I don’t have enough experience in Artificial Intelligence to provide you sufficient and promising answer buy as per my understanding:
- An Artificially Intelligent Machine has achieved some level of accuracy in Image Processing, machine learning is already 97% accurate in comparison to Human, sound recognizition is also on its hipe, Deep Learning has increased the performance of many tech giants, Internet of Things today is all about being of service with some intelligence.
But:
Artificially Intelligent Machine can watch a music video process each image frame and at the same time can recognize the voice and the stanzas. But it fails to perceive the sensation of the music and also lacks the knowledge about what message is the song trying to give.
Computer Science still needs much work in combining the existing achievement and make them a composite intelligence first before moving further. - Defining AI: May be there is still not a concrete vision about AI and what sort of Intelligence are we expecting from a machine. Everyone seems to be applying intelligence in their machine. But I would love to call most of such machine are just solving problems by fetching the solutions(patters) stored in their memory by their programmer).
- We also fail to feed the knowledge about cognitive science in a machine.
- Human Brain is a mystery box to lean into and dig deep for understanding how it performs the computation in fraction of second. And having understood in detail about human brain scientists replicate neural networks. It works very efficiently than anything better, but still solving the mystery of human brain could lead the breakthrough in AI
- The architectural design of the microprocessor limits the computation speed when it comes to the billions of information processing to answer a basic query.
(Read this point here somewhere in Quora)