In the era of industrial digitalization, what kind of ChatGPT do we need?
Recently, the popularity of ChatGPT has remained high, and its powerful human-computer interaction capabilities are staggering, and it has set off a wave of discussions in China. For a time, this technological hurricane caused by ChatGPT made the world's top technological power spotlight on artificial intelligence (AI), and the game contest between Chinese and foreign technology giants also kicked off.
Google's Bard and Baidu's ERNIE Bot launched similar products as competitors of ChatGPT. Whether it is Google, Meta, Apple and other Silicon Valley technology giants, or domestic Internet giants such as Baidu, Alibaba, HKUST Xunfei, Netease, and JD.com, they have repeatedly mentioned "artificial intelligence", "generative AI" or "artificial intelligence" in recent days. It means that relevant products and technologies have already been accumulated.
The hot ChatGPT has amazed many technology parties who have tried it, and at the same time it has once again aroused people's infinite imagination and enthusiasm for AI. Especially in recent years, the integration of artificial intelligence and thousands of industries has accelerated, and its value has been released in many scenarios. This also makes people wonder what kind of sparks will be produced if ChatGPT is applied to the industrial field of the leading industry of the national economy?
Online reply from ChatGPT
With a monthly activity exceeding 100 million, the explosive ChatGPT has made many people in the industrial circle wonder what answers ChatGPT will give to their industrial applications.
Q: What are the application scenarios of ChatGPT in industry?
ChatGPT : There are many applicable scenarios.
For example:
1. Cross-industry science popularization : ChatGPT can answer various questions related to technology and industry, and quickly simplify complex technical knowledge into an easy-to-understand form, thereby helping people outside the professional field to understand the field;
2. Engineering technical support : ChatGPT can help engineers find code samples and technical solutions faster to speed up the development process. For example, generate code according to the requirements provided by engineers, or provide code modification suggestions, and prompt best practices and common mistakes during the writing process;
3. Knowledge update : ChatGPT can help employees of industrial companies quickly learn and update technical knowledge, such as providing the latest technical information and research progress according to the needs provided by engineers;
4. Data analysis : ChatGPT can use large amounts of data for analysis to help industrial companies better identify patterns and trends and make decisions based on them.
In addition, ChatGPT has many other application scenarios, such as helping industrial companies complete language translation, generate files, speech synthesis, online customer service, and provide prior knowledge/experience references, etc.
For these answers given by ChatGPT, some people in the industry verified the authenticity of ChatGPT’s answers in the form of questions and answers. The result is that ChatGPT’s coverage and knowledge are indeed much wider than we imagined. It is indeed very meaningful , but its weakness is also obvious, that is, the depth is not enough , the specific application cases cannot be expanded in detail, and very intuitive guidance cannot be given.
A programmer who has experienced ChatGPT told the reporter that ChatGPT can indeed provide code samples and technical solutions for the needs . The fun of finding bugs in answers.
In addition, regarding whether ChatGPT, which has been hotly discussed recently, will rob some industries of laborers’ jobs, some netizens also asked ChatGPT about whether automation technology and artificial intelligence technology will completely replace labor in the future? ChatGPT also gave an answer :Won't.
Since advanced artificial intelligence technology cannot completely replace human labor, what challenges do industrial engineers face in skill transformation in the era of digital intelligence?
The Industrial Internet is an important tool to help the high-quality development of the manufacturing industry, and it is also the key engine of an important strategy to meet the challenges of industrial development and promote transformation and upgrading. When can we realize the Industrial Internet and Industry 4.0?
The answer given by ChatGPT is indeed in line with the current development status of the Industrial Internet.
Through the above test questions and answers, although ChatGPT gives high-quality answers, provides an efficient way to obtain information, has strong language organization capabilities, and brings users an interactive experience beyond expectations, it cannot completely replace humans, especially in complex industrial models. And under the requirements of high-precision data analysis, the limitations of ChatGPT are clearly visible.
The To B field may be the first to achieve a breakthrough application
**"In China, similar to ChatGPT artificial intelligence in the commercial field, the To B field may obtain breakthrough applications before the To C field." **Professor of the School of Journalism and Communication, Beijing Normal University, director of the China Media Economics and Management Professional Committee Chang Yu Guoming said in an interview with the media.
From the business trends of technology giants, we can also get a glimpse of Microsoft’s additional investment of tens of billions of dollars in OpenAI, extending ChatGPT to its most difficult 2B business. He Xiaodong, vice president of JD Group and IEEE Fellow, also said recently that AI will definitely take the industrial route in the future. JD.com has been advancing on the ChatAI and AIGC tracks, but is more focused on the 2B industry. Further breakthroughs will be made in the near future, including the launch of the next generation on this track. product.
In fact, for the 2B end, artificial intelligence is an important starting point for the development of the industrial Internet. The two have a natural combination. The most innovative applications in the To B field may be born in the industrial Internet field.
Regarding what artificial intelligence technology can do in the industrial field, ChatGPT gives seven application scenarios.
Q: What can AI technology provide in the industrial field?
