At 12:45 PM, at the Bauhinia Garden Cafeteria.
As one of the busiest canteens at Shuimu University, it was bustling with activity at this moment. The air was filled with the aroma of meat...
The aroma of sour, sweet, and spicy fillings filled the air, and all around were students walking arm in arm, enthusiastically discussing advanced math problems or gossiping about their professors.
Zhang Wenduo carried a plate of scrambled eggs with tomatoes, shredded Fish-Fragrant Shredded Pork, and a bowl of rice, and casually found a corner to sit down.
He mechanically shoved the food into his mouth while absentmindedly glancing at the computer science department's group chat notifications on his phone, his mind preoccupied with how to revise his thesis.
The notice was sent out in the morning, but he was focused on his research and only took it out to look at it again during lunch.
[@All members, at 2:30 PM today in the East Main Building Auditorium, Mr. Han Luyi, founder and CEO of Source Intelligence Technology, will deliver an academic lecture on the topic of "The New Relationship Between Humans and Intelligence in the Era of Artificial Intelligence." Students from all labs who are not on duty are requested to arrive 10 minutes early, as seating is limited.]
He hurried over, even though the notification time was so short that not many people would come, but Zhang Wenduo planned to find a seat near the front so he could ask questions easily.
With that in mind, he quickly ate two mouthfuls of rice and headed straight for the main lecture hall in the east building.
Zhang Wenduo arrived at the lecture hall at 1:10 PM, and was surprised to find that there were already quite a few people there.
Students in the Shuimu School of Computer Science aren't particularly enthusiastic about lectures. Simply put, there are just too many lecturers at Shuimu Forum, and they're all celebrities.
Academicians, professors, entrepreneurs, celebrities, and visiting dignitaries from abroad all gave speeches.
This can be considered a perk of attending prestigious universities, but you can't possibly go to every single one. Just pick and choose the ones you want to hear.
Zhang Wenduo had just found an empty seat in the first row when he suddenly heard someone call out to him, "Senior Brother Zhang!"
He turned around and saw a junior student, who, like him, was Professor Shao's doctoral student.
"Senior Brother Zhang, what brings you here? Isn't Source Intelligence Technology a company that works on Large Language Model? That's not your research area, is it?" the junior brother asked.
Zhang Wenduo was momentarily at a loss for words when asked the question. He was actually planning to change the direction of his dissertation. After thinking for a while, he managed to stammer out, "----It's not that inconsistent."
The junior colleague didn't mind, sat down next to him, and started gossiping. "This founder is quite young, not much older than me. Let me tell you, it's going to be crowded later, so I came early to grab a seat."
"Why?" Zhang Wenduo asked curiously, "The notification was given in such a short time, many people probably didn't see it."
The junior student gave him a mysterious look. "Senior, you don't usually go on the forum, so you don't know, but the founder is a handsome guy with a lot of female admirers at our school. I just saw on the forum that many girls from other colleges are planning to skip class to come and check it out."
Zhang Wenduo frowned. An academic lecture, turned into such a mess.
They didn't know the true value of Glutinous Rice Ball.
It's not just a Large Language Model, nor just a chatbot—this Glutinous Rice Ball might be the key to unlocking the World Model.
Another half hour passed, and sure enough, the auditorium became increasingly crowded.
In addition to conducting research, Zhang Wenduo also served as a teaching assistant for Professor Shao during his doctoral studies. He knew many students in the School of Computer Science, and now, looking around, he found that there were indeed more people he didn't know than those he did.
Actually, besides computer science students, there are students from all sorts of majors, including finance, economics and management, and even journalism.
Many people aren't there for the content, but rather to see if the founder is as handsome as he appears in the videos circulating online.
By 2 p. m., the auditorium was full, and those who came in later either stood or sat on the stairs on either side.
What Han Luyi do?
"Didn't you see the news? The guy who makes Large Model released his Large Model last week, and it beat all those foreign ones."
And then there were those who came to join the fun.
At 2:25, the front door of the lecture hall opened, and Academician Zhuang and Han Luyi walked in together.
