Turing Technology Founder Amir Mamaghani is here to talk about cybernetics and artificial intelligence…
I’m often asked very interesting questions by peers, one recent question that stuck out to me was ‘What are cybernetics in relation to AI?’.
I’ve used my expertise and experience to best answer this question below.
The word ‘cybernetics’ comes from the Greek ‘κυβερνάω’ (cybernáō), meaning ‘to steer, navigate, or govern’.
So as you might expect, this field is very wide and contains many different scientific disciplines.
Different perspectives of cybernetics:
- Technical & industry based
- Informatics & theory of information
- Management & theory of management
Personally, I’m heavily involved in the first and second category of cybernetics, so I’m going to pay close attention to these sections. First, let me put it into the correct order, because cybernetics has several definitions and categorizations that may cause confusion.
In the technology perspective, cybernetics is also known under a different name — AI, Automation & Measurement. This general field consists of several areas such as:
Control & Regulatory systems — (theory of automatic control, automation) for example, the thermostat (Honeywell are a well-known producer of these systems). This area aims to find opportunities for controlling the real-world. A control system manages, commands, directs, and regulates behavior through control loops.
Machine vision — Machine Vision (MV) is the technology and methodology used to provide imaging-based automatic inspection and analysis for applications based on image processing, recognition and perception (for example, industrial robot guidance). Machine Vision is a term encompassing a large number of technologies, software and hardware products, integrated systems, actions, methods and expertise.
In short, inputs for MV are imaging devices, like cameras or lasers, which produce an image, later to be pre-processed (like contrast adjustment, thresholding or colour calibration), then processed (feature extraction), then detection or segmentation (face detection, spot detection) and finally used as an input (neural networks or Haar cascade filters for facial recognition).
Sensors & Measurement — For robots to observe and interact with the real world, sensors have to be used. These sensors have different characteristics and properties, so this field is all about knowing how to choose and how to work with different sensors. This field is very important for robotics, so that robots have the proper and correct information from their surroundings.
Robots are machines that interact with the real world. Robotics looks closely at the design, construction, operation and use of robots in the real world, like in industry, military or entertainment use. Robotics also looks at the computer systems for control, sensory feedback and information processing. Robots are often created to replace human activity, especially in dangerous or repetitive situations.
AI is a general field consisting of many different approaches that give programmes the ability to learn without being explicitly programmed to do so. AI mainly describes the process of learning and working towards solving a problem by decreasing the error. It’s mainly used in fields where you cannot accurately describe the problem mathematically.
If you’re keen to learn more, I recommend reading these Wikipedia pages to get a good general understanding. Use these as the basis for further research, and take a look at the footnotes on each page for further academic resources to improve your knowledge.
Thanks for reading,