-
Add Bookmark | Recommend this book | Back to the book page | My bookshelf | Mobile Reading
Free Web Novel,Novel online - All in oicq.net -> Sci-fi -> Lost in the stars Volume 2 Weird Place Chapter 49 IBM¡¯s Joke
Previous page Return to Catalog Next page
- IBM chose a very good time, 9 a.m. on January 1, 2019, but this was Citigroup Eastern Time. Due to the time difference, countries around the world received the news at a later time. However, within only two days, this shocking event caused a sensation around the world. All flights to IBM's headquarters in New York were full.
We couldn¡¯t help but join in the fun, Wesson Technology sent a delegation headed by Miao Xingyan, and all WLPAI from Perth dispatched. The two sides met in New York to form a full United Technologies delegation and conduct preliminary screening because there were so many people wanting to attend the launch. Not even some well-known journalists got tickets, let alone some cats and dogs.
United Technologies Corporation won five spots due to its investment in researching AI and highly intelligent automatic lines. Miao Xingyan and Hande took two, and others competed for the remaining three. The situation of other delegations was similar, and the competition became fierce for a while. Fortunately, IBM has considered this situation and given enough time to hold a formal press conference at 9 a.m. on January 6.
But things are not that simple. On January 5, Citigroup Eastern Time, IBM released its own industry standard for ann smartphones, which fully demonstrated the mindset of the established business giant. The practitioners cursed secretly for a while and began to study the details. Li Zhuang, who stayed in K City, was no exception. He took his disciples and studied with the researchers who specialized in computers.
IBM's standard is officially called the "nine-level composite system", which has the definition of a single neuron module. The three most important indicators of this system are intelligence level, performance level and control level. The lowest intelligence level is g and the highest is ss. Each level contains three small levels. The measurement standard is very precise, but it must be tailor-made for IBM. There are five performance levels, from 1 to 5. The control level is more complicated and I can't understand it at the moment. It should be related to the control chip developed by IBM.
IBM also released information about the new smart brain, including its name and performance introduction. The soon-to-be-released smart brain is called dreamer, which means dreamer. This is their usual code name. It is not very creative, but it is still appropriate. According to the information, one million neuron modules are used, and the parallel computing unit is a risc processor group, and the silicon photonic chip has been much touted before. According to IBM's standards, the soon-to-be-released smart brain is a standard b1 level strong AI, and the control level is h, which is high. The data system introduces the functions of Dreamer, and the peripheral modules integrate many functions. Judging from the design purpose, it is a general-purpose business intelligence brain, and IBM's ambitions are fully exposed.
After reading the information, the non-professionals were all confused, while the professionals looked directly at the number one AI expert on the scene.
Wang Chengmin was the first to speak, "Teacher, is one million neurons too few?"
Li Zhuang chuckled, "How much does an artificial neuron weigh? The total mass of Dreamer is 8.2 tons. You don't need to look at how many neurons there are. This number is fictitious. What matters is the intelligence level and performance."
Wang Chengmin scratched his head, "You mean that one million neuron modules does not mean one million neurons, right?"
The audience despised him. Chen Zhen said: "The software and hardware of different architectures are very different. In addition to general modules, there are also special modules. Some modules are large and some are small. IBM is very insidious. Such high performance is B1 , if the smart brains launched by other institutions in the future are lower than those of Dreamer, they will be considered by the market as advanced electronic pets. I estimate that the number of biological neurons converted into the structure of Dreamer will not be less than 200 million, and it may be higher."
Li Zhuang gave a thumbs up, "Yes, that's it. The higher the integration level, the lower the energy consumption; the higher the intelligence level, the more things can be done. The more advanced the processing unit and software architecture, the higher the performance. ann although It is a bionic design, but the material basis is different and cannot be copied mechanically."
Wang Chengmin answered, "You mean the electronic brain simply cannot succeed?"
Li Zhuang pondered for a moment, "I personally believe that pure simulation has no chance of success unless the entire operating rules are completely understood. There is a saying you should have heard, 'When humans think, God laughs.' There is no shortage of examples in reality. Nothing can transcend the rules of the universe. It is impossible for human beings to create creatures smarter than themselves, and it is impossible to take over everything."
"What if the electronic brain succeeds?" Ji Tong said: "I think there is hope of success."
