WHAT ARE THE MERITS OF TURING TEST IN ARTIFICIAL INTELLIGENCE -

Discuss the merits and faults of Turing’s criteria for computer software being “intelligent“.

Turing Test is a form of investigation used to assess if a machine is capable of reasoning like a human being. The test was carried out by Alan Turing in 1950. It measures the capacity of a computer to demonstrate intellectual behaviour that is equal to or distinct from humans.

Turing Test is often related to here as Game of Imitation. It’s got three terminals. Each terminal is segregated from the other two terminals. One of the terminals is operated by a computer. The remaining two terminals are run by humans.

One human is acting even as respondent and the other human and the machine is acting as the respondent mostly during examination. To use a particular format and meaning, the questioner questions the respondents about a specific topic. During a certain lot of respondents or a specified amount of time, the questioner is asked to determine which respondent was a machine and which is the an actual being

The procedure is repeated multiple times. When the questionnaire is unable to differentiate between the machine and the human, the machine is known to have augmented reality.

Describe your own criteria for computer software to be considered “intelligent.”

Coupling the device to the physical world would be one change over the Turing test: to test a robot. At this point, the Turing test as such could be too complicated, because the outside observer could probably see the difference between a robot and a human being, and to know exactly what the robot is doing at all times, the observer would probably need a real image of the robot instead of just a summary of what it was doing.

would’ve been will be would also be would have been might be

would indeed be would really be More

The Survivor Test is then proposed. In this one, as in those famous reality TV shows, we stick a robot with a group of real people. For many weeks, the robot has to communicate with everyone else. As on television, to exclude individuals from the game, a vote must be taken at regular intervals. In this scenario, questions such as: We should correct the requirements for judging:

Sympathetic behaviour in comparison to others.
Capable of getting what he / she / it needs.
Cooperation capacity.
Knowledge of general kind

The parameters are all subjective, much as in the Turing test. Yet we are now focused on human intelligence as well as social intelligence..

Prove that modus ponens is sound for propositional calculus. Hint: use truth tables to enumerate all possible interpretations, then show that wherever the premises are true, the first line in the truth table below, the conclusion is also true. (10pts)

P Q P ⇒ Q P ∧ (P ⇒ Q) (P ∧ (P ⇒ Q)) ⇒ Q

4)Give aheuristics that a block-stacking program might use to solve problems of the form “stack block X on block Y.”

Explain your answer.
Is it admissible?
Is it Monotonic? .

As a running example for an object domain, we use the well-known blocks world (BW), which is arguably the most frequently used benchmark domain in the context of RRL. We use Event Calculus and generate a easy explainatio. The BW is scalable and no mre a trivial from the computational viewpoint. BWs are good to use with RRL, since BW is small, hence it can compactly represen by using a relation based anguage.

A BW is a relationally structured domain in which an agent observes and acts in a sort of grid with discrete positions. At each position there can be either a block (named as blockA, blockB, …), or the table, or nothing. Since some blocksx are on top of some other block y is expressed with on(x,y).on(x,table), which mean that block x is directly on the table. clear(x) denotes that there is currently no block on top of block x. Fluent clear seems not necessarily to be used but clear is useful short-cut in configuration of fluents with predicate on.The agent acts in the BW by moving blocks using a “stacking” action, conditioned by certain pre- and post-conditions (e.g., both the moved block and the target of the move need to be clear in beforehand). A stacking action which moves block x on top of block y is expressed with stack(x,y). If the action succeeded, subsequently on(x,y) holds.

The presented BW has a finite state and is observable. It is observable from the learning agent point of view and also as human interaction partners.

Using the EC, the truth of statements about which block is where is time-dependant and dynamic, that is, these statements are fluents. E.g., HoldsAt( on( block_C,block_B ), 25) which denotes that block_C lies on block_B at time 25. It allows to quantify over the blocks and the time steps:

∃block.HoldsAt(on( block, table), 12)∃block.HoldsAt( on( block, table), 12) :

Signifies that there is at least one block on the table at time step 12.

Happens(stack(blockA,blockB),T+1):−Happens(stack(blockB,table),T):

specifies that if blockB has been put on the table at time T (a universally quantified variable), blockA needs to be stacked on top of blockB at the following time step.

