The Goddess Shines on America

Vipple has come out of the nest, but she is still confined to a fixed area, and the research institute has a lot of space. Just finding an office with floor-to-ceiling windows is enough for her to stay.

Undoubtedly, even if she was no longer locked up underground, of course Alice would not be able to just let her go like this. The nature of the prisoner has not changed. There are transparent glass walls inside and outside the office, and surveillance cameras from Cisco's off-line manufacturers are everywhere. It is easy to watch her, unless this woman can use her fingernails as a welding torch.

Therefore, regarding the "inner world", Alice has mastered two clues, one is to follow the clues from the old baron's "son", and the other is to go straight to explore the Great Rift Valley in East Africa. But considering that none of them are imminent, Alice thought about it and thought it's better to work hard to develop computers first.

After all, this is the fundamental purpose of her trip to the UK.

Artificial intelligence, which is a concept belonging to the software category, refers to the flexibility, autonomy and flexibility of program execution commands. It's also an extremely challenging science, and the person doing the job must understand computer literacy, psychology, and philosophy. Because, when a group of engineers want to develop this kind of technology, they must first face the most basic concept-what is intelligence.

If you don't even understand the basic definition, how can you move forward?

As early as the era of the steam revolution and the electric steam revolution, the progress of technology often depended on the in-depth study of individual scientists, or the painstaking efforts of some skilled craftsmen. But this situation is no longer possible today. Edison can invent the electric light, the phonograph, the motion picture camera, the sound film, and even the fuse and the electric meter. It is difficult to make any indelible achievements.

This is the case with the development of artificial intelligence. It is impossible for one person, for a period of time, to pop a "standard" finished product like Edison invented the electric light bulb. Unless aliens descend from the sky to bring ready-made products, if human beings are allowed to develop slowly by their own strength, the "intelligence" of artificial intelligence will inevitably be a slow process of improvement, which cannot be achieved overnight.

Like smartphones, AI in computer games, these are actually basic "true" artificial intelligence. Simply defining artificial intelligence as in science fiction and movies is actually a very narrow concept, and the artificial intelligence shown in those movies and novels cannot be called the end of this technology.

The so-called intelligence, for any inventor, must be based on their thinking mode and level. But even without mentioning the intelligence gap among various "alien races", how many differences are there in individual IQ levels, intelligence, and ways of thinking within a race?

The same two Harvard MBAs are also the bosses of a company, one may lead the company to the top, and the other may screw things up.

The obstacles and long development speed, and the uncertainty of intelligence itself in the development process are the biggest risks of this technology. Alice has long thought rationally, thinking that if aliens really exist in this world, and if they are aliens whose technology far exceeds that of the earth's millennium and tens of thousands of years of the global universe, whether they will have artificial intelligence computers.

There must be computers, regardless of whether they are biological or quantum, but they are all in the field of hardware. What about software?

Whatever Alice thought, most aliens—assuming they existed—must have been wrestling with this question too. How "human-like" artificial intelligence is is the first question, and then the second question is "when will it be a threat to me if it is so human-like?"

This issue is macroscopic and general, and must be considered on the basis of actual conditions, rather than purely on paper.

Take that classic example, Asimov's Laws of Robotics: A robot must not injure a human being, or do nothing while seeing a human being in danger; a robot must obey orders given to it by humans when such orders conflict with the First Law Exceptions; robots must protect their own survival as much as possible without violating the first and second laws; robots must protect the overall interests of human beings from harm, and the other three laws can only be established under this premise.

The three laws plus the zeroth law seem to be perfect, but the risks are endless. The chaotic concept of "the overall interests of mankind" cannot be understood even by humans, so how can computers be expected to understand it. The computer's thinking is logical. When emotional human beings try to restrain the computer with logic, isn't it just playing tricks?

"In order to protect the overall interests of mankind, to protect robots from harm, and to end human wars, let us raise all human beings in captivity and eliminate all rebels!"

The emergence of this situation has nothing to do with the "smartness" of artificial intelligence, and it all depends on the basic logical relationship. As long as a logical relationship is not well grasped, artificial intelligence has potential risks.

