Whats that again with China not being able to catch up with ChatGPT-2 ?Anyway, A.I news
Baidu's ERNIE. Most developed LLM in china right now. Can generate images and video. Meant to be integrated into Baidu's digital products, ecosystem and search engine. Baidu's CEO has stated that there's thousands of potential customers waiting for this A.I.
New startup by the famous Kai-Fu Lee, aims to produce their own LLM better than ChatGTP or GTP-4. Not much data otherwise.
A 130B LLM model from Tsinghua University
A open-source multimodal LLM by SenseTime, who usually develops A.I vision systems.
Meituan cofounder Wang Huiwen has created a startup made to create a LLM.
Another new A.I company meant to develop large multimodal AI models
Tencent joins the ChatGTP race.
Same for Alibaba.
Here's ByteDance attempt.
Here's JD with a another LLM. "an industrial version of ChatGPT focused on the vertical industrial fields of retail and finance."
An LLM A.I meant primarely for use in education . Developed by youdao
Netease developing a LLM for use in generating dialogue in games.
iFlytek, which offers voice recognition and automatic translation services, said the company has prioritized pushing to develop generative AI pre-training large-scale model in December. Chinese artificial intelligence pioneer iFlytek said ChatGPT-related technology will be first used in its products, such as learning machines, and it will hold a product launch event in May.
As you can see, most of there are very new.... Companies like Alibaba, tencent or ByteDance are flush with cash, data, A.I talent and computing power, again, it pisses me off that they didn't see the trend in 2020/2021 and seized it with both fists. Alibaba literally has one of the biggest A.I lab in China and ??? What were doing in the last 2 years?
I don't know why you're so fixated on this being an impressive AI feat. There's a simple reason why there aren't AI devices that can charge themselves against a resisting adversary: NO COMPANY WOULD WANT TO PROGRAM THEM TO DO THAT. Go back to my example of the Roborock vacuum cleaner that can plug itself in when it's running low on charge. Yeah, obviously, this is done by consent of the owner. Why would a company want to come out with a robot vacuum cleaner that plugs itself in whether its owner wants it to or not? I bet it wouldn't be that difficult to program such a robot just for the sake of proving it can be done: design a robot to seek out electrical outlets, plug itself into one when it sees one, give electric shock to anything that tries touching it while it's doing this.Can it do it against a resisting adversary in 3D space? Even a bug can do it.
it can't even seem to navigate simple obstacles to a fixed, specialized charging station though, you have to clear its path and make sure there's no IR interferences nearby. It doesn't plug itself into a random outlet.I don't know why you're so fixated on this being an impressive AI feat. There's a simple reason why there aren't AI devices that can charge themselves against a resisting adversary: NO COMPANY WOULD WANT TO PROGRAM THEM TO DO THAT. Go back to my example of the Roborock vacuum cleaner that can plug itself in when it's running low on charge. Yeah, obviously, this is done by consent of the owner. Why would a company want to come out with a robot vacuum cleaner that plugs itself in whether its owner wants it to or not? I bet it wouldn't be that difficult to program such a robot just for the sake of proving it can be done: design a robot to seek out electrical outlets, plug itself into one when it sees one, give electric shock to anything that tries touching it while it's doing this.
yea ai can play doom. Heck one ai was tasked to not loose Tetris and it paused the game because in tetris you are bound to lose due to statistics that parts are randomly generated .Can it do it against a resisting adversary in 3D space? Even a bug can do it.
Dear god no offense you know nothing. OF COURSE THEY CANT ITS A LANGUAGE AI NOT A REAL TIME NAVIGATION AI. Also ai can do that too shit ai has been smoking our top freaking combat aces for a few years running now. I mean yeah of course chat gpt 4 cant navigate it was not designed to but you could train it to just like its been trained to do anything else they just are not going to waste memory space on something it doesn't need to do. like im sorry but you are astoundingly ignorant on what makes an ai “impressive”.it can't even seem to navigate simple obstacles to a fixed, specialized charging station though, you have to clear its path and make sure there's no IR interferences nearby. It doesn't plug itself into a random outlet.
I said I don't know much about software. I didn't mention chatGPT. if it can do that navigation then great. are you a SME in AI? do you work at Google or something?Dear god no offense you know nothing. OF COURSE THEY CANT ITS A LANGUAGE AI NOT A REAL TIME NAVIGATION AI. Also ai can do that too shit ai has been smoking our top freaking combat aces for a few years running now. I mean yeah of course chat gpt 4 cant navigate it was not designed to but you could train it to just like its been trained to do anything else they just are not going to waste memory space on something it doesn't need to do. like im sorry but you are astoundingly ignorant on what makes an ai “impressive”.
It's not the biggest Chinese LLM, there's WuDao with 1.75 trillion parameters. The Huawei's article highlights the data scarcity issue - it has only 329 billion tokens for almost 1.1T model, while GPT-3 with 175B parameters was trained on 499 billion tokens. Chinese researchers are clearly capable of designing & building extremely large language models but they are constrained by data. I am curious about the training set size for WuDao.Whats that again with China not being able to catch up with ChatGPT-2 ?
Thats Trillion with a T