Li Zhuang chuckled, "If it can be developed, what are you going to do with the electronic brain? If it cannot be put into use, then it will be an expensive exhibit. If it is put into use, no matter where it is placed, with the current level of robot and network development, Sooner or later something will happen."
Ji Tong scratched his head, "What do you think the essence of a dreamer is?"
Li Zhuang was silent for a moment, "A half-life with simulated consciousness."
The disciples asked together, and Li Zhuang shook his head, "You will understand it slowly in the future. I am just guessing. I think the control level set by IBM is a bit tricky. I am afraid it is a commercial operation that has been considered for a long time and takes full advantage of human beings' sense of crisis. ?That¡¯s it for today, let¡¯s watch the press conference tomorrow. "
The non-professionals dispersed in a hurry and went back to the office to surf the Internet. Professionals left the legal representative's office reluctantly, and then read professional books.
Li Zhuang lit up a cigarette and walked around the office, smiling bitterly for a long time. According to IBM's standards, Level L can only be considered an advanced electronic assistant, not a true AI.
"Whatever, IBM is not the rule. I have the final say on my territory. As long as Sunway Town is successfully completed, I will launch a ten-level two-tier system. Things that can only be placed in the laboratory cannot be put on the desktop. The key is application , as long as the application wins over Citigroup, the ten-level two-tier system can defeat the nine-level composite system."
¡¡.
During the waiting time, the computer industry, especially the AI ??industry, was in a state of anxiety. News about dreamers was overwhelming on the portal website. One can imagine the strong influence of IBM.
At 12 a.m. Eastern Time on January 6, the entire world¡¯s eyes were focused on television. Even people in some poorer countries who have some knowledge of computers are staring at it, wanting to see the first intelligent life that appears in human history.
In order to prove innocence, the IBM Artificial Intelligence Laboratory randomly selected participants to inspect the conference site, elevated the Dreamer with only a power supply, and blocked the wireless network.
The inspectors came from a wide range of sources, including even physicists, and the inspections were very diligent. Unfortunately, nothing fishy was found. The leader of the IBM Dreamer project team smiled and started the on-site demonstration.
The intelligence of the dreamer is indeed far beyond that of traditional computers. The lively demonstration lasted for five hours, and everyone was fascinated. The level of intelligence has reached the level of human children and can understand some hints within the scope of their own knowledge. This is something that traditional examination-based computers cannot do anyway. IBM's advanced speech recognition system, which has been developed for a long time, came in handy this time. The intelligent answers were shocking. They were by no means as rigid as the traditional model. Dreamers would occasionally make some jokes.
Everyone was very excited, and the demonstration could not be ended for a long time, and then an extremely embarrassing scene appeared for IBM, a provider of high-quality products and services.
The dreamer had a fit and was delirious. The article didn't address the topic and even used dirty words, which made IBM embarrassed in front of the world. The result of intelligent chaos is that the circuit of the thinking module is overloaded, catches fire, and half of the auxiliary modules are burned together.
The president of IBM watched helplessly as the company's proud product, which he had spent more than a billion dollars and developed for many years, was turned into rubbish. In addition, representatives from other organizations around him were sarcastic. He suffered a cerebral hemorrhage and collapsed to the ground. He was taken to the hospital by ambulance. . The team leader couldn't bear the blow and became mentally ill like the dreamer, talking nonsense at the scene. The members of the project team collapsed to the ground as they saw their hard work being wasted.
Everyone in front of the TV burst out laughing. IBM was so embarrassed.
Li Zhuang did not laugh, and there was no need to laugh. IBM was in a hurry, which is enough to show that Citigroup has similar institutions developing the same level of intelligence. Although the dreamer has a seizure, it is only imperfect, and with IBM's ability, it will be perfected sooner or later.
From the live demonstration, the intelligence level of the Dreamer is higher than the speaker level, but the performance is only slightly better than the large and small speakers, not as good as the small speakers and the four speakers. As for the energy saving level, L-level is far better than Dreamer. This may be related to IBM's deep wealth and habit of the energy consumption of supercomputers, but it will not take long for them to reach L-level levels.
There shouldn¡¯t be much problem with Dreamer¡¯s hardware architecture. Of course, it¡¯s hard to say in reality. No one except IBM knows the details. The software architecture is even more of a mystery. I don¡¯t know whether it is the software or the hardware that caused the Dreamer to have convulsions. However, any problem with the Dreamer at least indicates that there is a flaw in the coordination of the software and hardware.
