Skip to main content
  • Home
  • Library
  • Read
    • Popular
    • Charts
    • Browse
    • Categories
    • Wikis
    • Glossary
  • Request
  • Activity
  • Leaderboard
  • Help
Our Apps
  • DaoSearch
  • Narratorr
Social
  • Discord
  • Patreon
The AI Wave Tycoon's Rise
Chapter 266

Can't We Have It Both Ways?

1,863 words 6 min read 1 view 1mo ago Source ​
Skip to the end of the chapter
Prev
Next

Han Luyi sat at the desk in the hotel room, with a laptop on it, and Su Niannian's face on the video software.

However, Gu Siyue's voice could also be heard outside the video.

"Boss Han."

Han Luyi's heart skipped a beat.

Su Niannian at Gu Siyue's house?

When did their relationship become so good?

He took two deep breaths before speaking, "Just now, Lu Zhengping treated me, Academician Zhuang, and professors from the School of Shuimu Forum Science and Technology to a meal."

Su Niannian's expression also shifted.

Lu Zhengping wields considerable influence in Beijing.

Han Luyi's trip to Beijing was quite rushed; he left yesterday, Lu Zhengping was able to arrange a dinner of this caliber today, which is exactly what Han Luyi values most.

Han Luyi briefly recounted the dinner party at Shuimu Forum, and finally received an invitation to give a lecture.

The fact that Shuimu School of Computer Science was willing to let him give a lecture is itself an endorsement.

This could unlock an entire pool of talent in Artificial Intelligence.

After Han Luyi finished speaking, Su Niannian repeated the phone call from Haicheng.

"Deputy District Chief Chen didn't make any mandatory demands, but his attitude was clear: neither the district nor the city wanted Source Intelligence to leave Haicheng."

I made no promise, I only said I would pass it on to you.

Han Luyi: "Thank you."

Su Niannian then sent over the comparison chart of advantages and disadvantages that she and Gu Siyue had just compiled.

"This is what Si Yue and I just organized, take a look."

Gu Siyue also said off-screen, "It's obvious that both Beijing and Haicheng want you now. You need to make your own decision."

Han Luyi leaned back in his chair and rubbed his temples.

Then he straightened up and asked, "Can't we try to have our cake and eat it too?"

Upon hearing this, Su Niannian subconsciously looked at Gu Siyue.

Gu Siyue was also looking at her.

"Han Luyi is too greedy," Su Niannian thought. But this was something she and Gu Siyue had discussed.

If we want to structure Source Intelligence with a dual-headquarters architecture, how should the resources be allocated?

The most important part here is that the two sides may want different things, but how do you know what they want?

One wrong move and you not only fail to achieve the best of both worlds, but you end up offending both sides and losing everything.

"Si Yue and I have also discussed this. If what the two sides want doesn't overlap, then we'll put what the Beijing wants in the Beijing and what the Haicheng wants in the Haicheng. But how do we know what they want?" Su Niannian asked.

Han Luyi nodded after hearing Su Niannian's words.

Indeed, if you don't know what the other person wants, it's difficult to meet their needs.

But I know.

Brother Sight Shi told me everything.

Han Luyi opened the information Su Niannian had sent him.

What can Source Intelligence Technology offer?

The world's Large Model, with its Source Code applications as a source of commercial value, job opportunities, and the potential to become a city's calling card.

What is most important for Haicheng among these factors?

Commercial value includes revenue, taxes, employment, commercialization, and the industrial chain.

For an international business center like Haicheng, GDP is the most important thing, without exception.

GDP can drive the development of everything else.

As for the title of "Artificial Intelligence City," it's nice to have it, but it doesn't really matter if we don't.

But what about the Beijing?

Title.

If any city is to become the center of Artificial Intelligence in China, it can only be Beijing.

This is not something that can be discussed.

Therefore, since the two sides value different things, it is entirely possible to satisfy the needs of both sides and obtain the benefits from both sides.

"Haicheng prioritizes commercial interests, and this part can remain in Haicheng. Currently, Source Intelligence Company's main source of revenue is API calls to its Source Code, and in the short term, we are unlikely to have the capacity to provide external calls. This way, future revenue and taxes will all go to Haicheng."

"What the Beijing wants is completely different. In terms of importance, they don't care about those few extra tax revenues; what they want is the leading company in Artificial Intelligence."

After explaining his assessment, Han Luyi made his decision. "Design in this direction and divide the functions accordingly."

Gu Siyue's voice came from off-screen.

"Source Intelligence Technology can establish a second headquarters and R&D center in Beijing. Its greatest role will be to appear as the main R&D entity in academic papers and media reports. It can also serve as a platform for cooperation with various universities in Beijing."

Su Niannian also spoke up, "In Haicheng, the decision-making center, commercialization team, and productization team will all stay here; these are the departments that directly generate revenue. The R&D center in Beijing mainly handles the spending aspects; the R&D department is inherently a cost center and won't compete with Haicheng for tax revenue."

After Han Luyi the most crucial decision, the subsequent design process became quite simple.

However, none of the three people on either side of the video seemed to be considering what would happen if he was wrong.

Soon, the three of them finished all the necessary tasks, and Gu Siyue would be responsible for the legal details.

Su Niannian asked about another matter, "What do you think of Zhan Xin's investment?"

