The History of Automated Decision-Making (ADM)
The history of decision automation is closely linked to technological and computer breakthroughs. Here's an overview of significant events in decision-making automation history.
Timeline
- 17th - 19th centuries: first mechanical calculators
- 19th century: punch card systems
- Mid - 20th century: early computers
- 1960s: decision support systems
- 1970s - 1980s: expert systems
- 1990s - 2000s: machine learning and data mining
- 2010s: big data and analytics
- 2010 - present: artificial intelligence and deep learning
- 2010 - present: ethical and regulatory concerns
- 2022: breakthrough of ChatGPT
First Mechanical Calculators (17th - 19th centuries)
The first mechanical calculators, such as the Pascaline (1642) by Blaise Pascal and the stepped reckoner (1672) by Gottfried Wilhelm Leibniz, can be traced back to the concept of automated decision-making. These devices were created to automatically conduct repetitive computation, allowing for more efficient decisions in domains such as mathematics, engineering, and navigation.
Punch Card Systems (19th century)
The use of punch cards for automatic decision-making became common in the nineteenth century, particularly in industries such as textiles and manufacturing. These cards had holes that represented data, and machines read and processed them, making judgements based on the patterns of the holes.
Early Computers (mid-20th century)
The introduction of electronic computers in the mid-twentieth century was an important milestone in the history of automation of decisions. ENIAC (1945) and UNIVAC (1951) computers were used for a variety of applications, including scientific calculations, weather forecasting, and military operations.
Decision Support Systems (1960s)
The 1960s saw the emergence of Decision Support Systems (DSS). These computer-based solutions were created to help humans make decisions by providing data analysis, modelling, and interactive capabilities. DSS are still used today.
Expert Systems (1970s - 1980s)
In the 1970s and 1980s, expert systems, a subset of artificial intelligence, came up. In specialised domains, these systems are aimed at mimicking human expertise and decision-making. These systems were primarily rule-based, making decisions through a sequence of 'if-then' rules. They were used in the medical, financial, and technical industries. These systems had a number of drawbacks, including the fact that they were frequently domain specific, complex to programme and maintain, and had little adoption among experts and users.
Decision support systems vs. expert systems »
Machine Learning and Data Mining (1990s - 2000s)
As machine learning algorithms improved and massive datasets became more available, automated decision-making systems began to rely more on data-driven approaches. Techniques such as machine learning and data mining were employed for tasks such as recommendation systems, fraud detection, and predictive analytics.
Big Data and Analytics (2010s)
The decade of the 2010s saw a spike in big data technologies and analytics tools. Large volumes of data were collected and analysed by organisations in order to make data-driven decisions, optimise processes, and gain insights into customer behaviour. Since 2010, self-service analysis tools such as Microsoft's Power BI or Salesforce's Tableau have become popular.
Artificial Intelligence and Deep Learning (2010 - present)
Deep learning and artificial intelligence (AI) have transformed the automation of decisions. AI systems driven by neural networks and deep learning algorithms can handle and analyse complicated data, anticipate outcomes, and automate making choices in a variety of applications. Popular applications include: autonomous cars, natural language processing, and the recognition of images.
Ethical and Regulatory Concerns (2010 - present)
The growing use of automated decision systems has given rise to ethical and regulatory concerns. Bias, fairness, transparency, and accountability issues have become critical since these systems touch different parts of society, such as employment, banking, healthcare, and criminal justice.
Breakthrough of ChatGPT (2022)
Multilingual models, such as ChatGPT, are one of the most effective uses of AI to automate. Because they interpret natural language and provide access to a variety of information, these systems are extremely user-friendly. Moreover, they work quickly and cheaply. At the same time, this development emphasises the importance of ethical dilemmas in the field of artificial intelligence.
Symbio6 & ADM-History
Throughout history, technical advancements have increased and improved decision-making automation. However, these advancements have also raised worries about the ethical and social implications. Symbio6 continuously monitors these developments in order to deliver the best service possible to our clients.