Subconscious Artificial Intelligence system regulation: a global challenge?
The era we are currently in is an era full of paradoxes, just like the contradictory era in the past, such eras have brought us severe regulatory challenges. Artificial intelligence is described as a paradoxical existence because human intelligence does not get along well with AI machines with learning capabilities. So, in my opinion, the "Paraems of Our Time" proposed by David Held in his 2010 book summarizes this well, which reads: "The Path of Our Time" The paradox can be simply summarized as: the common problems we must solve are increasingly globalized, yet the means to solve them are limited to one country and one place, weak and incomplete. "This paradox applies to sustainable development by Many contemporary issues summarized by Objective , such as climate change and inequality in public health facilities during the global pandemic, but I think artificial intelligence also falls into this category, and I will try to explain the reasons below.
has carried out considerable regulatory activities in the past two years. These regulatory activities began last year, April 2021, when European Commission proposed a new regulation that stipulates unified rules for artificial intelligence, the so-called Artificial Intelligence Act, which is currently under formulation. Shortly thereafter, in November, UNESCO also adopted the Recommendation on the Ethical Issues of Artificial Intelligence in November with the broad consensus of its 193 members. While this is only a suggestion, it is also a very important document that shows a global concern: AI can not only bring opportunities, but also risks associated with it. This is also an initiative proposed by the European Commission, and of course, many other countries are also discussing legislative and regulatory behaviors that deal with the various consequences of artificial intelligence at the domestic level. However, the bill and suggestions of EU are also quite comprehensive, and are still a method of horizontal comparison, which is also the interesting point of taking it as the starting point.
Before this, I would also like to mention a specific aspect that is also emphasized in the UNESCO text, namely the huge positive and negative impact of artificial intelligence on society, environment, ecosystems and human life, with particular emphasis on human life Influence, because this is the reason why artificial intelligence has become a game changer to some extent, because artificial intelligence allows us to influence human thinking, interactive communication, decision-making, and ultimately change the way we behave.
In this regard, my book also discusses a specific aspect of the European Union Artificial Intelligence Act, namely Article 5 (1), which suggests that certain AI systems are prohibited from deploying subconscious technology outside of human consciousness , to substantially distort individual behavior, this technology may or may cause physical and psychological harm to the user or others.
So while this is only a proposed text, it seems to me that this is the core or key provision of the bill, involving the impact of AI on the brain, and its possibility of manipulating the brain in positive and negative ways. In this regard, through this clause, many important questions are raised, namely, first of all, what is subconscious perception? What is subconscious manipulation and is it effective? Secondly, above and below our consciousness, where is the threshold of perception? Does this threshold apply equally to every sense or all senses? Finally, what are these artificial intelligence systems and technologies that have been able to manipulate ours? In what form will they appear before us in the future?
is here, I would like to first review one thing I encountered in my research, that is, at this threshold, we can obviously only control 5% of cognitive functions, and 95% of them are the subconscious at work, which means Most of our decisions are made under subconscious control, and we basically make these behavioral decisions spontaneously, such as walking and driving. But this means that if there is technology that allows us to leverage this 95% of the subconscious, the technology will be a very powerful tool that can not only affect our thoughts but also our behavior.
This can also be traced back to experiments in the 1930s, first of which were some psychological experiments, and then some experimental findings: subconscious and sensory stimuli have an impact on the brain, and how we apply it to advertising, for example, After a quick broadcast of some shots about drinking Coca-Cola and eating popcorn in the cinema will make the audience consume more, although the audience themselves do not realize that these shots will have such effects. This has been controversial for years, and books in the 1950s warned that these subconscious strategies had a great impact on consumer behavior, and that these strategies could also be used for other purposes, such as political purposes, such as promoting politicians. Or change the election results. We know that the "Cambridge Analytica scandal" proves this possibility in a sense. But I don't want to discuss this further, I just want to summarize the subconscious problem, just like advertising, and I think there is a scientific consensus on this: many studies have proven that subconscious manipulation is effective and that it can affect our thoughts and behaviors. But to what extent it can play depends on many factors, but in short, it is generally believed that subconscious manipulation is possible.
Another question is: Where is the threshold limit for realizing subconscious manipulation? There has been a long debate about this. I think the latest findings scientifically summarize this problem: there is no such threshold limit.
