Do you genuinely think Baidu can succeed in autonomous driving? My main criticism of Baidu has always been their poor operational performance. They’re often first movers, yet consistently fumble the execution—just like they did with their large language model, opting for monetization instead of open source, allowing others to dominate.
Considering the intense competition from companies with proven operational strength now entering the autonomous driving market, I’m wondering if Baidu can really pull this off.
Hi. Thanks for your comment - I totally share your skepticism – Baidu’s track record of "pioneering initiatives with shitty execution" does raise doubts about their ability to win protracted battles in this space.
However, I haven't seen any standout achievements from Baidu's direct competitors yet - feel free to correct me if I'm wrong. Their main rivals are autonomous driving service providers especially in Robotaxi sector like Pony or WeRide（car makers tend to target Tesla), and Baidu appears to be moving toward profitability faster than these competitors (assuming their reports are truthful).
Additionally, like Waymo, Baidu has a significant and difficult-to-replicate asset in HD mapping, which requires special licenses in China and substantial investment. While most EV makers—following Tesla —claim they will move away from maps in favor of end-to-end solutions, but it looks hard to achieve in the short-term. I guess a dual solution (HD Map+real time AI) sounds more feasible? — if so, Map is still important.
That is why I believe this is the only thing they can showcase right now—like a durian among garbage—it's not particularly appealing, but it's better than their other nasty offerings.
I absolutely agree with you—I came across the same things when I looked into autonomous driving. That said, this might be more of a personal bias, but I just have a really bad feeling about Baidu. They’ve repeatedly failed to capitalize on early advantages. People made the same arguments about their large language model: first mover, longest in the game, best team. And then they ruined it by charging for access. Honestly, I don’t know anyone still using Ernie. There are so many alternatives now—it feels like it just faded out.
I agree - I feel the same way. nasty behavior, again, has depleted all remaining trust. Autonomous driving is their last chance - if they can't even get this right, the entire company won't be worth a glance anymore (as a former insider, I might be the one who's truly biased, though I didn't realize it at the time). Setting Baidu aside - the autonomous driving industry looks fascinating for the coming years - what do you think?
I completely agree. I was an early adopter of large language models, even before ChatGPT launched. But the one experience that truly amazed me was my first ride in a fully autonomous car—no driver at all. It was absolutely fascinating. I’m now watching this sector very closely.
Thank you so much for the shout out!
High Five
Do you genuinely think Baidu can succeed in autonomous driving? My main criticism of Baidu has always been their poor operational performance. They’re often first movers, yet consistently fumble the execution—just like they did with their large language model, opting for monetization instead of open source, allowing others to dominate.
Considering the intense competition from companies with proven operational strength now entering the autonomous driving market, I’m wondering if Baidu can really pull this off.
Hi. Thanks for your comment - I totally share your skepticism – Baidu’s track record of "pioneering initiatives with shitty execution" does raise doubts about their ability to win protracted battles in this space.
However, I haven't seen any standout achievements from Baidu's direct competitors yet - feel free to correct me if I'm wrong. Their main rivals are autonomous driving service providers especially in Robotaxi sector like Pony or WeRide（car makers tend to target Tesla), and Baidu appears to be moving toward profitability faster than these competitors (assuming their reports are truthful).
Additionally, like Waymo, Baidu has a significant and difficult-to-replicate asset in HD mapping, which requires special licenses in China and substantial investment. While most EV makers—following Tesla —claim they will move away from maps in favor of end-to-end solutions, but it looks hard to achieve in the short-term. I guess a dual solution (HD Map+real time AI) sounds more feasible? — if so, Map is still important.
That is why I believe this is the only thing they can showcase right now—like a durian among garbage—it's not particularly appealing, but it's better than their other nasty offerings.
I absolutely agree with you—I came across the same things when I looked into autonomous driving. That said, this might be more of a personal bias, but I just have a really bad feeling about Baidu. They’ve repeatedly failed to capitalize on early advantages. People made the same arguments about their large language model: first mover, longest in the game, best team. And then they ruined it by charging for access. Honestly, I don’t know anyone still using Ernie. There are so many alternatives now—it feels like it just faded out.
I agree - I feel the same way. nasty behavior, again, has depleted all remaining trust. Autonomous driving is their last chance - if they can't even get this right, the entire company won't be worth a glance anymore (as a former insider, I might be the one who's truly biased, though I didn't realize it at the time). Setting Baidu aside - the autonomous driving industry looks fascinating for the coming years - what do you think?
I completely agree. I was an early adopter of large language models, even before ChatGPT launched. But the one experience that truly amazed me was my first ride in a fully autonomous car—no driver at all. It was absolutely fascinating. I’m now watching this sector very closely.
Cool. I see. Looking forward to reading more about that in ur newsletter 😊