well in Chinese, BYD's nickname is Xiaodi, little Di, so calling it Didi is not an issue. Of course, I don't know what its English name actually is. I'm just translating the Chinese name directly to English.
Artificial Intelligence thread
- Thread starter 9dashline
- Start date
In Chinese Didi's 滴 is pronounced (dī) while BYD's 迪 is pronounced (dí), so not the same pronunciation or character.
Yeah it might be confusing for English only speakers but nobody cares about them.
News from today on AI PCB related stuff. The expansion across this full supply chain is quite impressive.
On top, I'm just commenting on the ridiculous news out of US on just how many data center projects are collecting dust at the moment.
China is just having huge growth right now in export for AI generated short films and dramas. The margins on the exports are also a lot higher than domestic market.
Meanwhile, the AI short drama industry is facing massive layoffs after platform subsidies end. This industry has experienced the five-year cycle of the EV industry in just one year.
Absolutely Crazy
do u hav article link?
Once DeepSeek came out, it was over in my opinion, simple because of economics of the principle of diminishing returns. In fact, how these LLM are constructed internal, apparently we see the same concept in action where increasing compute power does not lead to more better model, but the efficiency of the output goes down. Hence, it has it own non-economic diminishing returns.
All this means is that once DeepSeek came out the game was over. American AI being in the hands of only a handful of private companies, will only compete among themselves for customers, who probably in the long run the biggest customer will be the US government. While China and could be most of everyone else, will take advantage of open source AI that they control and tweak themselves.
It is too painful for the Americans to admit that. So we get this story of AGI.
AGI my ass. A recent story demonstrates the folly of the AGI god. I saw a recent video that a city in China was using AI with sensors mounted on a top of a car, that would scan the road for defects like potholes. That was unbelievable! They fix potholes in China! Now, when the Americans achieve AGI ... will they fix their potholes!? We already know the answer to that, and we know logically that most of this AGI talk is total bullshit.
Notice how we never hear or read about the word diffusion anymore. No one should be unclear about that.
Disclaimer. Although I believe American pursuit of AGI to be total bullshit, just saying crap to cover their asses, this AI boom is real.
AI will improve software, which humans use to improve productivity. Since the modern economy all runs on software of some sort, this AI boom will last for a while.
If this AI boom is a ten story building, then I would say we are at level 3, while the US stock market is at level six. But still a long way to go to reach level 10 of this building.
By the time the general economy reaches level 10 with AI usage, the American stock market probably be at level 20, while those derivatives at level 300 or more. Haha!
ByteDance's Redfruit has cancelled the guaranteed revenue for AI-simulated human dramas, switching to a pure revenue-sharing model.
"Top-tier companies favored, mid-tier companies squeezed out, bottom-tier companies eliminated."
Strange how nobody shared this here yet, since it's one of the biggest news these days
New benchmark dropped which actually reflect real world coding use cases and imo reflect my real world experience using them.
GPT-5.5 mogs Opus 4.7.
Open-source models way down the list.
Description on how the benchmark works:
New benchmark dropped which actually reflect real world coding use cases and imo reflect my real world experience using them.
GPT-5.5 mogs Opus 4.7.
Open-source models way down the list.
Description on how the benchmark works:
SWE-Bench Pro benchmark is garbage
It’s too early to write off the US AI tech companies . This talk of AI as a fait accompli China won is wrong . Nobody knows how things will play out or which groundbreaking breakthrough will be achieved by either side. So it’s wayyyy too early to claim victory. We don’t even know what will happen tomorrow my CV bless years from nowOnce DeepSeek came out, it was over in my opinion, simple because of economics of the principle of diminishing returns. In fact, how these LLM are constructed internal, apparently we see the same concept in action where increasing compute power does not lead to more better model, but the efficiency of the output goes down. Hence, it has it own non-economic diminishing returns.
All this means is that once DeepSeek came out the game was over. American AI being in the hands of only a handful of private companies, will only compete among themselves for customers, who probably in the long run the biggest customer will be the US government. While China and could be most of everyone else, will take advantage of open source AI that they control and tweak themselves.
It is too painful for the Americans to admit that. So we get this story of AGI.
AGI my ass. A recent story demonstrates the folly of the AGI god. I saw a recent video that a city in China was using AI with sensors mounted on a top of a car, that would scan the road for defects like potholes. That was unbelievable! They fix potholes in China! Now, when the Americans achieve AGI ... will they fix their potholes!? We already know the answer to that, and we know logically that most of this AGI talk is total bullshit.
Notice how we never hear or read about the word diffusion anymore. No one should be unclear about that.
Disclaimer. Although I believe American pursuit of AGI to be total bullshit, just saying crap to cover their asses, this AI boom is real.
AI will improve software, which humans use to improve productivity. Since the modern economy all runs on software of some sort, this AI boom will last for a while.
If this AI boom is a ten story building, then I would say we are at level 3, while the US stock market is at level six. But still a long way to go to reach level 10 of this building.
By the time the general economy reaches level 10 with AI usage, the American stock market probably be at level 20, while those derivatives at level 300 or more. Haha!