Name: Phi Nguyen


Assignment on Chapter 1 and 2



We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

Intelligence: to use logical thinking and execute best
course of action according to what goal one wants to achieve, which brings
about best result in term of utility.


Artificial Intelligence: a multi-disciplinary field
(mathematics, computer science, cognitive science… ) that creates a computer
system to solve problems, execute tasks that generally can only be performed by
human intelligence.


Agent: Subject of action, the one who “does”. Agent must
actively engage in the action and not controlled. Agent is independent and
executes actions based on its own rationality.


Rationality: a property of thinking that requires logic.


Logical reasoning: a type of thinking that processes through
the use of logic.



No, it would not. Because the goal of the ANALOGY model was
to score over 200 on an IQ test and that test alone, it can only be excellent
at one specific task but cannot do anything else otherwise. Intelligence, or
general intelligence, requires more than just being excellent at one task, but
to be able to do more, based on its own set of goals.



Because we can process a lot more than one single thought in
our brains. Our thought processes in not linear but dynamic direction in a
sense that we can have so many thoughts simultaneously.

Moreover, when being tested, different subjects provide
different results on the same stimulus, even those well-trained subjects are
used from trial to trial (Hockenbury & Hockenbury, 2006). Ethically,
introspection cannot be tested on children or animals. And lastly, complex
matters such as personalities, and so on.



Barcode scanner:
There is very little use of artificial intelligence in this case. The barcode
scanner basically decode the encrypted bar and locates the store as well as
determines the price, there is very little use of autonomous decision. It is
considered to be just another peripheral device.

Web-search engine:
There is a moderate use of artificial intelligence in this case. Take Google
for example, it most of the time detects typing errors and suggests the correct
ones. It also gathers data about what we have been searching lately, and
suggests potential search we might be interested in.

telephone menus: There is moderate use of artificial intelligence in this
case. Language is arbitrary; the sound we make has no reasonable connection
with the object we refer to. This machine has to make a decision referring to
what it is listening to and turns it into action (for example: “call Travis”)


Internet routing
algorithms that respond dynamically to the state of the network: this is a
complex task that contains a lot of dynamic problems, which needs an autonomous
agent to optimize the solution. So there is a large use of artificial
intelligence in this case.



The person’s brain has a large capacity for solving
complicated problems. There are trillions of neural networks for learning and
performing complex tasks like opening the door with a knob. The field for this
matter is convolutional neural network, which uses deep, feed forward network
to analyzes and solve such problem.



Evolution makes one stronger, because the one who is
stronger is the one who survives, who can adapt to new environment. And those
who survive are not only strong in the sense of muscle, but also of the brain
capacity to act rationally. The goal for them is obviously to survive so the
one who knows how to avoid preys, how to get food, mate… is the one who is more
likely to live longer and even transcend.

This biological fact has inspired computer scientists to
create Genetic Algorithm which has the same idea of mixing the best parents in
order to produce the best child (the ultimate solution)



There is one argument by John Searle supporting this claim
called “The Chinese Room”. The analogy goes like this: imagine there is a man
who only knows English inside a room full of Chinese word and matching English
word. People who turns in English sentences from the window and he would use
all the rules and grammars and matching vocabulary in the room to turn it to
Chinese. The question is, if he goes outside of the room, can he use Chinese
for general purpose?

If we refer to this analogy, we can see that the computer itself
is not “intelligent”, because all it does is following commands and rules and
grammar. Even though it performs better than human (Alpha Go), it only follows
what it was told to do.



Surely we are made of atoms, quarks and so on. So when we
move, it is the atoms that move, when we think, it is the chemical reactions
inside our brains – same go with the animals. When we take a human as an animal
(and yes we are), we still go with the same argument that we do not think or
act free but those actions are results of law of causation. However, there is
one argument against this claim called “The Mary Scientist”. It goes like this:
supposed there is a scientist named Marry who live in a black and white room.

She has never gone outside in her entire life, and never seen any other color
rather than black and white, but she knows everything about properties of
colors, in theory. Supposed one day, she is given a red apple, would she still
be surprised when seeing it?

This phenomenon is called qualia – to experience feelings.

Personally, I think human intelligence has more to do than only reasons. In the
other words, human intelligence also contains a large amount of emotion
capacity that artificial intelligence would take long to achieve.





I think the explanation above has answered this question. I
have nothing more to say here.



k. Performing a
complex surgical operation:

I think this has been able to achieve in the field of
Medical Internet of Things.

h. writing an
intentional funny story:

I think it is still attainable though I am not sure the word
“intentional” here makes a large difference to the task. I have seen a model
built by Stanford, in which they use the database of “How I Met Your Mother” to
compose new scripts which emulate Barney Stinson’s style of speech (the funny
guy in the show) and it did work.

j. Translating spoken
English into spoken Swedish in real time.

I think this is attainable though I would doubt the accuracy
of the translation. Simply it is composed of two artificial intelligence tasks:
voice recognition and text translation. And these two tasks have been solved
quite well in the real world.



a. The world in page 38 is partially observable,
deterministic and static. Hence the in completing the task, five functions are enough
(suck, right, left, up, down)

b. Yes. Because it needs to know whether the world has been

c. Yes. It can learn the geography to save time in cleaning
and learn from past experience to examine what squares tend to get dirty


a. False. Perfect rational agent will make good decisions
given that it has correct information from sensor

b. True. In an unobservable world, the agent would not be
able to navigate to act rationally

c. True. In a state where there is no action to be taken.

d. False. The agent programs takes input as agent current
percept. The agent function takes percept history

e. False. For example there is no agent that can solve both
searching, sorting, delete in constant time

f. True. When flip a coin

g. True. In a game with different map size, the agent would
still be able to perform regardless of the size

h. False. Because in unobservable world, the agent goal might
still be achieved.

i. False. This depends a lot on unknown probability.















Fully observable, multi-agent, stochastic,
sequential, dynamic, continuous, known

for used AI books

used AI books – cheap, usable




Partially observable, multiagent, stochastic,
sequential, dynamic, continuous, known

a tennis match


tennis court


agent, the opponent, the ball

Fully observable, multiagent, stochastic,
sequential, dynamic, continuous, known



a. No. The agent needs state of the environment to determine
whether it is clean or not.


c. A simple reflex agent would be fine since it has
omniscience knowledge about the world. And the reflex agent with state would
follow and update its state, then act accordingly


a. No. Since this agent does not have any knowledge about
the world, which leads to random changing direction through trials.


d. if the sensors stop working, the agent will start

Categories: Articles


I'm Garrett!

Would you like to get a custom essay? How about receiving a customized one?

Check it out