Top academic

For a moment, the scene was completely silent.

Tao Xuanzhi, sitting in the first row, was equally shocked. However, a long-standing habit made him subconsciously nod slightly to Qiao Yu in greeting…

The example given by Qiao Yu actually belongs to the predictive processing theory in cognitive science.

Although our understanding of the human brain is still very limited, there is no doubt that it is a very powerful predictive machine.

It can continuously generate predictive models for the future based on past experience and real-time information received. Or, in other words, it generates a picture of the future.

Humanity's unique creativity also stems from this. After all, if everything required rigorous calculation, there wouldn't be so many mathematical conjectures driving the entire system forward.

According to Qiao Yu, it seems that he has been researching how to enable machines to surpass the execution of preset algorithms and gain the intuitive ability to instantly integrate complex information, generate meaning, and make value-driven choices, similar to that of humans.

WTF? !
This is clearly no longer a simple mathematical problem. Of course, it still involves mathematics.

Ultimately, to achieve this, a mathematical foundation is required. The reason is simple: the essence of a machine is reading 0s and 1s.

At least in this day and age, any artificial intelligence implemented on silicon-based computers, no matter how advanced or sentient it claims to be, or how powerful its algorithms are...

Its underlying operation necessarily relies on mathematically defined algorithms and data.

There's no way around it; without algorithms, computers are nothing.

So, what Qiao Yu meant to many people was probably to open up a mathematical framework that was anti-mathematical, so that artificial intelligence could break free from the constraints of algorithms and achieve instantaneous reasoning and decision-making capabilities similar to the human brain.
Hmm, the anti-mathematical mathematical framework...

Undoubtedly, this concept sounds very convoluted. Many people have already realized why the report included the term "meta-mathematics"...

"Before I formally introduce my results, I would like to share my thought process with you."

First, we need to analyze human consciousness. That is, as I just mentioned, why can consciousness make behavioral decisions that align with optimal expectations without relying on calculation?

During our group discussion, some members used quantum theory to analyze this. They argued that at the level of consciousness, the most advantageous state is instantly selected from all possible superposition states.

Understanding it from this perspective is obviously much easier. After all, the framework of quantum physics is relatively mature.

We already have the hardware foundation for quantum-like analog computers. Solving the problem from this direction would undoubtedly be much easier.

But I don't think so. First of all, there is no superposition state at the level of consciousness; all choices are definite states.

Moreover, if we understand the decision-making process of consciousness from the perspective of quantum physics, the computational process cannot be omitted.

Please allow me to emphasize again that consciousness does not need to rely on any formula for calculation; it simply relies on experience to select the optimal outcome of future possibilities.

In other words, consciousness is only responsible for prediction; calculation is a process that occurs after prediction. Let me give you another example.

Everyone knows that China has many online shopping festivals, such as 618, Singles' Day, and Double Twelve. Xida University also has similar offline shopping seasons, such as Crazy Friday.

For consumers, when these discount seasons approach, they instantly realize that they can save more money by buying things during the discount season.

However, in practice, because merchants and platforms have various preferential policies, the brain needs to calculate to find the most favorable way to buy.

Do you all understand? Consciousness is responsible for making instantaneous predictions and providing the optimal choice, and then the other parts of the brain begin the specific calculations.

Similarly, when artists suddenly have an inspiration, or when a melody comes to mind for a musician, the first thing to do is to write it down, and then modify the melody accordingly.

A painter would pick up a brush and first sketch the general outline before starting to add details; a writer would jot down the outline and then begin refining it…

So do you all understand now? Decisions are made instantly, while calculations are deferred and serve the purpose of optimizing the decision.

But is consciousness really that simple? Obviously not! I believe everyone here feels the same way: when you consciously begin making optimal choices, your subconscious automatically ignores those choices that come at a huge cost!

Just like the example I gave at the beginning, when a young man sees his ideal partner, he thinks about the wonderful married life he will have in the future. He might even be thinking about the names of his future children.

