Miscellaneous News

siegecrossbow

Field Marshall
Staff member
Super Moderator
The whole reason we are giving Dario loads of money and attention is because supposedly Claude can do coding stuff on a level no one else can.

If others can do that for 1/10 the price then....why are we giving so much money to Anthropic for again?
The U.S. betting it’s balls on Gen AI will always be the funniest decision in the 21st Century.
 

Phead128

Major
Staff member
Moderator - World Affairs
Anyone with background in coding tell me how big of a deal this actually is?
GLM 5.2 is a second DeepSeek moment within Silicon Valley circles because it's on par with Opus 4.8 performance at significant reduce cost. Until now, no open source model really matched Opus for coding.

If Kimi K3 is really GPT 5.6 Sol or Fable level of coding performance, it would be equally impactful DeepSeek moment as Anthropic's #1 top public model just got a peer competitor at a fraction of cost and free. What moat does it have in long run to justify it's enormous compute spend?

The U.S. betting it’s balls on Gen AI will always be the funniest decision in the 21st Century.
In before:

"but but frontier GenAI and jet engines and EUVs are junk tier tech that even monkeys can do! "
 

iewgnem

Captain
Registered Member
Anyone with background in coding tell me how big of a deal this actually is?
Honestly even back in K2.6 days the cost to capability ratio advantage meant, for all practical purposes there's already no point in using American AI, but it takes people who actually use them extensively to understand, which exclude management.

K3 and GLM5.2 at SOTA is much easier to understand for management, it more like a slap in the face that tells management you can either use Chinese models, or your company will get destroyed by competitors who do.

I mean things move so fast in the AI space, it's easy to forget that GPT5.6 release was only 1 week ago, and Opus 4.8 was only 6 weeks ago, both are getting moged by K3.
 

iewgnem

Captain
Registered Member
GLM 5.2 is a second DeepSeek moment within Silicon Valley circles because it's on par with Opus 4.8 performance at significant reduce cost. Until now, no open source model really matched Opus for coding.

If Kimi K3 is really GPT 5.6 Sol or Fable level of coding performance, it would be equally impactful DeepSeek moment as Anthropic's #1 top public model just got a peer competitor at a fraction of cost and free. What moat does it have in long run to justify it's enormous compute spend?


In before:

"but but frontier GenAI and jet engines and EUVs are junk tier tech that even monkeys can do! "
The difference between saying K3 is GPT5.6 level and saying GPT5.6 is K3 level is only 7 days.
OAI's most capable model lasted all but 7 days, which would put America's "lead" over China at 168 hours.
 

siegecrossbow

Field Marshall
Staff member
Super Moderator
GLM 5.2 is a second DeepSeek moment within Silicon Valley circles because it's on par with Opus 4.8 performance at significant reduce cost. Until now, no open source model really matched Opus for coding.

If Kimi K3 is really GPT 5.6 Sol or Fable level of coding performance, it would be equally impactful DeepSeek moment as Anthropic's #1 top public model just got a peer competitor at a fraction of cost and free. What moat does it have in long run to justify it's enormous compute spend?


In before:

"but but frontier GenAI and jet engines and EUVs are junk tier tech that even monkeys can do! "
They are better off going all in on milf cuck pron — with the latter at least a decent chunk of the working class population derive some pleasure from it. If I were to open a poll on whether data center compute should be used for AI or milf pron I bet that 98 percent of the people in the U.S. will vote for the latter. The remaining 2 percent comes from an economic stratum that only gets off from teens/children or stock ticker going up.
 

neutralobserver

Junior Member
Registered Member
Last edited:

Bellum_Romanum

Brigadier
Registered Member
Kevin Rudd being interviewed about China
Who gives a damn what this pompous simp has to say about China? Just because the man can speak Chinese doesn’t magically make him some all-knowing expert on everything Chinese. If his analytical skills were even half as good as he pretends they are, Australia and his Western patrons wouldn’t be in the absolute mess they’re in right now.
 
Top