Artificial Intelligence thread

[Eventine]

The American AI companies pay lip service to open source but it is pure PR. In reality they are 99.999999% focused on building up and maintaining walled gardens and captive markets based on proprietary models. Chinese open source ecosystem and low AI costs (energy or otherwise) are the ultimate weapon to eventually win dominant market share globally. Western companies cannot even help but to outsource decent white collar work to India just to save a little money in the grand scheme of things. Are we really supposed to believe they can universally and forever stick with extremely expensive American AI? Especially with their energy generation bottleneck, cost of energy infrastructure, slow build times, high energy costs, etc.

They are simply taking the recipe they used to monopolize global internet platforms (Google Search, Facebook, YouTube, iOS, Android, Windows) and attempting to extend it into Big AI. This is entirely unsurprising and if not for China open sourcing models they would have almost certainly succeeded. Even now they are more successful than not due to how gullible the global population outside of China is; let’s not ignore the fact that Open AI is still by far the largest global market share on consumer AI and Gemini is their closest competitor, not any Chinese model, due to close integration with their existing platform monopoly.

 

[tphuang]

Any whispers on when the next Deepseek model is being released? Was supposed to be May, but still waiting. Trying to time buying put options on NVDA and semiconductors. Maybe just look for large put option contracts for NVDA since it may be the hedge fund side of Deepseek making a killing again.

they just released a major reasoning update on May 28th.

I think their latest version is as good as OpenAI latest models.

depends on what you use it for. Most people using it for English application probably find o3 results to be a little better. I used o3 for some work stuff that was quite difficult and compared it to deepseek R1. Neither could solve the problem, but o3 seemed to give better results. Although, o3 seems to have this nasty habit of being extremely confident while hallucinating all the time.

https://twitter.com/i/web/status/1944130898448069065

Honor using data from its manufacturing to create Luban large model for world's 1st L4 phone production factory.

 

[tphuang]

https://twitter.com/i/web/status/1944137228302590069

Huawei Cloud announces launching of its Horinger data center with CM384 super nodes. It is the first one in Northern China to have it and Huawei's largest infrastructure in Northern China. This region is a cluster in EDWC and already has 91 EFLOPS in computation across 46 projects. Utilizing > 80% green electricity.

 

[tphuang]

https://twitter.com/i/web/status/1944354233790218739

Alibaba Cloud showcases how its overseas Data Center infrastructure has helped the expansion of several Chinese AI firm expansion outside of China.