Events are not restricted to action events, but in our framework, the only types of events are actions performed by the learning agent in the object domain (such as stack(blockB,table)), speech acts addressed to its interaction partner(s), and speech acts performed by interaction partners.Actions in the object domains are precisely those events which transform the current Markov state (informally, a set of fluents from the object domain, such as {on(blockA,table),on(blockB,blockA)}) into another state during the search for the semantics of some concept. Speech acts also cause state transitions, but we will later see how to separate “object domain” states (e.g., blocks world configurations) from “interaction domain” states.

5)Explain the difference between a system that uses “dead reckoning” and one that is a “reactive” system.

DEAD RECKONING

On the basis of its rigid-body kinematic equations, a dead reckoning framework for a wheeled mobile robot was developed and the method for robot pose estimation in 3D environments was introduced. After studying the mobile robot locomotion architecture and the concept of proprioceptive sensors, the mobile robot kinematics model was designed to recognize comparative specificity.In view of the fact that the research on the dead count of the mobile robot was limited to 2-dimensional planes, the locomotion of the mobile robot in the 3-axis direction was considered in order to estimate its location on uneven terrain. Since the plane computing method is fairly mature, the height direction measurement is emphatically portrayed as a key problem..With the results generated by the simulation programme and the machine device, the position of the mobile robot can be reliably estimated as well as the optimisation accuracy can be effectively improved, thus displaying the effectiveness of this dead-accounting system..

REACTIVE

In artificial intelligence, reactive planning denotes a group of techniques for action selection by autonomous agents. … Although the term reactive planning goes back to at least 1988, the term “reactive” has now become a pejorative used as an antonym for proactive.

6)Explain why graph searching can be a useful approach to problem solving. Include an example.

Quest strategies are standard problem-solving techniques in Artificial Intelligence.

These search techniques or algorithms have often been used by logical agents or problem-solving agents in AI to solve a particular problem and provide the best answer.

The target-based agents are problem-solving agents that use atomic representation. We can learn different problem-solving search algorithms in this area.

PROPERTIES OF SEARCH ALGORITHMS:

Following are the four essential properties of search algorithms to compare the efficiency of these algorithms:

Completeness: A optimization technique is said to be complete if it promises that if there is at least one problem for every perused, it will return a solution.

Optimality: If a solution found for an algorithm is guaranteed to be the best solution (lowest path cost) among all other solutions, then such a solution for is said to be an optimal solution.

Complexity of time: Complexity of time is a calculation

of time for an algorithm to finish its job.

Complexity of space: It is the ultimate storage space needed

at some point during the search, when the problem is complex.

The different types of uninformed search algorithms below are:

Breadth-first Search Search

Depth-first Quest for Depth-first

Check Depth-limited

Deepening Iterative

Deepening-First Search

Uniform Quest for Costs

Bidirectional Quest Quest for

7) AI researchers often use knowledge and belief interchangeably even though they are technically different.

Explain the distinction between them, and, when they might be used ignoring the distinction.
Then, explain how you (personally) would apply the two terms within an AI system that you were developing?

a) Philosophers have defined knowledge as true, justified belief. The terms knowledge and conviction appear to be used more interchangeably by AI scholars. Awareness appears to mean general knowledge that is taken to be valid. Belief appears to mean facts founded on recent knowledge that can be updated. Beliefs also come with indications of how much they can be trusted and models of how the beliefs communicate. Information is usually not inherently valid in an AI environment and is justified only as useful. When one module of an agent may treat such information as real, this difference sometimes becomes fuzzy, yet another module will be required to revise the information.
b) Knowledge tends to mean general information that is taken to be true. Belief tends to mean information that can be revised based on new information.AI researchers tend to use the terms knowledge and belief more interchangeably. Awareness is the data regarding a topic that is used in that topic to address problems. General information that can be extended to individual circumstances can require knowledge. It is, however, more universal than the values of a given state. A information-based system is a system that uses a domain’s knowledge to work or solve problems.

