Symbols of the future, catalysts of promises for some, of fears and misunderstandings for others, artificial intelligence (AI) and the metaverse are two buzzwords. At first glance, AI and the metaverse seem to have in common only the complexity of their technology and their use cases. However, the metaverse is highly dependent on AI!
What is the metaverse?
The metaverse is a virtual world connected to the Internet in which individuals can interact with each other through their avatar or by handling digital objects. It therefore refers to the evolution of the Internet towards the Web3, with 3D or virtually integrated environments that users experience thanks to virtual reality and augmented reality. Thus, the metaverse is an interactive digital world.
What is Artificial Intelligence?
Simply put, AI refers to systems, algorithms or machines that mimic human intelligence. Its purpose is to enable computers to think and act like human beings. By combining AI with machine-learning, these computers learn over time from user behaviors and errors in their predictions. Often the AI analyzes a huge pile of data very quickly (big data) to then make better decisions or to act automatically.
AI is the most important technology used by the metaverse
The metaverse uses many technologies, such as virtual reality, augmented reality, blockchain, 3D, AI and even NFTs in some cases. Of all these emerging technologies, AI is arguably the most important in the puzzle. It can effectively help build the metaverse, develop it, adapt it to each user and improve their experience…
In a theoretical way: why artificial intelligence is essential for the metaverse?
The metaverse represents the digital future. Much more than a simple virtual environment, it would be a parallel world. Within this new world, individuals will be able to work, attend events, play, travel, visit, buy, interact with each other… In a parallel digital world, isn’t it normal and reassuring to have human intelligence through a computer program? In other words, to insert artificial intelligence within the metaverse?
Pragmatically: what are the use cases of artificial intelligence for the metaverse?
Let’s get to the most interesting and concrete part of this article. Five major benefits of using AI for the metaverse can be distinguished:
- Virtual characters;
- The avatar;
- Automation of repetitive tasks;
1. Virtual characters
This term refers to robots, characters existing only in the metaverse, and 3D versions of chat bots. It does not refer to avatars of individuals in the metaverse. These virtual humans will then be able to interact with the avatars by communicating and moving. They will have been previously programmed by AI algorithms. These virtual humans are essential to the functioning of the metaverse. They add a presence improving the user experience. A comparison can be made with the non-playing characters in video games like Grand Theft Auto.
AI can also make it possible to create an avatar that perfectly matches your appearance (face, build, size, etc.) rather than a completely different avatar. With AI, the user will be able to create his ” true avatar in the metaverse in addition to controlling it. AI is indeed a powerful tool that has become practically indispensable for certain beauty, cosmetics or clothing brands. AI actually allows a better choice of products offered by e-commerce brands, when they use it to have a 360-degree view of the physicality of users.
3. Automation of repetitive tasks
As we mentioned earlier, the metaverse should be conceived as a parallel digital world. Within this world, users (through their avatar) will do many activities. For example, they can send a letter, visit a brand’s virtual store, buy items… Technologies such as conversational AI platforms (CAIP), or more generally AIOps, will allow brands to design more attractive and engaging user experiences in an automated way. The term AIOps (Artificial Intelligence for IT Operations) refers to the use of Big Data, Machine Learning and AI to automate the identification and resolution of common IT problems.
The metaverse removes borders. However, language differences persist. And, they can interfere with user interactions. AI can then be the solution. It can immediately translate what is said by an avatar into the language of another avatar or a virtual character with whom it interacts. Technically, the challenge is to combine conversational AI (CAIP), voice recognition and self-supervised learning (self-supervised learning).
Concretely, the AI captures what was said by the avatar, analyzes it and converts it into the language of the second person and transmits it to him. If the AI has been trained (and collected enough data) in different languages, then it can translate easily. If you have trouble designing the creation of such a tool, I invite you to see the level of the translator Deepl. Having collected millions of texts translated into hundreds of languages, this translator instantly translates complex sentences, without grammatical errors, while respecting linguistic differences.
Historical data can be exploited, by combining ML (machine learning) and AI. This can then be used to make predictions about what the user is going to want or suggestions about what they might like. The more time the user has spent in the metaverse, the more personal data there is about them, the better the AI is able to pin down the person and become precise in these suggestions and predictions. Therefore, the scalability of the metaverse can only improve over time spent inside.
Zuckerberg, AI and the Metaverse
In this video from February 2022, Mark Zuckerberg highlights the importance of AI to the metaverse. The founder of Meta presents several ambitious artificial intelligence projects for the metaverse.
According to him, AI is the key to unlocking the growth and evolution of the metaverse. During this video, we see his avatar in a basic virtual world that he builds using the AI Builder Bot feature. He very quickly creates an island, trees and a beach within this world. Within the video, Mark Zuckerbeg also unveils a universal voice translator project developed using AI. With such a translator, the avatars of people speaking different languages will then be able to “to communicate with anyone in any language. It’s a super power we’ve always dreamed of.”.
Obstacles to overcome
It is essential to remember that the metaverse is only in the early stages of development. Many opportunities require time (and money) in research and development before they can be operational. In particular, the incorporation of AI into the metaverse must overcome several challenges. For example, we can cite:
- The protection of privacy. How to use AI ethically? What are the consent policies to put in place?
- Collecting a lot of data is necessary for artificial intelligence, which takes time. Until the metaverse is used frequently, AI models will run out of data. They won’t be able to learn enough to do effective forecasting and hyper-personalization.
- Creation and training of algorithms.
- Technological difficulties, which are addressed in this article.
- The designation of the owner of the data in a virtual world built with AI and who therefore has the right to use it. Who owns the copyright?
Thus, the metaverse could reach its full potential using AI. But, incorporating AI for different uses within the metaverse is not easy. Several challenges must be overcome and certain criticisms will no doubt persist. However, over time, AI will certainly emerge as the most important technology for the functioning of the metaverse. To quote Mark Zuckerberg, “ artificial intelligence is the key to unleashing the metaverse” .
Sources: VentureBeat, Coindesk, ITU, ArtificialIntelligence.com, VirtualReality.com, KDNuggets, Meta, Sensorium, Fintelics
Receive a digest of news in the world of cryptocurrencies by subscribing to our new service ofdaily and weekly so you don’t miss any of the essential Tremplin.io!