But he certainly wouldn't have imagined that he would use a method that violates ethical standards to quickly realize the scenario he had planned in his mind.

In mathematical terms, consciousness is still subject to constraints when making decisions.

These constraints vary depending on an individual's level of education, family environment, upbringing, and personality.

Moral constraints, legal constraints, emotional constraints, and so on... Of course, there are also the different hormonal stimuli that humans experience at different ages, so these subconscious constraints cannot be quantified at all.

But without a doubt, these constraints are also part of conscious decision-making. In fact, they can be said to constantly influence the decision-making process of consciousness.

I believe that by now everyone should have already formed a mathematical definition of consciousness in their minds.

In my view, consciousness is mathematically defined as a constrained instantaneous optimal feasible region selection predictor.

At the same time, we can also conclude that consciousness has the following mathematical characteristics: First, the decision time Δt approaches zero infinitely;

Second, it possesses a constraint-driven feasible region; third, it includes probabilistic evaluation based on instantaneous experience and selection of the optimization objective.

Therefore, in response to these characteristics, my team and I developed a completely new mathematical model that prioritizes prediction over computation and selects options beyond optimization.

This is the rudimentary form of meta-mathematics that I will focus on discussing today. We call it the Empirical Constraint Instantaneous Optimization System, or ECIO framework for short. You can see its basic structure on the screen.

As soon as she finished speaking, the large screen that had been prepared behind Qiao Yu lit up simultaneously.

This was the first time a PowerPoint presentation had been used. As is typical of Qiao Yu's presentations, it was simple and direct.

Not much explanation.

This is also the reason why a question was included when sending out the invitation in July.

The mathematical symbols above are all directly referenced from the concepts in Cho's algebraic geometry.

Obviously, this is very unfriendly to mathematicians who have not specifically studied and researched the axiomatic system of generalized modalities and the geometry of Georg algebras. Even if they come, they will not understand anything, which is tantamount to wasting their time.

Not to mention that although these formulas look simple, the derivation process is quite complicated.

While the ultimate goal is to transition artificial intelligence to general artificial intelligence, eliminating the need for parameter tuning and rigorous, complex derivations and calculations in decision-making, mathematics is, after all, mathematics.

The derivation supporting this decision tree process itself must be logical.

The next hour or so was undoubtedly the most boring for the more than one thousand mathematicians in the audience, especially the students and assistants who had come with their mentors to broaden their horizons.

There are no longer those vivid and interesting examples; it's all about mathematical explanations of conscious decision-making.

Especially the part involving modal space.

Unfortunately, the difference metric under feasibility measures inevitably needs to be introduced into a high-dimensional modal space. This is definitely the most abstract part of the entire theory. After all, discarding the traditional decision tree doesn't mean that general artificial intelligence doesn't need decision trees...

However, the decision tree required to achieve this kind of instantaneous decision-making is too abstract...

This is inevitable, of course.

After all, human consciousness is inherently abstract.

Moreover, when introducing these mathematical derivations, Qiao Yu showed absolutely no consideration for the feelings of the audience.

It's still his style; he always assumes that the mathematicians in the audience are at a similar level to him and will definitely be able to understand every step of the derivation.

Therefore, there was no detailed explanation; he simply recounted his reasoning at the time.

So, setting aside the half who were just there to observe, the remaining half, consisting of 90% mathematicians, also looked completely bewildered, frowning and seemingly not understanding at all...

Even the many prominent mathematicians sitting in the front row were in a similar situation.

For example, Tao Xuanzhi's furrowed brows never relaxed.

After listening for a while, he even took out his phone and started recording, which is quite rare for him.

Qiao Yu doesn't really care about this kind of behavior.

He was talking about theoretical things. And theoretical concepts really are universal.

Whether mastering the theory will lead to the creation of similar general artificial intelligence is a matter of opinion.

