The human brain serves as a model
Cognitive technologies imitate functions of the human brain. These include, for example, natural language processing, data mining and pattern recognition.
Watson “thinks” almost like a human
An outstanding example of cognitive technology is IBM’s Watson supercomputer. It “thinks” at least as well as the human brain. Its processing rate is 80 teraflops. The open multi-cloud platform is used, for example, to automate an AI lifecycle or accelerate value creation times with pre-built enterprise applications.
Human thought includes consciousness
Of course, Watson doesn’t really think, as it lacks an awareness of the commands it executes according to parameters set by humans. Nevertheless, cognitive technologies are increasingly capable to draw conclusions from the comparison of collected data.
At Google, the motto is: AI first
Another prominent example is Alphabet from the big concern Google. As reported by Handelsblatt,the CEO Sundar Pichai describes artificial intelligence as the most important thing mankind is currently working on and the strategy “AI first” applies to all nine of the company’s products. According to the report, around one billion active users benefit from it every month when they surf the Internet, let themselves be navigated by Google Maps or post pictures in Google Photos. In the future, they are to receive more and more active suggestions and alternatives to their search queries through cognitive technologies, over and above pure information queries.
Streaming services also actively provide user-related suggestions
Netflix is pursuing a similar strategy. The streaming media service uses cognitive technology to generate recommendations that are created according to the subscriber’s user list and are continuously supplemented and optimized. Much like the Spotify music streaming service.
Cognitive technology is creating a paradigm shift in business
Looking at the growing use of cognitive technology, consulting firm Deloitte sees a paradigm shift in business that has been driven by innovative enterprise architectures. Thus, it says, a very practical application area in the technology sector has emerged from a subset of the scientific research field on artificial intelligence. According to Deloitte, this is due to increased mergers and acquisitions (M&A) activities, among other things.
The most important areas of application at a glance
According to Deloitte, it is primarily network and semiconductor manufacturers, hardware producers, IT providers, software developers and Internet players that are interested in using cognitive technologies. According to the study, the main areas of application so far are defined as follows:
- Computer vision can filter complex patterns from unconstrained (that is, naturalistic) visual environments through special programming. This enables computers to recognize individual objects, scenes or activities, for example in road traffic, and react accordingly.
- Machine learning defines the ability of computer systems to use cognitive technology to autonomously sift through data about their performance on an ongoing basis, analyze it, and adapt it in an optimized way to meet requirements. Machine learning is divided into three areas: Supervised learning is used, for example, in text recognition or sales forecasting. Unsupervised learning, on the other hand, is found in cluster analyses. Reinforcement learning is used in traffic control, autonomous driving or even robotics.
- Natural Language Processing (NLP) enables computers to work with text. By processing natural language, the programs can identify meanings and assign synonyms. One of the intended goals is to produce computer-generated texts that are easy to read, stylistically natural and grammatically correct.
- Speech recognition programs are well known thanks to Siri and Alexa. Cognitive technology automatically transcribes human speech, provides queried information, and actively supplements it with appropriate suggestions, which in turn are generated in an individualized manner by cognitive machine learning technology, for example.
- Optimization is about using cognitive technology to find the best possible solution, e.g. for a work process. This involves automating complex decisions and trade-offs when resources are limited.
- Planning and scheduling should also be seen in this context. It is about using cognitive technology to develop a sequence of activities in such a way that the goals are achieved effectively and optimally, even while complying with, for example, time constraints.
- Rules-based systems use databases of existing knowledge to automate a process. Cognitive technology, however, further draws conclusions from this existing information, for example, to correctly classify larger and more complex texts in terms of content.
- Robotics is located in an area between artificial intelligence and machines. For example, in the production of vehicle parts, robots have long been an important component, e.g. in order to carry out activities that are harmful to health, such as painting, as a substitute for humans. Cognitive technologies are also used for collaboration or interaction between humans and robots. In doing so, they seamlessly integrate computer vision and automated planning, for example, with small high-performance sensors, actuators and other hardware.
- Brain-computer interfaces (BCI’s) and androids also spring from cognitive technologies. This involves translating the power of human thought into the digital. We reported on the methodology and developments from this field already back in 2018. Epilepsy patients can be warned of an impending seizure by BCI devices, for example. Developments have progressed rapidly since 2018 and BCI’s are not only used in e.g. medical research, but have now become suitable for mass use. For example, an end user of a BCI device can monitor and evaluate their own sleep or monitor the effects of their meditation sessions.
Magility ensures data security
But for all the blessings promised by cognitive technologies, magility Managing Director Dr. Michael Müller points out that this involves the processing of sensitive data that needs to be protected. According to the security expert, the installation of security systems and constant monitoring of the data flow could reliably prevent hacker attacks, industrial espionage and data theft.