This inquiry was made entirely out of concern. Most of the shares in Source Intelligence Technology are still held by Han Luyi alone, and whether or not to accept this investment is basically a decision made by him alone.

Han Luyi said, "I'm ready to accept."

"Is it all coming in cash?" Su Niannian asked. "Have you thought about where you want to spend the money?"

Han Luyi and said, "I already had this idea. If we have enough money, we can speed things up. I've already assigned a task to Lu Mingzhou."

Su Niannian knew that Lu Mingzhou was going to Pengcheng on a business trip and was about to ask him about it.

Gu Siyue cleared her throat beside her. "Am I not suited to listen to the next topic?"

Su Niannian smiled, then glanced at her watch. "You're the company's legal counsel, no one would hide anything from you. But it's getting late, everyone should get some rest."

Han Luyi still felt a bit of excitement.

Meanwhile, in the doctoral apartment at Shuimu University, Zhang Wenduo had just returned from his advisor, Professor Shao.

When he took Professor Shao home, he got a severe scolding from Professor Shao.

--

Professor Shao said he acted without any sense of propriety, had low emotional intelligence, and would inevitably cause a major incident sooner or later.

Although Zhang Wenduo did not refute it to his face, he did not take it seriously.

Doesn't everything in this world require a pragmatic approach?

When it comes to scientific research, do we really need to consider personal relationships or social niceties?

I love my teacher, but I love truth more.

He didn't drink much during dinner, but he has a very low alcohol tolerance, and he felt a little dizzy while walking. He only felt better after walking back to his apartment.

Zhang Wenduo didn't plan to go to sleep immediately. He opened his email and checked the API key for Glutinous Rice Ball -0.9.

After Han Luyi processed it for him during dinner, the key was sent to his email.

He's now eager to expose this fake.

He copied the key from his email and opened his frequently used test script.

The personal assessment set is small, containing a total of 127 questions.

But it was all money he had saved up little by little over the years.

Some were boundary cases he discovered while reading papers, some were counterexamples he used to test his model in his own experiments, and some were quirky questions he specifically used to test Chinese intent understanding.

These questions have never been made public.

It's not on any leaderboard.

It's also impossible for them to be trained in advance.

Zhang Wenduo named the document -

kill-model-v11.1. jsonl (kill Large Model version 11.1)

Although the name sounds a bit childish, the quality is quite high. Many so-called world-class advanced Large Model will get very poor scores when running this set of questions.

Zhang Wenduo filled in the API address of Glutinous Rice Ball -0.9, pasted the key, and called the Large Model evaluation script he had written.

Run.

As he looked at the terminal output, he leaned back in his ergonomic chair and put his feet up on the table.

I still feel a little dizzy.

Upon closer inspection, I realized I had passed the first ten questions.

I passed all twenty questions.

Zhang Wenduo frowned and put his feet off the table.

Thirty questions, and all correct.

What happened?

Zhang Wenduo pulled his chair forward a little.

The script has run up to question 50 and is still getting them all right.

Zhang Wenduo felt a bit cold, touched his head, and found a thin layer of sweat on it.

Without continuing to monitor the running script, Zhang Wenduo clicked on one of the questions and looked at Glutinous Rice Ball's output.

The logic is very sound!

An expert's skill is immediately apparent. These questions, specially collected by Zhang Wenduo, emphasize the complexity of the tasks. Under such complex conditions, it is easy to become distracted and overlook key conditions if the general Large Model is used.

However, this question did not appear at all in Glutinous Rice Ball's answer.

Zhang Wenduo watched quietly as the script in the terminal finished running. Out of the last 127 questions, he got 120 right and 7 wrong.

But there are unusual things in these seven paths.

One of the questions was answered by Glutinous Rice Ball with "I don't know".

As a researcher of AI models, Zhang Wenduo knows how difficult this is.

Can this be achieved simply by improving data quality?

Zhang Wenduo suddenly felt his face getting hot, and it was getting hotter and hotter.

He touched his face. Was this alcohol really that potent?

I'm getting hooked again.

Zhang Wenduo picked up his cup, filled it with water, and drank it all in one gulp.

He took out his phone and found Han Luyi WeChat friend.

The chat log contained the message: "I've accepted your friend request. We can start chatting now."

Zhang Wenduo stared at the input box for a while before starting to type.

[Boss Han, I was a little...]

He stopped there.

What's wrong with it? Offense? Lack of tact? Arrogance?

He frowned, deleted those words, and retyped them.

[Boss Han, I just checked the Glutinous Rice Ball; my previous judgment was too hasty.]

That won't do either; it seems too arrogant.

Zhang Wenduo thought about it and decided against it; perhaps the other party wouldn't mind.

Such a high level of skill, yet so humble.

Well.

Zhang Wenduo lay on the bed, feeling his face burning, and thought to himself, "Why is the alcohol getting stronger and stronger?"

Go to sleep; the effects of the alcohol will wear off when you wake up.

Suddenly, he sat up in bed.

Where is Boss Han's lecture tomorrow? I absolutely have to go, I can't forget!

-

End of Chapter
Prev
NovelSummaries
Next

Explore the wiki

Han Luyi
Su Niannian
Gu Siyue

Comments

No comments yet.
AboutStatsAPIChangelogStatusContactPrivacyTermsDMCA
© 2026 RandomTranslator·v3.26·RSS