(Original title: AI is used in the accounting industry, saving millions of dollars annually for companies; Chinese courts are introducing AI into the case-handling system | AI Nugget Evening News) AI is used in the accounting industry and saves companies millions of dollars annually In addition to the legal and medical fields, how will AI help accountants? From a personal point of view, accountants have many tasks in their daily work. There are many practical needs that cannot be solved. From an industry perspective, the number of employees in the accounting profession is currently saturated, but there is a shortage of high-level professionals. For many companies, It is difficult to obtain professional business consulting from the accountant side, and this directly affects the relevant companies to obtain the correct financial reports in compliance with laws and regulations. So how will AI help solve this problem. At present, AI-related companies help solve these problems through extensive training. But it is worth noting that companies still need accountants to "check" the results of the implementation, so it will only serve humans better, not replace the work of accountants. If related technologies can be developed and approved by users and used, it will save users a lot of time and money. So, how does an artificial intelligence system interact with accountants and companies? - Chat bots. Chat bots are becoming more and more popular. Users can use it to perform natural language queries or talk to them through voice recognition components. The query processor is connected to artificial intelligence. Then artificial intelligence is connected to the database of the enterprise and the Internet to bring users Rich professional information. These often have multiple components for speech input, natural language processing, knowledge of various business areas, and a variety of ways to present results. During use, some chatbots ended in failure. For example, Microsoft’s Twitter chatbot quickly became a “racistâ€. However, according to Lei Feng network (public number: Lei Feng network) understanding, has begun to see some successful cases, such as Sage's Pegg or the Netherlands Airlines Messenger robot. In addition, general-purpose robots such as Alexa, Siri, and Allo have become quite smart, and there are some great toolkits, such as Amazon Lex, which can be used to develop chat bots, so for more and more developers This will become very easy. Chinese courts are introducing AI into the case system Recently, at the International Summit of “Legal + Science and Technology†leaders held in Hangzhou, He Fan, director of the Planning Office of the Supreme People’s Court’s Office of Justice Reform, said that Chinese courts have been working hard to bring artificial intelligence into the case-handling system. He Fan said: "Artificial intelligence will have many changes to the model of the court's future. For instance, it changes the judge's judgment and the mode of decision-making. In the past, the judge judged according to his own courtroom and his understanding of the law. Judgments can be made, but it may be necessary to use artificial intelligence software to perform the deduction. In addition, artificial intelligence has changed some of our litigation process. With the advancement of artificial intelligence, it may be necessary to formulate an Internet litigation law in the future." At present, an ideal presentation mode of "AI Investigating Cases" is the Shanghai Intelligent Criminal Case Aiding Case System, abbreviated as 206 Project. The system will open up all data processing systems of the public security bureau, the procuratorate, and the courts. Based on the absorption of past mass evidence rules and legal rules, artificial intelligence will remind the investigators to reconstruct the case handling process of criminal cases. Since the launch of the 25 pilot units on May 3, the "206" trial operation has been completed for two months. After the “206†learned from the tens of thousands of criminal case files and document data in Shanghai, this “little baby†with big data, cloud computing, and artificial intelligence as its technical kernel already had preliminary evidence information and a school’s grasp. Test and logic analysis capabilities. Microsoft launches a cloud-based tool to detect software security risks with AI In the past, most companies would employ network security experts to solve these problems. However, in the face of ever-increasing and complex software programs, many experts are also unable to handle them. Recently, Microsoft has released an "AI security risk detection" tool that can solve these problems instantly and efficiently and protect the system from attacks. The tool leverages the capabilities of the cloud to extend the current test protocol internally, providing developers with solutions that they may miss. According to Microsoft, if there are any problems with the software, companies can use this tool to proactively handle and prepare for emergencies. The team led by Microsoft researcher David Molnar developed this risk detection tool. Molnar said that the risk detection service can be used as an aid to help developers find security issues with artificial intelligence. When a vulnerability is found, developers can use other tools to fix the vulnerability, reduce risk, or explore other solutions. Lei Feng network learned that the special feature of Microsoft's security risk detection service is that it can identify the factors that may trigger a crash and trigger security risks. Each time it is run, it focuses on the most critical areas to find loopholes that other tools that do not use smart methods may overlook. Molnar said that this tool is ideal for companies that independently develop software, modify off-the-shelf software, or use open-source product licenses. DocuSign is a company that helps users electronically and no longer sign documents in pen and paper. They participated in a small-scale trial of the Windows version of the risk detection service released in the fall of 2016. John Heasman, senior director of software security at DocuSign, said the tool helped them identify potential errors that might otherwise not be discovered. Anodot uses artificial intelligence to overcome people's limitations in data analysis According to foreign media reports, Anodot, Israel's big data startup company for real-time analysis and anomaly detection, is currently using advanced machine learning algorithms to overcome human limitations in data analysis. Artificial intelligence can provide more answers by analyzing all data, and Anodot believes that e-commerce companies that do not integrate machine learning into data analysis will suffer losses. Rebecca Herson, vice president of marketing at Anodot, explained the need for artificial intelligence in this area: "Because so much data is generated, humans cannot view all the data. Sometimes when we analyze historical data for a business, they can't distinguish the data. And don't know where to find the data. Obviously, companies know that once the server crashes, they find it difficult to find all the data." This is not just about falling sales, but also data on product supply disruptions and customer satisfaction. In many case studies, Anodot found that 80% of the outliers found in his machine learning software were negative, not positive. These companies lost money because they did not realize the specific problem. Anodot is using unsupervised artificial intelligence, which can autonomously find new ways to classify and understand data through deep learning.