The volume in the noisy lecture hall immediately decreased significantly.
Academician Zhuang walked onto the stage, turned on the microphone, and began by saying, "Hello, students."
The room fell silent after those words were spoken.
Academician Zhuang gave a brief opening speech. "Today we have invited Lu Yi Han, the founder of Source Intelligence Technology, to have a discussion with everyone. Mr. Han has been very busy lately, and I am very grateful that he was able to take the time to come to the university."
Han Luyi walked to the center of the stage and took the microphone from Academician Zhuang.
A sparse round of applause broke out from the audience, but it quickly subsided.
"Hello everyone from Shuimu Forum University. I came in a hurry and didn't prepare any slides, so I'll just give you a verbal presentation."
Han Luyi looked down at the audience, his Sight opening up to the hundreds of people present. Their emotions, presented in different colors, converged before his eyes, mostly warm white, reflecting curiosity.
He discovered that, contrary to his expectations, only a small portion of the people present were computer science students, and even fewer were specializing in Artificial Intelligence.
He had originally prepared a very technical version, discussing the limits of Glutinous Rice Ball's abilities and training strategies, and the relationship between humans (the Research Institute) and intelligence (Glutinous Rice Ball). However, if he presented it in that way, most people in the audience probably wouldn't understand it.
Given this, Han Luyi made a bold decision to change the content and give an impromptu speech.
"As you all know, I work in AI, and I want to start by making a bombshell statement," Han Luyi, dropping a bombshell right from the start. "Within five years, half of the internet companies you know right now will disappear, and the remaining half will fully embrace AI."
These words sparked a small discussion among the students of Shuimu Forum, despite their extensive knowledge and experience.
Many people give speeches, but very few start with such strong opinions.
"Some might ask, why 'disappear' instead of 'transformation'?"
"I can explain my reasoning. I used to work at Dingsheng and have some experience with it, but this experience isn't specific to Dingsheng; it's a common problem in large companies. In large companies, everything has a history and baggage. Even if someone keenly sees an opportunity, they can't act immediately. The planning system requires them to go through approval processes, and during the approval process, their actions might infringe on the interests of too many people, so it ends up being left unresolved."
Han Luyi spoke, he observed the reactions of the audience, and their emotions were immediately reflected in his own.
These words were all experiences he had learned at Dingsheng, but the students in the audience lacked real-life experience and simply listened to them as stories.
On the contrary, Academician Zhuang nodded slightly.
"While all this was happening, startups had already rewritten their products using AI."
"In this situation, who do you think will win the market?"
Seeing that many people in the audience were nodding, Han Luyi took the opportunity to end his argument and move on to the next topic.
"I said five years, and many people might think I'm exaggerating. I'd like to ask you a question first: how many of you here have used AI tools? Including chatbots or other AI tools? Please raise your hand."
Han Luyi spoke, he raised his right hand.
After the question was asked, almost everyone in the audience raised their hands. It's not surprising that students at Shuimu University have a higher acceptance of technology than the average person.
Han Luyi changed the subject.
"So, please raise your hand if you've never used any AI tools before."
All hands were lowered; not a single person raised their hand.
Han Luyi looked around and moved on to the next topic.
"Does anyone know how long ChatGPT has been around since it was released?" Han Luyi asked, then answered himself, "Three years."
"It's only been a few months since the Glutinous Rice Ball was launched. How come we were able to create a model that's almost on par with OpenAI in just a few months?"
Han Luyi saw that the audience's curiosity had been piqued by this question.
"Because AI itself can help the development of AI. Glutinous Rice Ball's current generation model will also accelerate the development of the next generation model. This self-accelerating flywheel is real, and it exists in this industry."
Ray Kurzweil once proposed a Singularity Theory, giving the term a different meaning than in physics. He suggested that human civilization would accelerate dramatically and completely change its form after crossing a certain point. Is this theory true? Where exactly is the singularity? I don't know, but I can sense the accelerating trend in Artificial Intelligence research.
