Neuromorphic computing is an approach to hardware and software modelled on the structure of the brain. Instead of moving data between separate memory and processor in clock cycles, artificial neurons and synapses work in parallel and respond only to events known as spikes. This promises very low energy consumption, for example in sensors or robots. Well-known research chips include Intel Loihi and IBM TrueNorth; broad commercial use is still at an early stage.
Artificial neural networks (ANNs) are computational models made up of many connected units (“neurons”) arranged in layers that process inputs via weighted connections. During training, the weights are adjusted step by step using examples, usually by backpropagation. ANNs are the foundation of deep learning and power image recognition, speech recognition and large language models. Despite their biological inspiration, they work far more simply than the human brain.
An application-specific integrated circuit (ASIC) is a chip designed for one particular task rather than general-purpose computing like a CPU. In AI, ASICs are tailored to the arithmetic of neural networks – Google's Tensor Processing Units (TPUs) are one example. For that task they run faster and more energy-efficiently than general-purpose chips, but they are expensive to develop and hard to adapt to new workloads.
Quantum computing uses effects from quantum physics such as superposition and entanglement to compute with qubits instead of classical bits. Major speed advantages are expected for certain tasks, such as simulating molecules, some optimisation problems or breaking today's encryption methods. However, the machines are still error-prone and mostly used in research and pilot projects (as of October 2026). For today's AI applications, quantum computing plays practically no role yet.
Cognitive computing is an umbrella term for systems intended to emulate human thought processes such as understanding language, reasoning and learning from experience in order to support people's decisions. The term became widely known in the 2010s, mainly through IBM Watson. It combines methods such as natural language processing, machine learning and knowledge representation; today people more often speak of AI assistants or language models instead.
