The AI can think. But it's thinking is different than human's thinking. The AI can collect new images from its hard disks like puzzles. But even if we think that AI makes impressive new entireties about things, that are stored in its memory. The AI can't make all things better than humans.
Humans recognize faces better than AI. The reason for that is in logic that the AI uses. The AI's way of using fuzzy logic is different than the human brain. In AI fuzzy logic is a series of precise logic algorithms. That means the AI has problems comparing two images like faces, if the images are taken from different angles. The AI compares two images by making a matrix of them. And then this system puts those images overlap each other.
The system doesn't see those images the same way as we see them. Artificial intelligence sees the color and brightness values of images as RGB codes. If there is a difference in light conditions between two images there are different RGB codes. And that thing makes those images seem different in the eyes of AI.
The human brain sees the image as an image. That makes the human brain less accurate. The brain doesn't remember all the details of the image like AI. The brain remembers some details or main characters about images. And then senses fill those frames. The brain uses more blocks than the AI. So that means the AI is not as good as the brain. If the image that it should compile is from another direction. Or there are some kind of shadows or some other differences from the original images.
When a certain part of the frame matches with memory, the brains check the databases that are connected with that memory. The brain handles memories as entireties. In the center is an image. And then the other memory cells connected with that image. Those other cells cooperate with cells that control movements or some other things.
Memory is one interesting thing in the human brain. The purpose of memory is to save people. This thing means that we remember bad things better than good things. The memory is not created for the past. It's created for the future. The memory's purpose is to make us learn from our mistakes. The memory is a library that the brain can use in the future for solving problems.
For the AI the memory is different. Memory is a static thing for computers. And that means the AI can remember things, that were once stored in a hard disk like those things that happened yesterday. Computers don't forget things. They don't learn new things like humans either. The traditional way to make computers learn new things was simply by making programs for them.
The automatic learning process is harder to make. The computer requires algorithms that determine details of what it should remember. Or if the computer stores every single detail from the day to the hard disks that system requires lots of hard disk space. The human brain has impressive memory capacity but in the same way, there are limits to its capacity.
https://bigthink.com/the-present/facial-recognition-ai/
https://bigthink.com/the-learning-curve/why-memory-is-more-about-your-future/
Comments
Post a Comment