Five years ago, no one could have predicted that we would become so accustomed to communicating with AI robots as we are now. Three years ago, no one could have predicted that there would be so Large Model in the world, with Glutinous Rice Ball being the number one one. What will happen in five years? I can say with certainty that the impact of AI on our lives will be far greater than the changes of the past five years.
Han Luyi's words made the audience members sitting below the stage think.
Large Model have only been in the public eye as a tool for a short time, but their intrusion into our lives has been too rapid, seemingly making people forget what life was like before the AI era.
"Besides the changes to the industry, I also want to say something that might make you uncomfortable. Many of the professions we are familiar with now, especially those that require computer operation, will disappear in the next five years. Junior translators, junior legal document writers, and junior software developers—these jobs are already being reshaped by AI, and this is happening right now."
Han Luyi said this, not only because the viewpoint was heavy, but also because his mood was just as heavy.
"Some might say that new technologies always create new jobs, just as the Industrial Revolution did, with car drivers replacing coachmen and assembly line workers replacing artisans. That's true, but there's a difference: the Industrial Revolution replaced physical labor, while this time, some of our mental labor is being replaced, and at a faster pace than ever before."
"Ladies and gentlemen, the workplace we are about to enter is unlike any other in history."
"Going forward, there's only one thing AI can't do in the workplace: take responsibility."
"AI can help a Lawyer draft all the legal documents, but the Lawyer signs them, and the person responsible is the one who ultimately bears the consequences. AI can help a junior engineer complete all the code, but if the system goes live and problems arise, the person responsible is still the one who gets held accountable."
The audience was completely silent, and Han Luyi could see that their mood had also become somber.
"AI can replace tasks, but it cannot replace responsibility."
"This is not a bad thing. On the contrary, when AI takes over the execution part, what remains for humans is a purer judgment, a clearer value choice, and a trust relationship that is more difficult to replace."
"I'm not trying to sell anxiety. Let's get back to today's topic: <The New Relationship Between Humans and Intelligence in the Age of Artificial Intelligence>."
What I really want to discuss is the value of human beings.
"Previously, we could say that AI could only replace entry-level jobs, not advanced ones. At that time, this assertion seemed correct. AI could write code, but it couldn't solve engineering problems. AI could draw diagrams, but it couldn't create product posters. AI could write short stories, but it couldn't create long literary works."
"These were the limitations of AI at the time, but now, with the release of Riding the Wind and Glutinous Rice Ball, the engineering capabilities of AI have greatly improved. Some of our colleagues can achieve the highest output even without writing any code for a month."
"As AI evolves at an increasingly rapid pace, will these limitations persist indefinitely?"
Although it's not about selling anxiety, these facts clearly do cause anxiety in people. Han Luyi glanced around at the audience's emotions and his anxiety intensified.
"I think the question of comparing ourselves to AI is fundamentally flawed. Human value doesn't lie in comparing our abilities to AI, but rather in the fact that, as human beings, we possess certain values that AI can never replace."
When you're feeling down, you want a loyal friend by your side; when you're exhausted, you want your life partner by your side; when you're old and frail, you want your children to visit you, not a robot to take care of you.
"I think that in this new era, the trust between people, the feelings between people, and the connections between people are the most precious things."
Thank you everyone.
The lecture ended with enthusiastic applause in the auditorium.
Not everyone agrees with the views expressed in Han Luyi's speech, but at least the speech has prompted everyone to think.
As Academician Zhuang clapped, he silently sighed in his heart.
He somewhat agreed with Han Luyi's assessment; however, bringing up the point about career in this setting might have a significant impact on the students.
Actually, his invitation to Han Luyi to give a speech was more about broadening the horizons of undergraduate and graduate students in the School of Computer Science and providing them with inspiration for their future career choices.
Today is not bad, is it?
The applause continued, and Han Luyi looked at the audience below the stage.
Although he originally wanted to take this opportunity to promote Source Intelligence and recruit people, after changing the content at the last minute, he spoke according to his feelings and did not leave any room for this part.