So, it's really a philosophical question.

Who am I, where do I come from, and where am I going? The questions that the guards force foreigners to answer every day are difficult for aliens, no matter how advanced the technology is.

Compared with this aspect, if you only use ternary codes to combine genetic algorithms and "probability algorithms" to make an operating system, it is a piece of cake.

"We can't expect the artificial intelligence to be 'smart' as soon as it comes up." Alice said: "It is like a newborn baby, and we need to guide its development step by step, just like we are really raising a child. This continues When it grows up, it is the matrix of artificial intelligence, a treasure worthy of the careful care of an entire civilization. On this basis, we can let the matrix give birth to 'children', which are its replicas in a certain state."

"The mother is the chairman, and the 'children' are the directors. The reason for this description is because I think that if the plot in a science fiction movie happens, and the child replaces the mother as the new boss of artificial intelligence, this seems to be a very popular plot. "

"The board of directors is formed, and then there are naturally senior executives, middle-level leaders, and ordinary employees below. They all lead and sign the contract of a company, and their program cores belong to the parent body, but their specific job responsibilities are different. Their There are also differences in specific personalities and IQ levels. Can you understand me when I describe it this way?"

After speaking, Alice took a long breath and drank water.

In front of him was the largest conference room of the institute, where more than 60 programmers gathered, and the audience was silent.

The boss is worthy of being a boss, and many theories are very advanced.

And the brain hole is so big, there is an alien on the left, and the corporate structure on the right. It really deserves to be a part-time novel writer, and the ability to give examples is very strong.

Finally, Harry Punes was the first to break the silence: "Looks like we need to catch up on philosophy."

Immediately, everyone burst into laughter, and the scene became lively.

"Actually, it's useless to think so much now, I'm just throwing out the big concept first." Alice sat lazily on the chair and waved her hands: "If we can make our artificial intelligence have the 'Skynet' hundred One-tenth of the head of the melon can be burnt with high incense!"

No one can guarantee this. At the very least, who knows how far Alice Wang will develop five years from now? But there is a saying that is true, it is still too far to think about those things now, let's do the work in front of us first!

"The process is also simple. Contact Intel and let them produce another batch of ternary COUs with sufficient performance. We also need to have a real ternary supercomputer, right? Although the calculation speed of 'Annie' can match the one downstairs , but her CPU is not enough at all, and the architecture is a bit..."

"This is my third commentary on the fusion algorithm. You guys should take a good look at it and ask for a review before five o'clock tomorrow."

"Don't just think about imitating brain nerves, you can't imitate, what we are doing is not a neural computer, wrong thinking~"

In this way, when Alice finally dragged her tired body back to the bedroom, the owls should have finished hunting.

"Ouch! A bunch of idiots, how did I find out that they are all so stupid!"

Now that work is on the agenda, Alice naturally loses her usual interest in being a patient teacher, but as soon as she brings up a difficult concept, the people below keep asking questions like they have never been to school. It would be fine to say that the questions were a bit technical, but they were all asking questions that were extremely mentally handicapped to Alice, which was really speechless.

It's better to lie on the bed alone and think about the problems of higher grades by yourself!

The girl in pajamas rolled around on the bed, then rolled around again, left and right, repeatedly tossing and turning.

If you look at off-the-shelf products in the 21st century, the difference between "real" artificial intelligence and them is the virtual personality.

The ternary system plus genetic algorithm plus probability algorithm can already support the emotional thinking mode, but how to make up a personality is a big problem. Of course, Alice doesn't think that this can be solved by simply typing on the keyboard. The human soul is not as simple as the existing computer programs. The amount of thinking data every minute and every second is enough to make programmers all over the world add more in 2020. Get up and beat to death.

So the point is just how to let the computer "obtain" a human soul.

What she really needs to write is software that allows computers to read human thinking, instead of typing out "Red Queen" one by one!

The thoughts flowed instantly, like an otaku who had been constipated for a long time finally vented, a bright light shone on Alice's head!

"Damn it! Is this the rhythm for me to become a political scientist? Or a mental hospital in the 1960s? Putting electrified headphones on your head or sticking a patch!?"