If there's one thing we've learned about technology over the past few years, it's that we don't anticipate the drawbacks until they get so severe that we can no longer deny them. The most prominent example is social media, which was once regarded as idealistic but is now mostly viewed as a harmful and disruptive influence on society that can magnify hate, propaganda, abuse, and divisiveness. It took more than ten years for the concerns to become truly apparent. However, nowadays, more than 68% of the world's population believes social media has a negative impact on society.
There is a new class of products and technologies requiring only affordable hardware that have the potential to endanger our cognitive freedom across enormous populations. These seemingly harmless technologies could be fatally mishandled and exploited without our knowledge or agreement. These products will be promoted for a variety of good purposes, from entertainment to education, and they are aimed at both adolescents and adults, such as the METAVERSE.
Naturally, these issues are not caused by technology itself but by social media platform's business strategies, which filter and magnify information in ways that can negatively influence our thinking and manage information across societies. This means it will get much worse since this is a sort of mind control.
It's possible that the metaverse may unleash an unstoppable flow of human invention and productivity. Nonetheless, it may become the most lethal instrument ever developed if unregulated. Humanity can benefit from the metaverse; however, if we wait until the issues are severe, like social media, it will be too late to reverse the harm. We should keep in mind that the Metaverse is far more dangerous than social media.
To demonstrate how dangerous the Metaverse is, we will introduce a fundamental engineering concept called Control Theory, which is the technique employed by experts to influence the behaviors of any system, to highlight the dangers as clearly as possible. Think of your home's thermostat. You specify a target temperature, and the heat comes on if the temperature drops below the specified target. If the temperature rises above the specified target, it turns off. It keeps your home near to the objective you set, almost magically. This is called control using feedback.
In the above example, the house represents the system, the thermometer represents the sensor, and the thermostat represents the controller—the temperature you specify as the target is an input signal referred to as the Reference. The target temperature is contrasted with your home's actual temperature. The thermostat uses the difference between the desired and measured temperatures to decide how to operate the heater. The heater turns on if the house becomes too cold and switches off if it becomes too hot.
Naturally, control systems can be incredibly complex, allowing driverless robotic vehicles to land on Mars and allowing autonomous cars to drive themselves and airplanes to fly on autopilot. These systems require specialized sensors to identify flying or driving conditions suited for the task and require robust controllers to interpret sensor input and subtly modify system behavior. Nowadays, controllers primarily incorporate more AI algorithms.
In the above diagram, you can see that no matter the complexity of the system being controlled, whether it is a basic thermostat or an advanced robot, only a few components are required. The most vital components are A SENSOR to track the system's real-time behaviors and a CONTROLLER to modify those behaviors. Other components are feedback loops that constantly detect behaviors and convey influences, directing the system toward the expected objectives.
As you would have surmised, when thinking about the risks of the metaverse, the system being controlled is you. After all, when you wear a headset and enter the metaverse, you could be submerging yourself in a setting that affects you more than it does the other way around. In other words, you enter a created environment governed by a third party that can observe and affect your actions in real-time. This is an extremely risky circumstance.
The immersive images, audio, and touch sensations that are delivered into your senses and body are referred to as system input to the human participant in the diagram above. Without requiring brain transplants, this is the most emotional and comprehensive input we could possibly envision. This indicates that the system's capacity for influence is both comprehensive and personalized. The System Output, which represents your responses and actions, is on the opposite side of the user in the diagram.
That leads us to the SENSOR box in the previous diagram. Sensors in the metaverse will continuously monitor everything you do, including your head, hands, and body movements. This includes the way you're gazing, how long your focus stays, the slight movement of your eyes, the dilating of your pupils, the changes in your position and body movements, and your vital indicators, such as your heartbeat, breathing rate, and blood pressure.
This means that sensors will virtually monitor everything you do and will be able to identify your actions and emotions while interacting with the metaverse. This is shown in the Diagram (replacing the SENSOR box with the METAVERSE)
The behavioral and emotional input will not only be tracked in an uncontrolled metaverse; it will also be kept over time, building a repository of data that depicts how people are likely to respond to a variety of stimuli. This vast amount of data might be converted into behavioral and emotional model types by AI algorithms, allowing systems to precisely anticipate how users will respond when exposed to specific stimulus from a controller.
