Evolución de la IA - (3/7)

Evolution of Artificial Intelligence: Key Milestones

Introduction to the Evolution of AI

• María Florencia introduces the third part of a video series on the evolution of artificial intelligence, emphasizing its connection to digital transformation and disruptive technologies.

The Law of Moore

• The Law of Moore, formulated in 1965, predicts that the number of transistors per unit area in integrated circuits would double annually.

• This law was later modified to suggest a doubling every two years, although some believe it is every 18 months due to Intel's executive influence.

• The Law is not a mathematical or physical law but rather a predictive guideline that has been largely upheld over decades.

• It highlights technological obsolescence, where older technology becomes outdated as newer versions with improved features are released.

Technological Progression and Communication

• The transition from copper to fiber optics illustrates advancements in communication technology influenced by algorithmic improvements.

Genetic Algorithms

• In the 1970s, genetic algorithms were introduced by John Henry Holland at the University of Michigan, inspired by natural selection principles.

• Holland's initial goal was for computers to learn autonomously; he termed his technique "reproductive plans."

Functionality and Limitations of Genetic Algorithms

• Genetic algorithms optimize problem-solving when parameters are extensive and non-parallelizable, transforming solutions through reproduction, selection, and mutation processes.

• They do not guarantee optimal solutions but rather provide multiple viable options based on biological inheritance principles.

Early Expectations vs. Reality in AI Development

• Initial expectations for AI included revolutionary impacts on computing; however, this led to what is known as the first AI winter due to unmet promises.

• Challenges arose from insufficient computational power at the time; computers struggled with complex problem-solving capabilities.

Decline in AI Research Funding

• A doctoral student's observation highlighted that early computers lacked strength for intelligent displays leading to reduced funding and stagnation in research during the first AI winter (approximately a decade).

Resurgence Post-AI Winter

• After this period, significant milestones emerged as researchers began focusing their efforts more strategically rather than exploring random avenues without clear objectives.

Knowledge Engineering Concepts Emergence

• Following the first winter's end, concepts like knowledge engineering gained traction as researchers sought structured approaches within AI development.

The Law of Mercalli

Concept of Systems and Knowledge Engineering

Understanding the Concept of a System

• A system is defined as a set of interrelated elements working towards a common goal, emphasizing the importance of relationships among these elements.

• The integration of concepts from McAlf and Moore suggests that digitalization significantly enhances the utility of devices without drastically increasing costs.

Emergence of Knowledge Engineering

• The end of the first winter in artificial intelligence leads to the introduction of knowledge engineering, which is crucial for implementing projects effectively.

• Knowledge engineering arose from challenges faced during early software development crises, where initial solutions could not be sustained over time.

Characteristics and Architecture of Knowledge-Based Systems

• Knowledge-based systems are characterized by an explicit knowledge base, which includes data and rules stored in a specific location. This forms the foundation for their architecture.

• Key components include an explicit knowledge base, input/output interfaces, and an inference engine that processes data and rules to facilitate reasoning.

Expert Systems: Emulating Human Expertise

• Expert systems aim to replicate certain behaviors or decision-making processes characteristic of human experts within specific domains. This emulation is central to their design.

• The spirit behind artificial intelligence has always been collaboration rather than replacement; it seeks to assist rather than undermine human expertise.

Conclusion on Key Concepts