As we know, researchers have been always publishing AI papers, but the final goal has been lost for decades because of the difficulty and complications of designing a real AI. This is what we called Artificial General Intelligence, short as AGI. In this article, we will give a quick introduction to AGI, mainly on the definition of AGI and its importance.
WHAT IS AGI?
Human’s ability to solve real world problems contains something general enough so that we can read, speak, write, or even program, medical surgery, drive cars, etc. AGI similar to the general ability, also requires a much broad range of abilities for an AI agent. I found a general tool in mathematics, that is Game Theory to talk about this.
AI gets rapid development since 2012 ImageNet competition. This is a game on vision by many players. This competition is designed to stimulate research on AI. It is like an improving cycle.
What kind of Game we need to design? AGI can be seen as a giant game more complete existence.
So in our setting, AGI is a game that can influence many aspects of human future. Therefore, we must gather as complete insights as possible to shape a complete game setting. Can we prove the convergence of this game or find the Nash equilibrium or something similar to it.? This is a rigorous way to search solutions to AGI.
AGI is a game that everyone can be a player in it.
Components of AGI game:
Solution to the AGI game, how to fast reach it.
Nash equilibrium definition.
Countries, institutes and researchers
WHY SHOULD WE TRY TO FIGURE IT OUT?
AGI is going to be one of the most influential science/technology developments in human history. Someone once said that we are the last generation of pure human. Now everywhere we can see AI products changing our life from daily life to NASA space traveler.
People like Elon Musk, Bill Gates, Stephen Hawking considering the bad effects of AGI if we can not control it.