Artificial Intelligence thread

[Some1Guy]

why did they go abroad tho? alsp if ever how will CN enforce their decision about it? like in Singapore?

As far as i know part of the company still exists in China, and most of their research happened in China meaning they're still under Chinese jurisdiction and have to follow Chinese export controls. Also the investigation is not about preventing this one acquisition but setting a precedent for other companies trying to do the same.

 

[Moonscape]

why did they go abroad tho? alsp if ever how will CN enforce their decision about it? like in Singapore?

If all of the high level people at Manus are okay with never setting foot in China (including Hong Kong) ever again, then yes it'll be hard for China to enforce any sort of decision on Manus. But unless all the top folks at Manus are okay cutting all ties with China forever, there'll be some incentive to make a deal.

 

[tphuang]

Exclusive: DeepSeek withholds latest AI model from US chipmakers including Nvidia, sources say​

• DeepSeek grants early access to domestic suppliers such as Huawei
• US official says DeepSeek model trained on Nvidia's chips
• China strategy may be to place US hardware, models at disadvantage, research firm says

Please, Log in or Register to view URLs content!

this is a deeply stupid article and report.


on a separate note, I asked Gemini, GLM, Minimax and Kimi to do a semi difficult task on the chatbot. Kimi had the best and correct solution, with Minimax coming next. Gemini free version had the worst.

the new small Qwen3.5 models are popular on HF

https://twitter.com/i/web/status/2027043336670924947

I have the 35B-A3B model running on my MacBook locally using Unsloth GGUF and llama.cpp

https://twitter.com/i/web/status/2027037216720269563

Minimax is spreading its word to more places

https://twitter.com/i/web/status/2027027163078336885

Minimax Agen & Maxclaw going forward.

 

[siegecrossbow]

why did they go abroad tho? alsp if ever how will CN enforce their decision about it? like in Singapore?

Take advantage of the sale and add some backdoors.

 

[PandaAI]

Please, Log in or Register to view URLs content!

Meta Platforms is moving forward with integrating the newly acquired artificial intelligence agent start-up Manus, according to two sources, despite Beijing’s probe into the US$2 billion deal.

Some members of the Manus team in Singapore had moved into Facebook parent Meta’s offices and were granted Meta corporate accounts and other access, one of the sources said.

Meta was also offering its employees the opportunity to transfer to Manus, preferring Chinese-speaking staff, a second source said.

The merger not only involves talent but also operations. Meta recently embedded Manus into its Ads Manager so that advertisers on the platform could get performance analytics and optimisation advice.

Neither Meta nor Manus immediately responded to requests for comment on Thursday.

The integration has proceeded despite an ongoing investigation by the Chinese government.

In early January, just days after the deal was announced, the Ministry of Commerce said it would review and investigate the deal to assess whether it was consistent with China’s regulations on export controls, technology exports and external investments.

There has been no update since.

The South China Morning Post reported on the investigation a day before the official announcement, citing two sources.

One of them said Beijing feared the Manus case could set an uncomfortable precedent for other Chinese AI companies to follow by moving their operations abroad.

The start-up rose to fame in March 2025 after releasing what it described as the world’s first general AI agent – software that can complete tasks on a user’s behalf.

The team initially operated in Beijing and Wuhan, but moved to Singapore by mid-June, laying off some China-based staff and shutting down its Chinese social media accounts.

China updated its technology export controls in 2020 to cover certain algorithms, a change widely viewed as strengthening its legal tools to intervene in transactions after Washington pressured ByteDance to divest TikTok’s US operations.

Meanwhile, Washington has also launched stringent rules restricting American capital from supporting Chinese AI and semiconductor firms.

China must update its laws urgently. Companies originally registered in China should not be allowed to deregister in China and move its operations overseas. They have data from China. This is a national security issue for China. Manus and Meta exploited a loophole in the laws in China to do this.

First is to prevent other companies from doing this again. Second is to sanction all members of Manus from doing business with any Chinese (including HK) company or individual again. I’m sure these Manus employees hold Chinese passports and are citizens of China.

 

[henrik]

China must update its laws urgently. Companies originally registered in China should not be allowed to deregister in China and move its operations overseas. They have data from China. This is a national security issue for China. Manus and Meta exploited a loophole in the laws in China to do this.

First is to prevent other companies from doing this again. Second is to sanction all members of Manus from doing business with any Chinese (including HK) company or individual again. I’m sure these Manus employees hold Chinese passports and are citizens of China.

