Chapter 258: Everyone Has Their Own Obsession

Chapter 258 -195: Everyone Has Their Own Obsession

Pan Si’s tone grew heavy as he spoke the last three words, enunciating each syllable.

Bai Qian’s face fell the moment he heard this.

“Dr. Pan, what do you mean by that?”

His voice was clearly tinged with anger.

“My architecture is not some toy model!”

“I’ve already verified its feasibility on medium-scale datasets!”

Hearing this, Pan Si sneered.

“Medium-scale? How big is that? A million? Ten million?”

“Dr. Bai, real large-scale AI models start with training data in the trillions of tokens.”

“Your little ‘medium-scale’ isn’t even in the same league as real large-model training.”

Bai Qian was left speechless, his face flushing red.

“So you’re saying we should just keep using an architecture like Transformer, which has already been proven to have significant flaws?”

“Yes!” Pan Si nodded without hesitation.

“Transformer has its flaws, but it’s been validated by the industry countless times.”

“Its training pipeline is mature, its engineering implementation is simple, and there’s a massive amount of open-source code and tools we can use directly.”

“We’re a startup, not an academic institution.”

Pan Si’s tone became incredibly serious.

“Our goal is to build a viable, competitive product as quickly as possible, not to publish papers and win awards.”

“Therefore, I believe we should focus our energy on data cleaning and model alignment, not waste time on architectural innovation.”

After hearing this, Bai Qian was so angry he laughed.

“Waste time?”

He repeated the two words in a tone of utter disbelief.

“Dr. Pan, do you have any idea what you’re saying?”

“Architectural innovation is the true core competency in the field of AI!”

“If we just follow behind giants like CloseAI and Doggie, using the architectures they’ve already validated, then we’ll forever be followers and never catch up to them!”

“Only by making a breakthrough in architecture can we possibly leapfrog the competition!”

Hearing this, Pan Si snorted coldly.

“Leapfrog the competition? Dr. Bai, that’s easy for you to say.”

“But have you ever considered that to leapfrog the competition, you first need a car that can actually run?”

“If you don’t even have a car, what are you going to overtake?”

“We don’t even have a single viable model yet, and you’re already thinking about architectural innovation?”

“Isn’t that putting the cart before the horse?”

The two went back and forth, arguing until they were red in the face.

Sun Yu, Huang Yiyi, Yunqing, and Shen Yan, who were sitting in on the meeting, were all completely dumbfounded.

They never would have imagined that their very first technical meeting would turn out like this!

Chen Lin sat at the head of the table, his brow furrowing tightly as he watched his two tech titans argue heatedly.

He had originally thought that by recruiting these two, they could join forces as a powerhouse duo and rapidly advance the AI project.

He never expected their technical philosophies to be so diametrically opposed.

And from the looks of it, neither side was willing to back down.

Bai Qian insisted on an “architectural revolution,” believing that only by fundamentally changing the model’s underlying architecture could they create a truly competitive product.

Pan Si, on the other hand, advocated for a “data is king” approach, arguing they should build upon existing, mature architectures and focus their energy on data quality and model alignment to quickly produce a viable product.

Both viewpoints had their own merits.

The problem was, a project can only have one technical path.

If they couldn’t reach a consensus, the project couldn’t move forward at all.

Chen Lin was silent for a moment, then suddenly raised his hand in a gesture to calm them down.

“Alright, both of you, let’s take a pause.”

His voice wasn’t loud, but it carried an undeniable firmness.

At his words, Bai Qian and Pan Si both stopped and turned to look at Chen Lin.

Chen Lin began to speak slowly:

“I understand both of your viewpoints.”

“Dr. Bai believes we should innovate on architecture, while Dr. Pan thinks we should use a mature architecture to get a product out quickly.”

“Your reasons are both very compelling.”

He paused, his gaze sweeping over the two of them.

“However, I’m not going to make a final decision on which approach to use right now.”

