Adam Rousselle's avatar

You’re right about bubbles: the crashes are exactly what usually ushers in the most technological development due to reduced barriers to entry. With innovation in China, I’ve seen some there first hand, but there are a lot of hurdles to overcome with something like this. Liang has made comments that say the same thing myself and many Chinese and foreigners living there have said over the years: innovation comes from a culture of curiosity and desire to create. However, fostering things like curiosity and critical thought are especially challenging there, something I’ve also witnessed first hand. Funny enough, Liang’s approach was apparently similar to my own when I worked in commodities research: he wanted fresh graduates instead of seasoned professionals to avoid bringing bad habits to his team. Many technical aspects are still unclear, but it seems likely they did at least some of what they set out to do and deserve credit for that.

Expand full comment