```
Even so pessimistic predictions have been made, asserting that, due to the limitations of human lifespan and physiological rules, scientific developnt on Earth would stagnate for a long period of ti.
Unless there is a major breakthrough in biotechnology that significantly extends human lifespan, or at least slows down the aging process. Otherwise, relying solely on the current human lifespan, which barely lasts a century, just to learn from predecessors' theories and integrate knowledge, one would need to spend an entire lifeti studying, thinking, and at the peak of their creativity.
Furthermore, to achieve anything on top of that foundational knowledge is imnsely difficult, as one can imagine.
If the standards for publishing a paper were really that high, those scientific journals that mainly publish theoretical results might as well beco annual publications, as there likely wouldn't be enough content to fill their pages.
Therefore, to address this issue, Qiao Ze planned to use the Qunzhi Frawork as the underlying logic to design an intelligent assistant that could help him quickly index academic papers.
Qiao Ze's requirents for this intelligent assistant were that it could rapidly crawl the main content of journal papers on the internet and intelligently tag each paper according to its different values, incorporating the most valuable articles into the system database for his perusal at any ti.
For papers of lower value, it could independently summarize their valuable parts for direct archiving, while categorizing others deed worthless for imdiate blocking.
In this way, after the system had been running for so ti, Qiao Ze would no longer need people to painstakingly pan for gold in excrent.
Often, what they panned for was not gold, but rely sand.
...
"Qiao Ze, are you still busy?"
At four o'clock in the afternoon, when Su Mucheng arrived at the research group's new location after her supposedly confidential course, Qiao Ze was still busy at his computer.
The web crawler program had already been written, but it still required thought to endow the software with the ability to make intelligent judgnts.
"Yeah, I probably won't be able to finish it today."
"It's okay, there's no rush. Tan and Liu took the day off for themselves today, aren't you going to take a break?" Su Mucheng said as she walked up to Qiao Ze, then saw the programming interface and the familiar, simple dialogue box.
"Huh, you're working on Qunzhi again?"
"Yeah, I'm planning to make an intelligent assistant."
"Like Xiao Ai and Yi?"
"It has to be smarter than those."
"Hey, that would be great, Qiao! Our Qunzhi was smart enough before, so just make a smarter one, and ideally, add a voice module so it can play gas and chat with like a smart speaker. That way, when you're too busy to pay attention to and I happen to be bored, I'll have sothing to do."
"Well, when I'm busy, you can study, watch movies, watch videos. How could you have nothing to do?"
"But watching movies alone can be boring, too. If I had soone to discuss the plot with while watching, it would make the experience twice as enjoyable. For instance, if I'm watching a mystery movie, your intelligent assistant could analyze the plot with ; it would be like you were watching the movie with . That would be so much fun, wouldn't it?
I can't possibly ask Chen Yiwen, the exceedingly dull one, to discuss the plot, right?"
Carrying a bunch of fruit and following Su Mucheng into the room, Chen Yiwen wore a puzzled expression and couldn't help but retort, "How am I dull?"
"You've never even watched Agatha. And you say you're not dull?" Su Mucheng glared at Chen Yiwen, speaking with conviction.
"Qiao Ze probably hasn't watched it either. How co he's considered interesting?" Chen Yiwen muttered.
"First off, how do you know Qiao hasn't seen it? Second, even if Qiao hasn't, it's because he can guess the ending just from the beginning, so he doesn't bother; can you do that? Third, even if Qiao guesses wrong, it must be because the author's clues themselves have logical flaws. Do you have the confidence to question that?"
Su Mucheng delivered her explanation with airtight logic.
Chen Yiwen decided to stay silent.
Only a fool would try to reason with a woman in love; he almost made that mistake.
"Okay." Qiao Ze casually agreed after the two had finished their argunt.
...
Qiao Ze was willing to satisfy Su Mucheng's small request.
After all, for the Qunzhi Frawork, which had already preliminarily passed the Turing test, implenting these minor features wasn't too troubleso—certainly much less so than enabling the machine to discern which papers had value and which didn't.
This was not just about simple keyword indexing; the machine needed to reject content that was excessively homogeneous through countless comparisons of academic paper contexts.
Put simply, to realize Su Mucheng's desired features, the machine only needed to "know."
But to achieve the functionality Qiao Ze envisioned, the machine needed to possess a certain degree of "understanding."
Don't underestimate these two words; the difficulty of realizing the latter exponentially increases, because it involves a problem of computational complexity.
Before the Qunzhi System incorporated causal models and decoupling design into machine learning algorithms, the so-called machine learning basically ant pouring a bunch of data into a frawork made up of linear algebra components in a manual manner, then silently waiting for the system to produce a result.
If the result was correct, OK, the operation ended.
If the result was wrong, then a stick would be thrust into the pile of data to stir things up until the machine finally produced the right outco.
Whether it was the Alpha Dog that had once dominated the world of Go or the recently popular ChatGPT4.0, the technological ans they employed were rather crude—simply more advanced algorithms than before.
```
Reviews
All reviews (0)