Hearing this, both Bai Qian and Pan Si were slightly taken aback.

Chen Lin continued:

“I want each of you to go back and prepare a detailed technical proposal.”

“Dr. Bai, you will detail your new architecture, along with the projected training pipeline, resource requirements, and timeline.”

“Dr. Pan, you will also detail your plan based on existing architectures, including the data cleaning process, model alignment methods, and expected results.”

“Three days from now, we will meet again to discuss.”

At this point, Chen Lin’s tone became even more solemn:

“At that time, I will make a final decision after thoroughly considering the pros and cons of both proposals.”

Bai Qian and Pan Si glanced at each other. Although both were somewhat displeased, they could only nod in agreement.

“Alright, President Chen.”

“Understood, President Chen.”

Seeing that they both agreed, Chen Lin finally breathed a sigh of relief.

“Then let’s end today’s meeting here.”

He stood up and looked at everyone present.

“Everyone is dismissed.”

Sun Yu, Huang Yiyi, Yunqing, and Shen Yan immediately scrambled out of the conference room as if granted a pardon.

Bai Qian and Pan Si also gathered their things and walked out, their faces grim.

Soon, Chen Lin was the only one left in the conference room.

He sat in his chair, rubbing his temples, feeling a headache coming on.

‘Managing a team really isn’t an easy task. Especially a team composed of tech geniuses like these. Everyone has their own convictions, and everyone thinks they’re right.’

Chen Lin sat in the conference room for a while longer, then suddenly remembered something. He took out his phone, opened his contacts, found the number for Xiang Wentian’s assistant, and dialed it.

The phone rang a few times before it was answered.

“Hello, Mr. Chen?” came the voice of Xiang Wentian’s assistant.

“Hello, Assistant Hang, this is Chen Lin,” Chen Lin said politely. “Is Elder Xiang available to take a call? I have something I’d like to ask his advice on.”

“Please wait a moment while I ask Elder Xiang.”

A rustling sound came from the other end of the line. A moment later, Xiang Wentian’s familiar, slightly weak voice came through.

“Ah, Chen Lin, what is it?”

“Elder Xiang, my apologies for disturbing your rest,” Chen Lin said apologetically. “I’ve run into a bit of a problem and was hoping to ask for your advice.”

“Go on,” Xiang Wentian said in a gentle tone.

Chen Lin then briefly recounted what had happened in the conference room.

After he finished, he sighed in frustration:

“Elder Xiang, I’m not sure what to do now.”

“Dr. Bai and Dr. Pan are both exceptionally talented, but their philosophies are just too different.”

“I’m worried that if I don’t handle this properly, the team might split apart.”

The other end of the line went silent for a moment.

Then, Xiang Wentian’s voice slowly came through:

“Ah, Chen Lin, the problem you’re facing is actually quite common.”

“I ran into it often when I was younger.”

His tone carried a hint of reminiscence.

“People in scientific research and cutting-edge engineering, especially the truly skilled ones, often have their own brand of stubbornness.”

“They become convinced their technical path is the right one, and they’re very difficult to persuade otherwise.”

“But,” Xiang Wentian’s tone shifted, “as the team’s leader, you can’t let them lead you by the nose.”

“You need to know clearly what you want, what your goal is.”

“Then, you establish the technical path based on that goal.”

“If both of their proposals have merit, then find a way to combine them.”

“If they absolutely can’t be combined, then you proceed in phases—which one comes first, which comes second. You need to have a clear plan.”

Chen Lin listened to Xiang Wentian’s words, nodding thoughtfully.

“I understand, Elder Xiang.”

“Mm,” Xiang Wentian affirmed, then added, “There’s one more thing to remember.”

“Tech geniuses often need a clear goal to constrain their creativity.”

“If you don’t set boundaries for them, they’ll just wander further and further into their own worlds, and what they end up creating might be completely different from what you originally intended.”