Question

As a human and using your own human Intelligent Analysis (IA), you would have noted that as humans we can practice learning and adaptation and we can also learn by recognition versus understanding because understanding is different to recognition. Based on this activity and reflection on your to date knowledge of smart/Ai agents and current smart systems you are familiar with, do you think if we can transfer the way we learning and create sets of learning and instructions to train smart bots or Ai agents by representation, evaluation, and optimization, we can then train computers to do just about anything we are capable of learning ourselves? Why or why not?

Answer 1

Answer:

We just can't train AI to do anything that we are capable of because an AI system learns from historical data and parametrises that learning into higher order or logic, or pattern recognition and does the job. Apart from processing information logically, human brains also have other capabilities. Let's list them down:

Sensory information: Humans use all the five senses all the time. AI will operate only in the realm of information processing.

Creativity: We can train a machine to train like us, but we can't train it to develop it's own personality, and that breaks the chain of creativity.

Social behavior: A machine can't make friends or love someone.

Deductive reasoning: Computers are good at processing logic and data at higher speeds, but they can't create stuffs and break new grounds of knowledge like we do.

Above all, we reproduce to survive but a machine can't do that.