ChatGPT : AI technology can provide many benefits in the industrial field.
include:
1. Predictive maintenance . AI can be used to predict when equipment is likely to fail, enabling proactive maintenance and reducing downtime.
2. Quality control . AI can be used to inspect and test products for defects, improve product quality and reduce costs.
3. Process optimization . AI can be used to optimize industrial processes, such as scheduling and resource allocation, to increase efficiency and reduce costs.
4. Robotics . AI can be used to control and automate industrial robots, which can increase efficiency and reduce the need for human labor.
5. Energy efficiency . AI can be used to optimize energy usage, for example, to manage and control the use of resources to save energy.
6. Self-driving vehicles . AI can be used to control and navigate autonomous vehicles, such as drones and self-driving cars, which can be used for tasks such as inspection, transportation and mapping.
7. Predictive analysis . AI can be used to analyze data in industrial processes and predict future behavior for better decision making.
We know that many scenarios in the industrial field have complex industrial mechanisms and high data analysis capabilities, and artificial intelligence technologies represented by deep learning and knowledge graphs can fundamentally improve system modeling and deal with complexity, uncertainty, The ability to solve common-sense problems has significantly improved the analysis ability and efficiency of industrial big data, and further expanded the depth and breadth of the industrial Internet that can solve industrial problems. Artificial intelligence has also become the starting point for redefining the logic of industrial Internet products.
For example, after learning and refining more than 100,000 sample data and related quality knowledge in a hot rolling production line of a steel company, Baidu Smart Cloud Kaiwu launched a software-hardware integrated hot rolling system from model services to terminal platforms. The steel surface defect quality detection program has a defect recognition accuracy rate of over 99%. Under the same detection environment, the accuracy rate is 300% higher than that of similar foreign systems.
With the gradual deepening of industrial transformation, the application mode of "AI+Industrial Internet" presents two characteristics: first, based on a large amount of data collection, with machine learning or deep learning algorithm as the core, through the establishment of artificial intelligence models to solve specific diagnosis, prediction, etc. Second, it is user-oriented, realizes the coverage of the entire industry chain, and provides decision-making support for enterprises in terms of production control optimization, supply chain optimization, logistics scheduling optimization, and market sales forecasting.
At present, artificial intelligence technology represented by deep learning, knowledge graph, and natural language processing is in a period of multi-party innovation and breakthrough. Through the continuous deepening of integration with industrial knowledge, AI technology is gradually accelerating its penetration into the industrial Internet. In industrial enterprises Many landing applications have been formed under the business chain of "research, production, supply and marketing management".
Industrial ChatGPT Conjecture
"At least in the 30 years I have been working in technology, this is a technology diffusion (artificial intelligence) that I have never seen, and I don't think this kind of diffusion has happened during the industrial revolution." Microsoft CEO Nadella said, maybe this Once (ChatGPT), for knowledge workers, this is exactly equivalent to the industrial revolution.
Although mentioned above, there are still many limitations in the application of ChatGPT in the industrial field, but in the future, in the industrial field, can a chat dialog box solve industrial problems with the fastest speed and the highest efficiency? Will the artificial intelligence technology behind it completely change the way people obtain information and trigger an efficiency revolution?
"Using industrial ChatGPT, equipment maintenance does not need to study maintenance manuals, quality assessment does not need to read a large number of reports, and factory novice does not need to ask the master's production experience." An industry expert said that if we use ChatGPT technology in complex industrial scenarios , Can create a real industrial AI assistant, let AI answer these questions. As long as workers submit their requirements to the AI center, a dialog box can solve all problems . For example, the on-site maintenance workers ask the AI system whether the equipment needs maintenance; the factory management personnel help me write a report on improving lean production, etc.
He said that if feasible, industrial ChatGPT may change the existing working mode, especially in manufacturing enterprises with relatively mature informatization and digitalization construction . Workers no longer need to switch between various systems every day to view different data information various operations.
He said that if industrial ChatGPT is to be truly applied to industrial scenarios, in-depth model transformation and continuous optimization are needed to provide more professional industrial knowledge, so that the answers generated by AI are more relevant and accurate to questions, and provide effective answers to questioners. Help, so that AI can give full play to its best advantages in the professional field. But he also said that if the ChatGPT model in the application and industrial field is placed in industrial big data for enough training data and model configuration , it may really help industrial scenarios to mine data value and realize the conjecture that a chat dialog box can solve industrial problems , and has promising applications in the industrial field.
The idea is good, but it must be combined with reality. "Expert but not proficient is the overall feeling that ChatGPT gives me at this stage." The reporter interviewed a worker in the industry who also said that the industrial barriers are high, the scene requirements are complicated, and it is difficult to mine the value of massive industrial data. Some of the codes provided by ChatGPT are inaccurate, and it is still too early to be used in industrial scenarios.
The industry believes that if ChatGPT is to be connected to the industry, the configuration of the industrial model is very important, because it will affect the accuracy and efficiency of the model. How to train and configure industrial models is more reasonable, and how to obtain more diverse real data to fully nourish the model. I believe that ChatGPT still has a long way to go in the industry.