AI is everywhere. Be it healthcare, transportation, predicting natural disasters, farming, strengthening economy AI is doing everything for us through supervised learning. But what if AI becomes so advanced that it could enable a robot to think on its own?
Just like a human, computers learn mostly through observation or trial and error. Through supervised learning, researchers feed the computer data, and from that, it can learn things. But this restricts the knowledge, making AI far from achieving human-level intelligence.
Researchers then turned to Reinforcement Learning, wherein AI sets a goal and keeps working towards it through trial and error until it is consistently receiving a reward.
The researchers agree on working towards Predictive learning which recognizes patterns, predict outcomes, and choose a course of action. Some intend to get there by reinforcement learning, whereas some by supervised learning.
But to achieve human-like intelligence, they need to work with “self-supervised learning.” By this, they mean a system that can ingest vast amounts of unlabelled data and make sense of it without supervision or reward just like a human can do through common sense. That would mean it can teach itself, hence creating a mind of its own. It’s just a matter of time before humans could build robots as Intelligent as themselves.
#AIMonks #AI #ArtificialIntelligence #SupervisedLearning #ReinforcementLearning #SelfSupervisedLearning #PredictiveLearning
0 Comments