Recently, while experimenting with TensorFlow, I messed around with a few predictive algorithms. Right now, I’m trying to write a program to identify flowers given a fed-in dataset. Because of such predictive programs, I learned about how neurons connect pathways in machine learning, and hence, how neural networks form. Network neurons can be compared to the actual neurons in our brains. They send synapses, or signals, to the neurons around them in order to relay and process data more efficiently and accurately. Furthermore, these neurons work together to create what’s known as a neural network, a core foundation for machine learning. In a previous blog post, I discussed Dall-E 2’s drawing creation and uses. The project itself had millions of neural connections, and can hence, generate images in many mediums and styles specific to given instructions. After experimenting with IBM’s Watson AI Bot, I noticed that neural interfaces could be created to separate out genres of text. Different neural groups can be attached in order to create multifunctional bots to do a desired thing. Furthermore, scientists experimenting with AI robots, such as Japan’s Sony Aibo Dog robot, implement such techniques in order to help the robot learn. In the case of Aibo, the dog is able to perform tricks, respond to commands, and etc.