If they violated national security laws, then they should be tracked like those HK separatists.

 

[tphuang]

https://twitter.com/i/web/status/2027376994233536662

my experience with Kimi yesterday was really good. Although Minimax was surprisingly good at deep research also.

I don't see why people are so obsessed with copilot type of tools by Opus. They are surrendering control of their computer. I can do everything I need to be done right now with Kimi 2.5

 

[tphuang]

https://twitter.com/i/web/status/2027443974055551017

ByteDance and Doubao is the clear leader in China's AI app war right now.

 

[Kalum Pupeter]

Pentagon moves to build AI tools for China cyber operations​

US holding talks with companies to identify vulnerabilities in Chinese networks

The Pentagon is seeking to make AI-powered cyber tools to identify infrastructure targets in China as part of an effort to improve US capabilities in any future military conflict with Beijing. The department was in talks with leading AI companies about partnerships to conduct automated reconnaissance of China’s power grids, utilities and sensitive networks as well as those of other adversaries, said several people with knowledge of the plans.

The US has already created powerful cyber-espionage weapons but is seeking to deploy new AI-powered tools to identify software flaws in opponents’ systems that could then be exploited to enhance infiltration and degrade those systems in any conflict.
The proposed system would use

Please, Log in or Register to view URLs content!

to penetrate computer networks, map vulnerabilities and integrate potential targets into US war planning, the people added. The Pentagon declined to comment.

OpenAI, Anthropic, Google and Elon Musk’s xAI have been awarded contracts worth about $200mn to partner with the US government for military, cyber and security applications. Which companies will be involved in the new cyber initiative is yet to be determined.
A senior US official on Thursday warned the administration would rip up all its existing agreements with

Please, Log in or Register to view URLs content!

if the company failed to reach a deal with the Pentagon, after chief Dario Amodei said he would reject a “final offer” for the terms under which it would work with the military.

The Pentagon’s effort reflects recognition in Washington about the increasing importance of cyber operations in any war with China and the view that AI could help tilt the balance in a conflict. But the move also comes at a time of

Please, Log in or Register to view URLs content!

with some of the country’s most advanced AI companies over how far their technology should be used in military operations.
Dennis Wilder, former head of China analysis at the CIA, said AI cyber tools would help solve the problem posed by the huge amount of manpower needed to scan and identify vulnerable infrastructure.

“It’s equivalent to the thief in the night who tries the front door to homes until they find one that has been left unlocked,” said Wilder, now at Georgetown University. “AI-assisted cyber hacking can exponentially increase the number of doors tested and thus allow for much more efficient and accurate mapping of targets for selection.” Military cyber experts already conduct work identifying vulnerable targets, but the envisioned new AI tool would perform these tasks faster and at a higher volume with much less human involvement.
“They have been building cyber offence strategies for infiltrating power grids. You have to build both the offence and the defence, you can’t have one without the other,” said one of the people familiar with the plans.

Another person said power plants near data centres could be targeted to disrupt adversaries’ AI capabilities. Defence secretary Pete Hegseth has sought access to powerful generative AI technology for what he describes as “lawful use”. But AI labs have hesitated to give the Pentagon open-ended control over the technology. Anthropic had sought to block the use of its model Claude in lethal autonomous weapons and wanted restrictions on AI use for mass domestic surveillance, people close to the company said.

Amodei was told by Hegseth on Tuesday that his company could be branded a supply chain risk or have its technology co-opted by the government if it failed to agree to his terms by Friday. Claude is currently the only AI model used in classified operations, but staff at other AI labs, including OpenAI and Google, have raised similar concerns, according to multiple people with knowledge of the matter.
“This is a response to what’s coming out of China and the lack of guardrails there,” said one person familiar with the Pentagon’s stance. “It’s open-ended use, we can’t have shackles on us when it all kicks off.”

Please, Log in or Register to view URLs content!

Please, Log in or Register to view URLs content!

I completely agree with Teortaxes and Zephyr. They clearly have above-average knowledge on this topic. The next paradigm could involve AI agents used either for deterrence or for crippling a country’s civilian and military infrastructure to the core.

 

[tphuang]

Please, Log in or Register to view URLs content!

Please, Log in or Register to view URLs content!

You can just post the original article and your own thoughts about relevant part of the article, what is the point of including x posts from random people online?

This is a serious thread, so please keep it to relevant material.