Artificial Intelligence thread

Randomuser

Randomuser

Captain
Registered Member
I'm beginning to think soon there is gonna be an AI bubble burst that's gonna be real ugly for America given the amount of hype and money put into it.

China on the other hand which is more focused on application over getting investors hyped will make a slower but more steady approach that will see genuine improvements to society based on AI.
 
tphuang

tphuang

General
Staff member
Super Moderator
VIP Professional
Registered Member
nativechicken said:
I just returned from a business trip, and my development is mainly focused on embedded environments, as well as tinkering with NPU AI applications. The solutions I have been exposed to include
rk3399/rk3568/rk3576/rk3588/ax630/ax650（1-18TOPS int8）
What I am doing in the future will basically involve the targeted application of all domestic AI SOC, including those discussed by the general public and those not discussed
I haven't had much exposure to AI SOC now, purely because I don't have the money to buy too many development boards and carry out development (I am currently working on mass production customization of AI hardware, with an estimated investment of 120000+USD), because I will add NPU/AI functionality to a specific customized Linux solution. It is 10 times more troublesome than on official custom Linux distributions (usually Ubuntu/debian/armbian).


I'm no amateur in AI hardware/software, though my projects primarily use the YOLO series—not LLM-type models. Truth is, current-generation NPUs like those in the RK33/RK35 series SOCs aren’t well-suited for LLMs.

It’s the next-gen RK36 series that’s designed for it.

I respect the information shared by the person in the video, though I wouldn’t consider them an expert either. That said, their insights are particularly intriguing.

I’ve engaged with acknowledged heavyweights in China’s internet circle—many now diving into AI. Technically, I’m confident we have a far clearer grasp of the underlying tech than these industry leaders.

But commercially speaking, early-stage innovation ventures’ survival—fundamentally reliant on a CEO’s fundraising prowess—means individuals like those in the video prove indispensable in the commercialization process.
Click to expand...
Yes I have tested with pretty much every board you mentioned in there. And rk3688 won’t be ready until end of 2026 based on my conversation with rockchip guys a month ago. Axera is actually proposing something with higher tops for int4 and they have in fact better software support. They told me their next chip will do 72 tops. But okay, even with 18 tops, I am not fully utilizing the NPU because the memory they put on those boards are too small. So what’s the point?

There is also horizon but there is no LLM support for them.

A huge % of people in Shenzhen think the current SoC can’t run LLMs on edge but guess what, a lot of people in China are trying it out. Do you know how I know this? Axera told me. My former partner factory in Indonesia said this too.

and I am pretty sure you can get it working reasonably. Whether people actually can sell the product is a different story.
 
tphuang

tphuang

General
Staff member
Super Moderator
VIP Professional
Registered Member
T

tamsen_ikard

Senior Member
Registered Member
Hyper said:
Oof, LLMs are difficult to monetize. I wonder what happens to all the investments. Some are more vulnerable to dud investment than others. Expect Meta to stop burning cash by end of 2026. GPT will have to improve for Microsoft to continue deals.
Click to expand...
AI bubble will cause one of the biggest crashes in history for the US.

They still don't realize AI will be commodity thanks to Chinese open sourcing l
 
siegecrossbow

siegecrossbow

General
Staff member
Super Moderator
superdog

superdog

Junior Member
Bellum_Romanum said:
What are the area Gemini 2.5 excel and beat the rest of its competitors (wether closed or open source) that makes you think that the forth coming version is going to be the one to beat?
Click to expand...
I know you're asking someone else but in case you're interested:

I use Gemini 2.5 Pro for to help with clinical work in psychology, such as case analysis, summarizing progress, treatment planning suggestions, integration of different theories, creating meditation scripts and worksheets, etc., I found it performing better in most cases than Claude 4.1, GPT-5, and other Chinese models like Deepseek R1, Qwen 3 etc. The advantage is not necessarily in insight or knowledge (most SOTA AI would cover similar keypoints), but it often has the most polished output, with the right amount of detail (others are sometimes too brief), with the most natural tone and pacing when making verbal scripts, and overall very good instruction following.

I have a personalized tool that let me consult multiple AI at once using the same prompt, so I cross compare them every time I submit a query.

For programming work I still use Claude and Qwen Code.
 
You must log in or register to reply here.
Top