At least in Qiao Yu's view, without more than 100,000 engineers with relevant experience and quantum simulation computing resources available at almost no cost, designing another July is nothing short of a pipe dream.

The core code for July was written by Qiao Yu himself.

But there were too many codes on the periphery in July.

Each server handles different decision types, and different code must be uploaded according to the manual it has written.

It took a whole year just to train the relevant AI engineers to understand and use the instruction manual he wrote.

That's why the development of general artificial intelligence was included in the five-year plan at that time.

Otherwise, it would have been impossible to cultivate July. Even if all the money that Fine Rain Technology earned from selling patents were spent, it still wouldn't have been enough...

It's true that a generous reward will surely attract brave men, and while FineRain Technology has a lot of resources at its disposal, it's still not enough.

It's impossible for it to be online and running in July so quickly.

It can only be said that the weight of the phrase "the strength of the entire nation in China" is something that people from other countries can hardly imagine.

Finally, around 11:10, the obscure math part was finally finished.

Clearly, not many people at the scene could understand it; in fact, almost none could.

It's already good enough if a few people can understand half of it.

However, this does not mean that all mathematicians in the world are much less accomplished than Qiao Yu.

Let alone directly engaging with a completely new mathematical system, it's perfectly normal for mathematicians to encounter many unfamiliar things in the early stages when switching to a specific research field.

Even though the generalized modal axiom system has unified mathematical notation, the empirically constrained instantaneous optimization system proposed by Qiao Yu is still a new thing, and it is also the most cutting-edge research field.

It's unrealistic to expect to fully explain the results of two or three years of research in just over an hour.

Therefore, the content of the report meeting was already highly condensed.

The leading mathematicians sitting in the front row naturally had a host of questions and even doubts popping into their minds at that moment.

After all, the idea of ​​using rigorous mathematics to construct a consciousness system that is so abstract that it often does not conform to logic is simply too crazy!
Even if you only half understand it, it subconsciously makes you want to question the correctness of this theory.

Unfortunately, Qiao Yu directly discouraged these people from asking questions on the spot.

"Alright, that's all for the theoretical derivation of the mathematics for now. You may have many questions, but due to time constraints, I won't be able to answer them all today."

As I mentioned before discussing this proposition, the initial intention behind constructing this mathematical system was to express consciousness through mathematics.

The most direct manifestation of expressing consciousness through mathematics is undoubtedly artificial intelligence. In other words, the essence of metamathematics is to build a set of mathematical foundations that can be directly applied to general artificial intelligence.

This has always been my view: although mathematics is the most fundamental subject, every mathematician should serve the purpose of applying it.

Whether you're an applied mathematician or a theoretical mathematician! So after building this system, I led my team to conduct research on artificial general intelligence, also known as AGI, and we've already made some progress.

Originally, this information was supposed to be kept confidential for the time being. However, prior to this presentation, I obtained authorization to showcase some of the relevant findings.

These words once again left the mathematicians present speechless.

Especially those bigwigs who had initially intended to raise their hands in questioning. This even included Tao Xuanzhi…

There have already been results?

In other words, has Qiao Yu's team developed general artificial intelligence with preliminary consciousness attributes?

For a time, many people's brains were truly overwhelmed...

Keep in mind that this was a professional mathematics conference. Most of the guests were leading figures in the mathematics field.

Their brains are absolutely among the best in the world. But everyone also perked up.

Especially those students who are brought by their mentors to broaden their horizons.

Young people are naturally curious about cutting-edge things, not to mention that they couldn't understand anything Qiao Yu had said for over an hour.

But if you give a demonstration, they will definitely be able to understand it.

However, after a moment of silence, Qiao Yu's next words once again overturned the mathematicians' worldview.

"Actually, everyone should have dealt with them before. They were responsible for sending out invitations for this conference, including providing the topics."

That's right, it's July, the user of the email address mentioned in the answer to that question. It's also the first general-purpose artificial intelligence product developed by the Yanbei International Mathematics Research Center. (End of Chapter)