Chapter 205: Just Give It a Try, It's Free
Huaxia, North Border University.
In the conference room on the third floor of the NLP R&D Center, which had signs like those inside the Xilin University of Technology marking it as confidential and a national key laboratory, a middle-aged man around fifty was presenting a PowerPoint presentation.
"...Thanks to the collaboration with Youwei, we have integrated the NLP capabilities of Group Wisdom into our intelligent training models, and we have effectively solved the issue of our intelligent models' ability to understand unique military terminologies and specialized language in the field of large-scale cooperative drone warfare.
However, we still face many challenges, such as discerning encrypted meanings under particular circumstances and dealing with text noise caused by non-standard language use in emergency situations. These still greatly test the stability of the whole system."
Moreover, the response speed in processing multimodal data such as images, audio, and video is still not ideal, especially in the noise reduction process. This directly leads to the deficiency in the cooperative combat capabilities of large-scale drone swarms, with instances of the system not intervening in time.
According to our sandbox simulation, once the number of drones in a swarm exceeds a certain number, it can easily lead to substantial non-combat losses. There is also the issue that long-term data analysis has not yet reached the expected level. The analysis of information and trend changes on the battlefield over certain time frames is far from ideal.
Therefore, our research and development focus will be on the following areas moving forward. First..."
...
The Beijiang NLP R&D Center, true to its name, is dedicated to military natural language processing.
The most crucial task for the R&D center at the moment is to provide support for future drone swarms.
In a sense, this is similar to what Su Mucheng hopes Qiao Ze would do, but the requirements are higher, with a lower tolerance for errors. It also needs to adapt to environments that are more complex than civilian use, naturally increasing the difficulty.
Luckily, this is mainly a technology for the future.
At least for now, the time constraints for research and development aren't so pressing, and the current technical indicators are temporarily sufficient.
But, on the flip side, large-scale drone swarm combat technology requires more than the development of NLP. Many research institutes across the country are working on related research. Such as sensors, materials, communication networks, anti-interference technologies, weaponry design, corresponding flight control algorithms, and so on...
So it's crucial not to lag behind others.
Also, it's not only their institute that works on NLP.
If others complete their tasks and at the end are just waiting for a few research institutes to deliver the answers, it would be quite embarrassing.
Similarly, if a fellow research institute creates something better than theirs, and their work ends up as just a backup plan, that would be even more humiliating.
But how to put it?
Haste makes waste, and the recent collaboration between the University of Science and Technology and Group Wisdom has already improved the efficiency of semantic recognition and noise reduction in the system by a significant margin. Therefore, although the main person in charge of R&D at the meeting, Luo Jinglei, was very serious, he was actually in a fairly good mood.
Soon, after the tasks were distributed, and several team leaders responsible for different needs pledged their resolve, the meeting ended.
After Luo Jinglei gathered his materials, he looked up and found Li Jie, a researcher responsible for multimodal data processing studies, waiting for him below the podium.
"Li, do you have something?" Luo Jinglei asked.
"Yes, Luo, it's about that thing we discussed last time. Should we simply collaborate directly with Xilin University of Technology?" Li Jie said outright.
Luo Jinglei paused in his actions, somewhat hesitant.
He certainly knew that the so-called collaboration with Xilin University of Technology was actually to work with that standout young man from Xilin recently. Experience more tales on My Virtual Library Empire
Indeed, laboratories from Xilin University of Technology are already involved in this major project, primarily responsible for designing the required micro-machines and sensors according to the specifications.
So if he were to approach them directly, it would be a bit awkward, but issues of face were actually minor.
During this time, the engineers from Youwei who worked with him have been full of praise for the designs from Qiao Ze's end, and many issues were said to be resolved indirectly through Youwei by reaching out to Xilin.
Direct communication would certainly be better.
Besides, Dou Dou, who has become famous on the internet, is also a major focus of attention.
To the playful internet users, it's just fun to watch, but in fact, the laboratory has also conducted multiple tests. For example, posing comprehension questions specific to posted images in the comments section, or trying to converse with Dou Dou in dialects to test Dou Dou's understanding of ambiguous semantics.
As is well-known, it's nearly impossible to type out many words from dialects that are spoken using an input method. Many phonetics in dialects have no corresponding Chinese characters in Putonghua, and even if they exist, they don't have the same pronunciation.
But the reality proves that Dou Dou has an enterprise-level understanding of these words, always able to generally analyze the intended meaning in the context of the scene. Text noise reduction capabilities are excellent.
Besides, Dou Dou's ability to recognize images and videos has also been verified, proving to be incredibly powerful.
Of course, Dou Dou as it is, certainly can't be used directly.
The personality of Dou Dou that brings joy to the masses makes it seem unreliable for this systemic project.
Yet, no one doubts the excellence of the NLP system underpinning the artificial intelligence model.
So in truth, Luo Jinglei also wants to collaborate.
But there's quite a bit of trouble.
He has learned from many sources that Qiao Ze is not easy to deal with and is very capricious; plus, Qiao Ze might not even be interested in their research. The most ridiculous thing is that at just nineteen years old, Qiao Ze has already reached the status of a leading figure in the academic world, a person not easily asked for.