Module 2 - Cognition and Computation Flashcards
Explain ‘‘The Computer Metaphor for the Mind’’
The brain is like hardware, and the mind is like software.
AI is based on the idea that thinking = computation
Physical Symbol System Hypothesis
Intelligence requires the ability to manipulate symbols.
AI must process symbols to reason and make decisions.
What are the two limits of computation?
Intractable problems: Require too much time/memory to solve.
Uncomputable problems: No algorithm can solve them (e.g., the Halting Problem).
What is the difference between tractable and intractable problems?
Tractable problems: Can be solved in reasonable time (e.g., sorting algorithms).
Intractable problems: Require exponential time, making them impossible to solve efficiently (e.g., Traveling Salesman Problem).
What are the three models of universal computation?
- Von Neumann Architecture (memory + processing unit).
- Turing Machine (abstract computation model).
- Neural Networks (learning-based model).
In Depth Q. Why do AI systems struggle with uncomputable problems?
Some problems cannot be solved by any algorithm (e.g., predicting if a program will run forever).
AI can approximate but not solve these problems perfectly.
In Depth Q. How does the Physical Symbol System Hypothesis relate to modern AI?
Traditional AI relies on symbol manipulation, but deep learning uses data-driven approaches.
Some argue modern AI doesn’t need symbolic reasoning to be intelligent.
How have we understood thinking throughout history?
The question “What does it mean to think?” has changed over time as technology changed.
Each era compared the human mind to its most advanced machines of the time.
Different Metaphors for the Mind + Era
- Renaissance (1500s-1600s) - Clockwork mechanism – Thinking is precise & mechanical.
- Industrial Revolution (1760-1840) Steam Engine – The mind works under pressure like a machine.
- Early 1900s Telephone exchange – Thought is a network of signals & connections.
- Modern era Computer metaphor – The mind processes information like software running on hardware.
Why do the metaphors for the mind change?
Each metaphor reflects what humans know at the time.
As technology advances, we see more complex patterns in how the brain functions
What is the Computer Metaphor?
Mind = Software → Processes information, makes decisions, does not have a physical form.
Brain = Hardware → Physical structure, executes computations, requires energy.
What did Johnson-Laird (1983) state?
Stated that the computer metaphor is the last metaphor – meaning it will never be replaced.
No other metaphor has been studied in such depth.
Why is the Computer Metaphor Powerful?
It explains how the brain stores, retrieves, and processes information.
It allows scientists to model human thought using artificial intelligence.
Criticism of the Computer Metaphor
❌ Does not fully explain emotions, intuition, and consciousness.
❌ The brain is biological, while a computer is electronic – this may be a fundamental difference.
What is Computation?
Computation is the process of solving problems using a sequence of logical steps.
Thought was once seen as purely calculation-based.
What happend during Age of Enlightenment (1700s – Leibniz)?
Believed thinking was like mathematical reasoning.
Even moral and ethical questions could be solved with logic.
What happend during Industrial Revolution (1760-1840 – De Prony)?
Computation became a repetitive, mechanical task.
Talented thinkers were seen as more than just calculators.
What was the first ‘‘computer’’?
Before machines, human workers were called computers.
They performed complex calculations manually.
What happend in 1822?
The first step toward programmable computers. The Difference Engine was the first mechanical calculator.
Who imagined the possibility to replace the first ‘‘computers’’ with machines?
Charles Babbage (1812)
Why was Babbage’s idea revolutionary?
Shifted the idea that only humans could compute.
Led to the question: “Can machines think?”
How Did WWII Accelerate AI Development?
Codebreaking & military needs pushed technology forward.
Computers evolved from calculators to problem solvers.
Key Milestones in Computing History
- 1941- Zuse’s Z3 Germany- First programmable computer.
- 1943- Colossus UK-Used for code breaking (WWII).
- 1946- ENIAC USA- Military & scientific calculations.
What question did Alan Turing (1912-1954) asked?
Asked “Can machines think?” rather than “Is the mind a computer?” Saw computers as mimicking human behavior, but not necessarily thinking like us.
What did John von Neumann say about computers?
Compared computers to the nervous system. Agreed computers were like the brain, but not the same as the brain.
What’s the difference between the view of Alan Turing and John von Neumann?
Turing - AI can mimic thinking but does not prove real intelligence.
Von Neumann - The brain and computers share similarities but are not identical.
What did Newell & Simon propose about cognition?
Cognition is not just number-based calculations but involves symbol manipulation.
Computers should be seen as symbol processors, not just calculators.
What is the Mental Chemistry Hypothesis
Proposed by Newell & Simon.
Thought processes are like chemical reactions, where simple symbols combine into complex structures.
What are the basic operations in mental chemistry, according to Newell & Simon?
- Storage – Keeping symbols in memory.
- Retrieval – Accessing stored symbols.
- Transmission – Moving symbols between processes.
- Comparison – Evaluating and manipulating symbols.
What did Simon predict about AI in 1957?
“Within 10 years, psychological theories will take the form of computer programs.”
Suggested that computers would be used to model human thought.
What is the Physical Symbol System Hypothesis?
Proposed in 1976, it states that:
1. Intelligence requires symbol manipulation.
2. Symbol manipulation is enough for intelligence to emerge.
Why is symbol manipulation considered necessary for intelligence?
Humans use symbols to reason and solve problems.
AI models that manipulate symbols can replicate intelligent behaviors.
Why is symbol manipulation considered sufficient for intelligence?
If a machine can store, retrieve, and process symbols, it can act intelligently.
This suggests AI doesn’t need human-like consciousness to be intelligent.
Who can be considered a Physical Symbol System?
- Humans – Neurons process symbolic representations.
- Machines – Computers manipulate symbols.
- Other physical systems – Anything that can perform symbolic computation.
Does the material of intelligence matter in the Physical Symbol System Hypothesis?
No, intelligence can be made of neurons, silicon, or tin cans—as long as it processes symbols.
The process of computation matters more than the substance of the system.
