Artificial intelligence and big data.
With the development of AI technology, the impact of AI systems on our daily lives is ubiquitous. Scientists develop AI technologies and systems not only to solve functional or technical problems, but also to better help human beings. However, any technology has two sides, and AI technology is no exception. How can we achieve goals that benefit humanity instead of abusing or even endangering them?As a developer of AI technology, only if ethical standards are followed, can the better development and use of AI technology be guaranteed. Overall, in order to design, develop, and apply these technologies ethically, the following general principles should be followed:
Human rights: Ensure that they do not violate internationally recognized human rights.
Well-being: Indicators that prioritize human well-being in their design and use.
Accountability: Ensure that their designers and operators are accountable and accountable.
Transparency: Make sure they operate in a transparent manner.
Use with caution: Minimize the risk of abuse.
Data Privacy and Security.
Artificial intelligence technology is often based on big data collection and algorithm analysis, which not only appears in the technology development stage, but also continues in the whole process of research and development and application. Once it involves the continuous collection and analysis of large amounts of data, data privacy and security have become an inevitable problem for artificial intelligence technology.
The user's right to informed consent is very important. When users use AI technology or products, they need to understand what data the technology or product collects about them, where the data will be used, and freely choose whether to consent to data collection. Similarly, many AI technologies or products involve data sharing and access. For this function, the user must also have the right to informed consent. Of course, this data includes not only the raw data collected by the user, but also the data of a conclusive nature after analysis. After all, this data may include a series of privacy issues such as the user's health, religious beliefs, personal Xi and preferences, and personal attitudes. Once these private data are used illegally, the economic and personal losses caused are huge.
The security of data storage is also an important topic. With the development of cloud computing technology, many companies, scientific research institutions and other organizations will store data in the cloud, and even carry out data analysis and algorithm updates in the cloud. This information is at risk of being attacked and compromised. How to use Internet security technology to ensure data security in the cloud as much as possible is also a key point that needs to be considered in the field of artificial intelligence technology.
Ethics of AI Application.
Artificial intelligence can be applied in many fields, such as: user portraits, algorithm decision-making, autonomous driving, robots, etc., and many ethical and moral issues will arise in the process of artificial intelligence technology application.
2.1 Ethics of AI Decision-makingAI decision-making is often based on algorithms, and in general, we think that algorithms themselves are not biased. However, before discussing this issue, we need to understand how AI algorithms operate. AI algorithmic models for decision-making are not created in a vacuum. Just as children need to learn a lot Xi of knowledge and experience when they establish the three views, the algorithm used for decision-making needs to be trained by learning a large amount Xi of relevant data before it really works, until a good model is formed, and then it will be used for decision-making. If a child learns too many things that violate ethics or even violate the law (such as lying, stealing, etc.) from an early Xi age, then the three views he establishes are contrary to ethics and even the law. Similarly, if the data we use to train the algorithm is itself biased and unethical, then the algorithms we train for decision-making will inevitably make biased and unethical decisions. For example, in recent years, the MIT Media Lab has created the world's first psychic artificial intelligence, Norman.
on the Internet.
In addition to the bias of AI decision-making itself, AI-based decision-making can also have ethical issues in other contexts. For example, in the face of a moral dilemma, whether AI decisions should be accepted or not. The most typical moral dilemma is that when a train goes in one direction, and there are 5 people standing on the train track in that direction, and the train can't brake for various reasons, and the masses can't avoid it, the only way is to reroute the train to another direction at the fork in the road, and there is only one person in the other direction who can't be avoided, will you let the train be rerouted?If the AI algorithm makes a decision on this issue, will it be accepted?Or does AI have the right to make these decisions?There may not be a clear answer to these questions, but what we need to be clear is that AI should be transparent in making decisions and use the results of decisions with caution.
2.2 Problems caused by the misuse of artificial intelligenceWith the development of artificial intelligence technology, artificial intelligence technology will be used in more and more important fields, such as: education, driving, medical care, and even the military. As we mentioned earlier, AI computing in the cloud often has the risk of being attacked and controlled. Assuming that hackers attack the cloud and successfully control all the driverless vehicles on the road, it is not impossible for the scenes in Fast and Furious 8 to happen.
on the Internet.
If the application of artificial intelligence technology in the field of transportation is only a problem of controlled risks, then the application prospect in the military and advanced fields is frightening. For example, the combination of artificial intelligence technology and drone technology can deliver bombs with great precision. Now, there are already a lot of manufacturers of civilian drones who find that their products are used to carry out terrorist attacks and fight against the military. Once artificial intelligence technology is added, the consequences are even more unimaginable. Recently, Google has been in such a predicament. A large number of employees oppose Google's participation in military AI projects in the form of ** and resignation. It can be seen that researchers have a certain consensus on the bottom line of the application of artificial intelligence technology.
In addition, because artificial intelligence algorithms can not only be used to judge the existing situation, but also to predict, in terms of ***, it is even possible to use this kind of technology combined with personal behavior, physiology, psychology, etc. Big data to predict individual criminal behavior, just like described in the movie "Minority Report". But is this application really ethical?Is it abuse?We need to be very cautious about this.
2.3 Ethics and Ethics of Super Artificial Intelligence BodiesWhen artificial intelligence technology develops to a certain extent, they may have the same learning Xi, thinking ability, and even emotions as humans, so how should we treat such super artificial intelligence at this time?As early as Asimov's science fiction, the three laws of robots were mentioned (zero, robots must not harm human beings as a whole, nor can they sit idly by and watch human beings as a whole be harmed. 1. Robots must not harm people, nor do they stand idly by while people are harmed, but they must not violate Law Zero. 2. Robots should obey all human orders, but must not violate the zero and first laws. 3. Robots should protect their own safety, but they must not violate clauses 0 and 1.
1. The Second Law. ）。Of course, whether these laws are binding or not we will not discuss here.
on the Internet.
In addition to the mandatory definition of the nature of the law, are super-AI subject to the same ethical and moral standards as humans?Or even more stringent legal constraintsAnd after being bound by ethics and law, can super artificial intelligence enjoy the same rights as humans?These are questions that need to be carefully considered.