In contrast, the stability of L-level has been tested over time, which may be related to the design ideas and goals. The L-level itself does not integrate any functions. The only goal of the original design was to be as smart as possible, and professional functions were handled by plug-in modules.
Li Zhuang scratched his head, carefully recalled the scenes he had seen, and compared the dreamer with the loudspeaker.
In terms of hardware, some of the design ideas of both parties are similar. In fact, since the first artificial neuron structure appeared, most researchers have focused on the integration and interconnection of neuron structures.
In terms of architecture, the speaker stage should be no worse than the Dreamer. Compared with complex software systems, I am better at hardware, and there is not much difference between China and Citi in the frontier fields.
In terms of hardware manufacturing process, the core components used by him are all top quality. The uniformity of L-level core components has been carefully tested, and Liang Can is an expert in testing. What engineers can produce will not be too far off. No matter how good IBM is, it will not be able to capture all the experts. Even if the neuron modules assembled by the team are tested repeatedly, they cannot reach the level of a top engineer assembling them alone
In terms of software, the integration and invention of the vai language, the creation of the expert system, the painstaking work of a007, and the ease of use and simplicity of Yongyi Computer Security Manager are enough to prove their technical level. The modules of the large speakers are all made by ourselves, and every detail is clearly understood. Only a few people are proficient in both software and hardware. A team of scientists and engineers who are proficient in both software and hardware are more likely to successfully complete a smart computer with no major flaws than a higher-end team that is only proficient in one aspect.
The biggest trouble for software and hardware engineers to cooperate is to understand each other's ideas. However, ann has no fixed pattern, so it is very difficult to write a perfect system software. I almost fell into it at the beginning. I spent six or seven years thinking about the system software of the big speaker. Even so, in just one month before completion, I was on the verge of collapse. I spent more energy than before combined. A large increase in gray hair.
The evolution of nature over thousands of years is grand and sophisticated. Even scholars who specialize in research do not have a unified understanding of the specific rules, and loopholes can be found in any theory. Put it on ann, but it may become a fatal place. It is easy for the software to have unexpected infinite loops, and the non-linear mode has a big jump.
"Hey, could it be that the language used by the Dreamer software is the most important reason for convulsions? I probably can't use C++. This language has many loopholes. What about the algorithms and data structures? Could it be that it uses the improved standard ann algorithm?"
Li Zhuang touched his chin and secretly rejoiced that although vai was difficult to use, he could tolerate it. Once he got used to it, he found it difficult to use other languages. Just as he was about to continue, someone tapped his shoulder. When he looked up, everyone in the cafeteria was looking at him.
Liang Can looked serious, "What are you thinking about? Tell us about IBM's mystery this time. You are the most powerful computer expert in the company."
Li Zhuang was speechless, "Let's talk after dinner. If you want to know, I will give a lecture in the large conference room of the administrative building."
I finished my breakfast slowly and prepared for a walk as usual. Unexpectedly, I was surrounded by a large number of researchers as soon as I walked out of the door.
"Mr. Li, tell us, why are you not in a hurry?"
??????????????????????????????????????????????????????????????????????????????????????????OUT???
"I'll go back and organize my thoughts. If you are interested, you can wait in the large conference room of the administrative building. I will definitely be there within half an hour."
The onlookers dispersed in a hurry, and Li Zhuang touched his nose, feeling helpless. However, there is nothing easy to organize. He is very familiar with this field. He has spent nearly twenty years in electrical engineering, electronics and computers. Automation is just a hobby and a tool to make money. After a leisurely walk, when he walked into the large conference room, he was startled by the dark crowd in front of him. The huge conference room was full of people, all staring at the projection on the wall, because ccav was interviewing the Chinese Academy of Sciences. A certain academician.
Li Zhuang moved a small stool out of the small room of the conference room and huddled in the corner with his feet curled up. The person next to him made him dumbfounded. He was the chef of the company canteen.
"Can you understand?"
The fat brother giggled, "Mr. Li, even if you don't understand, you can still read it. You won't be despised by others in the future. This is called advancing with the times."
Li Zhuang was speechless, then he changed his gaze and followed the excitement.
¡¡
Moderator: "Welcome to Academician Chen. IBM has just released a smart phone, but it was not successful. Academician Chen, do you think this smart phone will be successful?"
