The AI ​​system used by Meta will be improved

Meta is currently conducting research based on the human brain model to improve artificial intelligence (AI). It should be noted that Facebook’s parent company plans to use AI to detect harmful content, develop translation tools, etc.

.

Most of AI systems that exist on the market, in particular those built by Meta are still far from satisfying the manufacturers. Until now, the AI still can’t learn the language as well as the humans. The giant of social media then conducts new long-term research that may change this reality. The project makes it possible to know more precisely the differences between the human brain and theAI in matter language learning.

AI system will be upgraded to promote futuristic products

The meta searchers try to bring an improvement to the AI ​​by betting on futuristic equipment. Among these products is a pair of augmented reality glasses that would allow you to find directions, display messages or play virtual games. To do this, the augmented reality combines digital images with a person’s real-world view. Jean-Remi King,

Woman wearing augmented reality glasses.

Jean-Remi Kingresearcher at Meta AIsaid the project could evolve theAI used for various tasks. The program would help improve the tasks of anticipate the continuation of a sentence or a expression such as words or ideas. Unlike humans who have no difficulty learning from small amounts of information, the AI ​​fails to perform simple tasks correctly. It is still difficult for these robots to do simple things even if they have large datasets.

Meta smart glasses in pictures

To specify that the virtual assistants better understand what a person wants if the task is well defined. However, dealing with these assistants can be frustrating, especially when they don’t know what to do. The modification that Meta is about to make to the AI ​​will therefore be able to improve its way of process language. Moreover, products like virtual assistants will experience great changes in the future.

Researchers compared AI language models to brain responses

To conduct this research, Meta partnered with the Center for NeuroSpin neuroimaging et al’INRIA in France. In fact, their main objective is to study how the brain reacts to certain words. As a result of this study, researchers were able to see the differences between human brains and models ofAIs trained to anticipate the next word.

Meta's Mark Zuckerberg in pictures

“Even though we are seeing tremendous progress in AI, it is still apparent to many that we still have a long way to go before we achieve human-level intelligence”

Jean-Remi Kingresearcher at Meta AI

In a study with INRIA, 345 people participated in the research program. This allowed the researchers to compare AI language models to the brain responses of volunteers. The latter listened to stories while their brain activity was recorded using functional magnetic resonance imaging.

In short, theAI improvement by Meta could revolutionize the world of augmented reality, but still change the lives of humans. Either way, researchers still have a long way to go before theartificial intelligence can reach the level of human intelligence.

We wish to say thanks to the writer of this article for this amazing web content

The AI ​​system used by Meta will be improved


Find here our social media profiles , as well as other pages related to it.https://www.ai-magazine.com/related-pages/