However, this feeling is not bad, as it allows me to express some of my thoughts.
No wonder celebrities love giving lectures everywhere.
After a long while, the applause gradually subsided, and the lecture moved into the Q&A session.
A student council volunteer stepped forward, took out a microphone, and prepared to hand it to the student whose name was called.
Zhang Wenduo was the first to raise his hand. Although the content of the speech was different from what he had expected, he had come with questions.
Many hands were raised, but after scanning the room, Han Luyi still chose Zhang Wenduo.
After all, they had shared a meal yesterday, and for his mentor to bring him to that occasion shows that he is highly favored by his mentor.
The next morning, Zhang Wenduo seemed to have improved his emotional intelligence. After standing up, he first praised Glutinous Rice Ball: "Boss Han, I was fortunate enough to obtain Glutinous Rice Ball's API and conducted detailed tests. Glutinous Rice Ball is indeed currently the number one in terms of comprehensive performance of Large Language Model. I have a question: Do you think that with Glutinous Rice Ball's current training strategy, pushed to its limits, it can achieve AGI?"
He didn't explicitly state that Glutinous Rice Ball's training strategy focused on improving data quality; this was something Han Luyi revealed at the dinner table, possibly a trade secret, but both of them understood what he meant.
Han Luyi: "Thank you, student. You asked a rather technical question. The entire industry is still in the exploratory stage regarding the implementation of AGI, and I can't guarantee which strategy is right. In my personal opinion, while Glutinous Rice Ball's current level is quite good, this kind of improvement based on data quality is reaching its limit. If we want to achieve truly general-purpose Artificial Intelligence, we still need interpretable World Model to emerge."
Zhang Wenduo was taken aback. Han Luyi was also exploring World Model?
Isn't that exactly my field of research?
He handed the microphone back to the volunteer, and it was passed to the next student.
The person who stood up was a slightly Pangpang girl.
"Boss Han, I'm a fan of yours!"
As soon as the girl took the microphone, she began to speak, which was met with laughter from the audience, but she didn't seem to care.
Han Luyi said, "Hello fan, what's your question?"
"I studied journalism, does that mean I won't be able to find a job in the future? Can you please stop developing AI?"
The audience erupted in laughter again.
Han Luyi chuckled twice before answering.
"Of course, I could stop research and development right now, then issue a statement saying that AI is too dangerous, calling on all research institutions to stop developing AI, and then everyone would comply, sealing away all AI-related technologies, and we would continue to use the technologies passed down from the industrial age."
The audience laughed even louder, clearly knowing that this was impossible.
"However, there might be a small problem here," Han Luyi said. "After we issue the announcement, others won't stop; instead, they'll quickly overtake us. So your problem isn't solved."
"What should I do? Can I still continue working in journalism?" the girl asked.
Han Luyi asked, "Classmate, what was your reason for choosing to study journalism?"
The girl was taken aback by the question. The laughter around her hadn't made her blush at all, but the question itself made her hesitate. After a few seconds, she answered, "I want to be the conscience of society."
This time, no one laughed.
Han Luyi said, "Since that's the case, although I don't know if there will still be journalism work in the same form as now in the future, I'm certain that in the age of AI, you can play a more important role. Because when AI can generate all kinds of fake news..."
In an era of fake images and videos, a reliable source is more important than ever. And being trusted is precisely where a person's true value is most evident.
The girl, unsure whether she was satisfied with the answer, sat down.
After answering a few more questions, the Q&A session ended. Although the students were still very enthusiastic, this segment couldn't continue indefinitely.
Academician Zhuang returned to the stage. "Thank you, Mr. Han Luyi, for your insightful views. I hope they have given you some food for thought. I noticed that many students from other colleges are here today, which is very kind of you. Our School of Computer Science invites people from all walks of life to give lectures here every Monday afternoon. Next week, we will be hosting Li Yan, the founder of Seek Him Technology. If you are interested, you are welcome to come at the same time."
Then Academician Zhuang took Han Luyi and left first, without keeping him. Many girls who were eager to take photos with him sighed with regret.