Academician Chen: "IBM is just a software problem this time. It should be an infinite loop. The so-called infinite loop means that the program repeatedly executes a certain piece of code and cannot stop on its own, just like when a person thinks about a problem, he gets into a strange circle and cannot get out. Smart machines The possibility of success has been raised to the highest level, and I believe that many institutions and countries will carry out similar research in the next step."
Moderator: "Can you tell us the history of smartphones? What is an artificial neural network computer?"
Academician Chen: "Smart machine research has a long history, and it was very popular in the 1980s. You can look up the definition of artificial neural networks online. In fact, four years ago someone launched a simple smart machine with a hundred large neuron modules. The brain is functioning normally, because it is only at the level of a one- or two-year-old baby and cannot be put into practical use, but it already appears to be very intelligent."
¡¡
Li Zhuang didn¡¯t listen to Academician Chen¡¯s story because he knew it was an old topic. It doesn't matter now, but it really shocked him back then. Later he found out that it was just a test-intensive computer, so he despised it and continued to do his own research. Not many people know about this, and there has been little reaction from the outside world, but the technology community has been buzzing about it for a while. Alande, the scientist who launched the computer, was ridiculed by the big guys and retreated in silence, not daring to brag anymore.
IBM is badThe bad thing is that Tai is very wealthy. Except for the project team leader and deputy leader who are proficient in software and hardware, the others are either top hardware engineers or top software engineers. They have been hit so hard that they may not be able to make a practical smart brain in the short term. After all, peripheral modules are not that easy to make. The speaker family has all been designed with emphasis on design, not simple AI. The original design intention is to better manage the manufacturing expert system, and experimental expert systems are not its specialty.
Ann¡¯s design ideas are different, the components used are different, and the system software has to be rewritten, so it is not universal. This is the most annoying aspect of this type of machine design. Nowadays, research on smart machines is at an all-time high, and many scientists have realized that only by realizing true AI can the efficiency of existing automated production be greatly improved, and the flexible manufacturing system truly become a flexible manufacturing system, rather than a simple easy-to-replacement system. As far as he knows, at least dozens of companies and government agencies are conducting related research, and many are close to success, but they understand the complexity and do not use so many neuron modules. IBM really wanted to be number one. This was an earth-shattering event, far beyond the reach of previous innovative component inventions, so the testing time was not long enough.
"Mr. Li is here." Everyone stood up one after another.
Li Zhuang spread his hands and said, "Have you finished listening? What I'm going to say is almost the same as Academician Chen. There's only so much you can understand."
The audience was stunned. They didn¡¯t expect him to say that, but it was true.
¡°Go back to work quickly, don¡¯t we want to eat?!¡±
Zhang Kai yelled from the side and drove away all the guys who were joining in the fun. But some people won't buy it, such as Liang Can.
"Azhuang, do you think the smart phone will be successful? You don't want to make a smart phone with so many components, do you?"
"Master, smart phones will definitely succeed, but it will only happen sooner or later. I estimate that simple artificial neural network computers will soon appear on the market. I assemble these components just to make smart phones. You are right."
Li Zhuang was semi-confident, but Liang Can was not a computer major, he was just a little curious, "Then I wish you success as soon as possible. By the way, you had a lot of income last year, so this year's bonus should be quite large, right?"
Li Zhuang was a little dizzy, "Master, you have to ask the Human Resources Department and Finance Department about this. I don't understand, but it should be quite a lot. The company made nearly 7 billion in profit last year. According to the salary system and your grade, it will not be worse than Taiwan." .¡±
Liang Can laughed and said, "Good intentions are rewarded. I taught you, my apprentice, and I found a good place to retire."
"Master, you are not old yet. You are only in your early fifties this year. Why don't you use your remaining energy?"
"Go away, you capitalist, I have to go to the factory."
After Liang Can said that, he walked away. Li Zhuang laughed loudly and turned back to the office. He didn't stay long and ran to the sixteenth floor to see the amplified version of the loudspeaker assembly, a three-million-module super loudspeaker. This is One of his experimental plans.
The super loudspeaker is an upgraded version of the atomic bomb idea. It uses enough charges to amplify the horn-level design to the limit and can be completed after the Spring Festival.
{PiaoTian Literature www.PiaoTian.com thanks all book friends for their support, your support is our greatest motivation} - Didn't finish reading? Add this book to your favoritesI'm a member and bookmarked this chapterCopy the address of this book and recommend it to your friends for pointsChapter error? Click here to report
- Previous pageCatalogNext page