Kartenstapel

• For any problem which can be translated algorithmically

• A Turing machine is a theoretical concept of a machine which reads in information and executes orders accordingly and gives back some output, it is said to be the basis of the concept of modern computers

• The Turing machine can be visualized, by a little box on wheels which moves along a tape of text. It moves on by one and reads in the content of each position. If there are certain orders for the specific content at a certain position the Turing machine is able to execute them (e.g.: erase, print, move)

• The Von Neumann architecture is a concept how a universal PC should work

• It should be seperated into different logical and spatially parts, i.e. CPU and memory

• The CPU communicates with the memory as it gives it the state information, then the instructions are transfered from memory to CPU, followed by the results from the CPU to the memory

• Data has a certain adress (i.e. the storing position) and accordingly a value

• An instruction is defined by the two parts, what should be executed (operating code) and with what should an order be executed (operand)

• A personal computer (PC) is a multi-purpose computer whose size, capabilities, and price make it feasible for individual use

• Turing: depends on the definition of thinking and machine

• Media: Articles about new technological develpments, such as Alpha Go, Deep Blue, etc. Often ambiguous and unprecise articles or statements

• Technology: Incredibly fast and large computational power

• Biology: Deep neural networks are based on the template of biological neural networks

• Cognitive AI

• Quantum AI

• Neuromorphic AI

• Nano-AI

Step 1: Naive algorithms –> repeating

Step2: Machine Learning –> imitating

Step 3: Deep learning –> learning

Step 4: Deep reinforcement learning –> Learning to learn

Step 5: Distributed agents and swarm deep reinforcement learning –> contribution and exchange