Super thorough and thoughtfully put together. Thank you for sharing these insights.
One line that stayed with me: “Completely disrupting the previous pace and inertia is the first step back to the right track.” Something so universal even across all the innovation driven teams :)
It is. Though I’d argue they don’t have much choice… The pace of this race doesn’t allow for any other option.
From what I’ve seen with big tech companies as someone who worked there, the ones winning have kept their startup gene intact. Easier said than done:) Thanks for the exchange, Tony!
The three-kingdoms angle is solid. But I think ByteDance wins this not because of model quality - they're probably similar -> but because they ship relentlessly. Alibaba and Tencent are powerful, but they're still locked into enterprise bureaucracy.
ByteDance's entire culture is built on speed. Real question: does AI dominance come from better models or from who puts features in front of users first?
For chatbot apps, fast feature iteration can matter more than model quality, it might be the opposite for agent where model capability becomes the core.
Re: bureaucracy, Alibaba’s AI lab is pretty bottom-up with real research freedom. As Alibaba went all-in on AI, they moved fast and poured resources in, that’s a big reason Qwen chatbot scaled so quickly.
Super thorough and thoughtfully put together. Thank you for sharing these insights.
One line that stayed with me: “Completely disrupting the previous pace and inertia is the first step back to the right track.” Something so universal even across all the innovation driven teams :)
It’s amazing to see big tech moves even faster than small startups in LLM race
It is. Though I’d argue they don’t have much choice… The pace of this race doesn’t allow for any other option.
From what I’ve seen with big tech companies as someone who worked there, the ones winning have kept their startup gene intact. Easier said than done:) Thanks for the exchange, Tony!
Great post. I see many adverts for Tencent & Alibaba AI services in airports around SE Asia. They are in a hyper growth phase
Great post. I see many adverts for Tencent & Alibaba AI services in airports around SE Asia. They are in a hyper growth phase
Yes, yes, and YES. I discuss this in detail in my podcast, so feel free to check it out
https://substack.com/@redeyereturns/note/c-213959248?r=5jmutn&utm_source=notes-share-action&utm_medium=web
The three-kingdoms angle is solid. But I think ByteDance wins this not because of model quality - they're probably similar -> but because they ship relentlessly. Alibaba and Tencent are powerful, but they're still locked into enterprise bureaucracy.
ByteDance's entire culture is built on speed. Real question: does AI dominance come from better models or from who puts features in front of users first?
For chatbot apps, fast feature iteration can matter more than model quality, it might be the opposite for agent where model capability becomes the core.
Re: bureaucracy, Alibaba’s AI lab is pretty bottom-up with real research freedom. As Alibaba went all-in on AI, they moved fast and poured resources in, that’s a big reason Qwen chatbot scaled so quickly.