Artificial Intelligence thread

L

Legume7

New Member
Registered Member
Kimi k1.6-IOI-high from Moonshot AI based in Beijing is now the top coding model on LiveCodeBench, beating out O1-high. The model has not been released to the public, but k1.5 was very impressive, just overshadowed by R1. k1.5 is also closed source, but they released a detailed technical report
Please, Log in or Register to view URLs content!
for those interested. Look forward to testing k1.6 when it becomes fully available.GkspMuMXUAAANjN.jpg

Edit: RL lead at Moonshot AI confirms that k1.6-IOI will be coming soon:
https://twitter.com/i/web/status/1894581740968317133
 
Last edited:
A

AndrewS

Brigadier
Registered Member
Wrought said:
I agree that it does make sense, but that's a political rather than economic conclusion. Same reason why industrial policies exist. Similarly, global demand can also be artificially constrained by the same politics (i.e. US restrictions).
Click to expand...

Political and economic decisions are intertwined.

You can't really separate the two
 
Wrought

Wrought

Senior Member
Registered Member
AndrewS said:
Political and economic decisions are intertwined.

You can't really separate the two
Click to expand...

Yes that's right when it comes to real-world decisions, but your original assertion was about economic theoreticals, and it was wrong.

AndrewS said:
But remember that comparative advantage doesn't work if:
1. China has an absolute advantage in both manufacturing and service industries
2. Plus the spare resources to fully apply AI to both manufacturing and service industries

Comparative advantage only applies if both sides are limited or face resource constraints
Click to expand...

Comparative advantage applies in every case, even though the correct decision (accounting for politics) may be to avoid it.
 
O

OptimusLion

Junior Member
Registered Member
Kimi's latest model k1.6 tops the list, with programming capabilities surpassing GPT o3mini and o1

Kimi's latest model k1.6 was exposed today. According to the global dynamic benchmark platform LiveCodeBench, Kimi k1.6 surpassed GPT o3mini, o1 and other models and topped the code generation test.

Zhang Yutong, co-founder of Moonshot, said in WeChat Moments, "k1.6 has not been released yet, but I am happy to see the results of LiveCodeBench, and I look forward to the continuous evolution of intelligent models."

001ZzMwgly1hyz997ov37j612z0g70w802.jpg

Please, Log in or Register to view URLs content!
 
O

OptimusLion

Junior Member
Registered Member
Tencent Hunyuan's new generation of fast thinking model Turbo S is released, supporting "instant reply"


Tencent officials said that unlike Deepseek R1, Hunyuan T1 and other slow-thinking models that need to "think before answering", Hunyuan Turbo S can achieve "instant reply", output answers more quickly, double the word speed, and reduce the first word delay by 44%.


001ZzMwgly1hyzamurythj60u00kogsh02.jpg

001ZzMwgly1hyzamuymwkj60u00grwib02.jpg

Please, Log in or Register to view URLs content!
 
You must log in or register to reply here.
Top