I... Fuck!
Laymen can hardly imagine what went through the professors' minds after Qiao Ze humbly asked for their guidance.
If it weren't for their habitual restraint, there probably would be a tea room full of "Holy shit!" at that moment.
If this topic could actually be accomplished, to hell with becoming a Yangtze River Scholar, one could proudly go for the National Science and Technology Progress Award and then waltz into the Science Academy to become an academician.
Going for the Yangtze River Scholar would be way beneath this endeavor.
But no matter whether this topic could be accomplished or not, Qiao Ze had already caused quite a stir among the professors just by voicing these concepts.
Regardless of whether Qiao Ze was talking big...
Those who are new to Group Theory could never combine a self-supervised feature decoupling framework with Group Theory.
Even now, half of the professors still had not completely grasped what Qiao Ze's research topic was supposed to mean.
This was already a highly specialized research direction.
And it indeed had practical implications.
Chat GPT had already become a hit sensation recently.
Every industry was getting bombarded by chat GPT, even to the point of questioning if their jobs could be replaced by this cutting-edge artificial intelligence technology.
If Huaxia could really develop an AI program that the industry recognizes as superior to chat GPT, and get a few more people to rally behind this kid, the National Science Progress Award might very well make an exception.
But should this really be something a high school student who hasn't even started college be considering?
Though the aroma of tea was rich in the air, the tea itself had suddenly become hard to swallow.
...
"How did you come up with this proposition?" someone asked after calming down from their shock.
Qiao Ze patiently explained, "Because I have used chat GPT. I think the current models have problems with black box characteristics and biased predictions, so I believe incorporating a causality model could improve the model's robustness and interpretability. Current similar technologies depend too much on big data, which causes issues with high-dimensionality, continuity, and coupling.
"Particularly with the selection of many variables, the variables that we observe are naturally decoupled, like Simpson's paradox. If we could apply Group Theory to introduce the traditional causality model framework, we could change this situation."
Zhang Chunlei then asked, "Do you realize how difficult this proposition is? Have you thought about how many challenges will need to be addressed in researching this topic?"
Qiao Ze nodded and said, "Yes. The biggest difficulty is finding a universally applicable invariant, or devising a way to separate variable from invariable to enable computer understanding; the next challenge is decoupling the variables themselves, which requires finding a method to balance them; and how to integrate it into the current deep learning models without causing conflicts. Those are probably the three toughest points."
The professors were stunned...
This really seemed like he had thought it through thoroughly!
Zhang Chunlei looked at Qiao Ze with a serious expression for a good while, then raised his head to glance at Li Jiangao, who had been quiet beside Qiao Ze, before directly asking, "Li, how long have you been mentoring this kid?"
Li Jiangao picked up his phone, glanced at it, and then let out a wry smile, "Roughly four and a half hours, I guess."
Zhang Chunlei was taken aback, then turned his gaze back to Qiao Ze, who remained expressionless, and asked, "Qiao Ze, when and from whom did you start systematically learning mathematical knowledge?"
Qiao Ze thought for a while before shaking his head and saying, "I haven't systematically studied it."
Zhang Chunlei frowned and asked, "Then how did you come to understand Group Theory, and how did you come up with the things you just mentioned?"
Although Qiao Ze's reply was a bit slow, it was very meticulous, "I saw a very interesting question online, and just thinking about it seemed to clear my mind. Then I got hooked. Later, I realized that the knowledge from textbooks simply wasn't enough to grasp it, so I started looking for materials and learning online.
Basically, first understanding what I needed to learn to solve this problem, I would search for video courses in that area online. So far, I have watched some online calculus courses, as well as calculus, linear algebra, geometry, algebraic geometry, affine geometry, projective geometry, basics of Group Theory, foundations of modern physical theories, high energy physics, analytical particle physics, and an in-depth mastery of quantum mechanics.
Oh, and also computer-related courses, because the computations needed to solve that really difficult problem are quite substantial, and require the aid of computers."
"Finding courses online and self-studying?" A professor asked in surprise.
"Isn't that what you all do?" Qiao Ze asked back, just as surprised.
"Is there a chance that one could learn these courses more efficiently in a classroom?" A professor asked with a wry smile.
"Impossible! Even in the videos there's a lot of nonsense from the teachers, reemphasizing a straightforward concept as if everyone's an idiot. With videos, you can just skip ahead, but in a classroom setting, you can't do that—you just zone out instead," Qiao Ze earnestly retorted.
That made sense, and the professors found themselves unable to argue back, even beginning to silently reflect whether they had been too verbose during their own lectures. They wondered if they had underestimated the intelligence of the bright young students beneath the podium.
"Haven't you ever come across a problem that wasn't explained clearly in the online courses? Something that you didn't understand after watching and needed to ask questions about?" another professor pressed on, not quite ready to let go.
Qiao Ze nodded, admitting openly, "Of course I have."
He then carefully explained, "But there are so many similar videos online, if one teacher doesn't clarify something, then I just switch to another video to see if that makes it clearer. Generally, by comparing a few, I can understand. If, after trying three different videos, I still don't get it, then I'll stop and systematically review the key points before and after the knowledge in question, and I am sure to understand it then. This is a little learning tip I'm sharing, and I want to take this opportunity to thank Professor Zhang for answering my question just now."
After saying this, Qiao Ze instinctively glanced at Li Jian Gao beside him.
He felt that his performance today had definitely been excellent.
He remembered the benefits he had received from others and did not forget to repay them.
This must be what Uncle Li had just taught him—the way to express gratitude after learning to be thankful.
This great learning experience should be helpful to them, right?
Unfortunately, Li Jian Gao hadn't noticed Qiao Ze's glance; his mind was completely blank at the moment.
He'd thought that preventing Qiao Ze from mentioning the topic would save him from going overboard, but who knew that the kid's head was clearly packed with too much stuff…
He had a feeling that he would not get a good night's sleep tonight.
Even if they were late for tomorrow's meeting, these people would not let him off the hook.
Noticing Li Jian Gao zoning out, Qiao Ze felt slightly unhappy, but he quickly adjusted his attitude.
Actually, he was quite happy today.
Although these uncles didn't seem very bright, at least they genuinely understood what he was talking about, and they were able to give feedback.
Not like those people he'd encountered before. It's one thing for them not to understand what he said, but the way they looked at him, with all sorts of scrutiny—it was so stupid, it was off the charts...
Ah, just thinking about it is exhausting!