Additionally, because the metaverse combines virtual and augmented reality, user monitoring and profiling will take place not just in completely created digital environment but also in the real world that has been enhanced with virtual elements. In other words, from the time we get up till the time we go to sleep, metaverse will be able to monitor and analyze our daily activities and feelings.
Undoubtedly, the risk is in what the metaverse can do with the data they collect, not in the fact that they can monitor and analyze us. We now arrive at the CONTROLLER box in the previous diagram. The controller collects a Measured Error, distinguishing between a Reference Goal (an expected behavior) and the Measured Output (a perceived behavior). If the metaverse is permitted to embrace economic models similar to social media, The Reference Goal will be the AGENDA of third parties seeking to exert influence over the users. The third-party may be a remunerated sponsor who wants to influence a user to purchase a good or service or to accept disinformation, agendas, or political ideologies.
Advertising and propaganda have always existed and may be very successful with conventional business strategies. The ability to build rapid feedback loops in which user actions and feelings are constantly sent into a controller that can acclimate its effect to maximize influence makes the metaverse special. This method can quickly go from marketing to manipulation.
The CONTROLLER's primary goal is to "lower the error" of both the system's expected behavior and the measured behavior. It accomplishes this by relaying System Input, as shown in the diagram above, as an innocent-looking arrow. This arrow describes the platforms' capacity to change the immersive virtual or augmented world in which users engage in the metaverse.
In other words, in an uncontrolled metaverse, the controller can modify the user's environment by changing what they hear, see, and experience to influence them toward the expected goal. The controller will also be able to continuously modify its strategies, maximizing the persuasive effects, similar to what a thermostat does to regulate a home's temperature.
To clarify, let's take a look at this example:
Think of a user in a coffee shop in the metaverse (virtual or augmented). The goal of third-party is to persuade the user to purchase a specific item or a service or to accept some propaganda or false information. Instead of the pop-up advertisement, marketing in the metaverse will take the form of interactive experiences that are effortlessly incorporated into our environment. In this instance, the controller creates a fictional couple and places them at an adjacent table. The input utilized to influence the user will be that virtual couple.
The controller will design the virtual couple for maximum effect. Based on the user's data, AI algorithms will choose the couple's age, race, sex, religion, clothing, language, and other characteristics to be the most persuasive. The couple will start talking to each other under AI control while the target user listens. The fictional couple may be discussing an automobile in a conversation about how pleased they are with the product.
Moreover, the user may think that the virtual couple are avatars controlled by other users. This could lead the target user to believe they are listening to an authentic conversation between real people and not realize that it is a personalized advertisement designed specifically for them to achieve a particular goal.
It is important to mention that adults are not the only ones who the metaverse will influence. Children who already have difficulty telling the difference between what is real and promotional content will also be exploited in this way.
It is well-established that artificial intelligence can beat the best players in the world in strategic games like chess. From this viewpoint, what chance does the typical user have when immersed in promotional conversation with an AI model that can access that user's data and preferences and adapt its conversational strategies based on quick physical and emotional changes? This kind of feedback system that poses a significant risk to users' cognitive freedom.
The below diagram shows the complete metaverse feedback control system. Notice that the "Controller" is replaced by "AI Agent" which influences the target user by altering the environment or inserting conversational avatars that convey personalized and immersive experiences to the target user.
As stated above, the general public should be informed that big metaverse platforms can be exploited to build feedback-control systems that track users' actions and emotions instantly and utilize AI agents to alter their immersive experiences for maximum impact by modifying their environment in a controlled and flexible manner. Powerful and massive platforms could track millions of people and influence them. After all, the idea of mind control in the metaverse is terrifying and plausible. It might even be the closest thing to “playing god” the social media platforms have ever accomplished. Before platforms embrace business models that are harmful to the users, industry representatives, legislators, and regulators need to take steps to prevent this situation by implementing safeguards, establishing industry standards, and ensuring Immersive Rights to users.
Sources:
Louis Rosenberg, Dec 28, 2022. Retrieved from: https://medium.com/predict/mind-control-in-the-metaverse-48dfbd88c2ae
No comments.