AI-Presentation.github.io

The Evolution of Artificial Intelligence: From Logic Machines to Agentic AI:

Introduction:

• Defines AI and outlines its journey from rule-based logic to autonomous agents.

Early Foundations (1940s–50s):

• Alan Turing’s work and the Turing Test establish AI’s conceptual base.

Dartmouth Conference (1956):

• Birth of AI as a field, led by McCarthy, Minsky, and others.

Symbolic AI (1956–70s):

• Logic-based systems aimed to mimic human reasoning.

First AI Winter (1970s):

• Overhype and technical limits cause funding cuts.

Expert Systems (1980s)

• Knowledge-based AI achieves success but struggles to scale.

Machine Learning (1990s–2000s)

• Shift to data-driven models; notable success with Deep Blue.

Deep Learning (2010s)

• Neural networks and GPUs drive breakthroughs in vision and speech.

Transformers & Generative AI (2017–2020s)

• GPT and similar models enable text, image, and code generation.

Agentic AI (2020s–Present)

• Autonomous, reasoning agents emerge (e.g., AutoGPT, OpenAI o1).

Key Figures

• Turing, McCarthy, Minsky, Hinton, LeCun, and Bengio shape AI’s development.

Future Outlook

• Focus shifts to ethical, safe, and collaborative human–AI systems.

Large Language Models work by:

1. Predicting the next token using the Transformer architecture.

2. Learning from massive amounts of text data.

3. Scaling up with billions of parameters and advanced hardware.

4. Being fine-tuned with human feedback for reliability and alignment.

Era	Dominant Approach	Paradigm Shift
1950s–70s	Symbolic reasoning	Logic → Knowledge Representation
1980s	Expert systems	Rules → Domain Expertise
1990s–2000s	Machine learning	Hand-coded → Data-driven
2010s	Deep learning	Manual features → Representation learning
2020s	LLMs & Agentic AI	Prediction → Autonomous reasoning