First of all, this may be due to different people having different consciousness thresholds, some people having better hearing, some people having better vision, and some people focusing more on different senses. Most importantly, even in the same person, the awareness threshold is constantly changing; when you are awake, your vision is better, your hearing is better, and people are more alert, so this awareness threshold will become higher according to different factors. Or gets lower. Therefore, the above discussion is only from a regulatory perspective, and if there is more time, a comprehensive discussion should be conducted.
What are these artificial intelligence technologies that operate in the subconscious? Here I just want to give an example, there is an experiment about hotel website booking. When the experimenter accidentally adds tiny emojis to the video of the hotel room, that is, when the tester sees the video of the hotel room, it will not Fully aware of the existence of the smile emoji. In videos that accidentally embed smile emojis, consumers have high recognition rates, so this shows to some extent that simple emojis , simple stimulation, such as smiling emojis, can be used to consumers Behaviors have subconscious effects. There are many examples. In reality, some people have also done some experiments that invade the brain and extract personal information. They mainly use different combinations of technologies to extract personal information, including your password, your birthday, your address, etc. . The artificial intelligence mentioned here can coordinate learning, allowing machines to collect data, and then combine the data with different sources, such as computers, brains, interfaces or various medical devices, such as functional magnetic resonance imaging, etc. Therefore, the powerful means we see today are based on the fusion of these different technologies, which have been recognized by neuromarketing to some extent, and of course this means can be used for many other purposes as well.
Therefore, to some extent, people have realized that the general trend of artificial intelligence has come, and we have changed from programming computers to programming humans. This also summarizes the fact that we already have very powerful technologies, and as more devices are created, innovated and recombined, AI technologies will become more efficient in the future, such as augmented reality and virtual reality middle. In addition, we are also creating and copying the human brain, and through these means, we can further understand what is happening in the human brain. In this regard, we need a heated discussion, a discussion not only limited to a local perspective, but a global perspective, because to some extent, there is a danger in artificial intelligence: that is, through these artificial intelligence technologies, in human personality The personal thoughts of the most private parts will become open and transparent, and these technologies, both in positive and negative ways, will become powerful tools for manipulating human behavior.
In addition, artificial intelligence itself is not contradictory. I think artificial intelligence is not "intelligence", and we should not refer to it as "intelligence", because such a title is an exaggeration in terms of its actual situation. Interestingly, computers can also be manipulated, computers are not completely trustworthy, and computers can also make mistakes. For example, our laptop will also crash or other failures. The same situation may also happen in autonomous vehicles, and once this happens, the consequences will be unimaginable. Therefore, we should truly pay attention to the current and future intelligence and trustworthiness of artificial intelligence. Therefore, we need a discussion, a global discussion, and the focus of the discussion is to focus on the trustworthiness of AI and how we can make AI compliant with ethics and ethics . We also need to discuss how much human scrutiny AI needs to be, and we should not give everything to machines. We must also ensure that these systems are interoperable in different legal systems, different jurisdictions, etc., and there are no human barriers we know from the Internet, such as geographical blocking. Therefore, we must also protect our brains and minds through basic rights such as freedom of thought and privacy protection, and we must also rethink the meaning of “causing harm”. There are very interesting comments on the "scope of harm". Harm should not only refer to psychological and physiological harm, but we also need to define harm more broadly: social harm, cultural harm, financial harm, or a manipulated election will also It has an impact on all people. Therefore, we need to listen to all parties to expand our understanding of harm.
Finally, we need a larger scope of regulatory consistency, which means consistency between different departmental laws , which cannot be regulated by only one departmental laws, such as only unfair competition laws or only intellectual property rights Law regulates it. This does require us to rethink the entire legal system and how these different departmental laws work together. Most importantly, we need an interdisciplinary discussion because we need the input of psychology, neuroscience , and all the people involved in the development of these AI technologies and AI systems that play a role in the subconscious, and together Work to ensure that artificial intelligence technologies and systems are safe in the future and to a certain extent can protect the privacy of our ideas. Ultimately, this is because if we can’t control our own thoughts (although often difficult), then it’s very likely that someone or algorithm will become the controllers of your thoughts.
Luo Zhen Jin Huizhu