Fig 1. Knowledge base and inference engine

Source:- Artificial Intelligence: Foundations of Computational Agents, second edition, Cambridge University Press 2017,

8) Explain what makes an intelligent agent and how such an agent can be utilized to develop an intelligent system. Include an example.

An agent can be anything that perceive its environment through sensors and act upon that environment through actuators. An Agent runs in the cycle of perceiving, thinking, and acting. An agent can be:

Human-Agent: A human agent has eyes, ears, and other organs which work for sensors and hand, legs, vocal tract work for actuators.

Robotic Agent: A robotic agent can have cameras, infrared range finder, NLP for sensors and various motors for actuators.

Software Agent: Software agent can have keystrokes, file contents as sensory input and act on those inputs and display output on the screen.

Hence the world around us is full of agents such as thermostat, cellphone, camera, and even we are also agents.

Before moving forward, we should first know about sensors, effectors, and actuators.

Sensor: Sensor is a device which detects the change in the environment and sends the information to other electronic devices. An agent observes its environment through sensors.

Actuators: Actuators are the component of machines that converts energy into motion. The actuators are only responsible for moving and controlling a system. An actuator can be an electric motor, gears, rails, etc.

Effectors: Effectors are the devices which affect the environment. Effectors can be legs, wheels, arms, fingers, wings, fins, and display screen.

INTELLIGENT AGENTS:

An intelligent agent is an autonomous entity which acts upon an environment using sensors and actuators for achieving goals. An intelligent agent may learn from the environment to achieve their goals. A thermostat is an example of an intelligent agent.

Following are the main four rules for an AI agent:

Rule 1: An AI agent must have the ability to perceive the environment.

Rule 2: The observation must be used to make decisions.

Rule 3: An action can follow from a decision.

Rule 4: A fair decision must be the action taken by an AI agent

An agent software typically scans all or any part of the internet, collects information that the user isinterested in, and delivers it to them on a periodic or requested basis, using criteria given by the user.Intelligent data agents may retrieve any stated information,such as keywords used or the date of publication.

User information is obtained using sensors, such as microphones or cameras, for agents who use artificial intelligence ( AI), and agent feedback is transmitted by actuators, such as speakers or displays. Push technology is called the process of getting knowledge brought to a customer by an employee.

Popular aspects of intelligent agents include experience-based adaptation, problem solving in real time, error or progress rate measurement, and the use of memory-based storage and retrieval.

For companies, smart agents can be used for data mining, data management and customer care and assistance (CSS) applications.Consumers may also use smart agents to compare the pricing ofrelated items and, when a website upgrade happens, alert the customer.Intelligent agents, but is intelligent computersystems, are often similar to software agents.

INTELLIGENTAGENTS EXAMPLES

Examples of intelligent agents are AI assistants, like siri and alexa, since they use sensors to interpret arequest made by the user and immediately gather data from the internet without the help of the user.They can be used to capture data about their perceived climate, such as weather and time. Anotherexample of an intelligent agent is Infogate, which alerts users to news based on particular topics of interest. Intelligent agents may also be called autonomous vehicles as they use sensors, GPSand cameras to make reactive decisions to navigate through traffic depending on the surroundings.

9)Describe a task that can be solved using constraints.

Distinguish between the hard constraints and the soft constraints of the system.
Develop a model of the constraint based solution using a table or a diagram.

Consider a Sudoku game with some numbers filled initially in some squares. You are expected to fill the empty squares with numbers ranging from 1 to 9 in such a way that no row, column or a block has a number repeating itself. This is a very basic constraint satisfaction problem. You are supposed to solve a problem keeping in mind some constraints. The remaining squares that are to be filled are known as variables, and the range of numbers (1-9) that can fill them is known as a domain. Variables take on values from the domain. The conditions governing how a variable will choose its domain are known as constraints.

A constraint satisfaction problem (CSP) is a problem that requires its solution within some limitations or conditions also known as constraints. It consists of the following:

A finite set of variables which stores the solution (V = {V1, V2, V3,….., Vn})

A set of discrete values known as domain from which the solution is picked (D = {D1, D2, D3,…..,Dn})

A finite set of constraints (C = {C1, C2, C3,……, Cn})

Please note, that the elements in the domain can be both continuous and discrete but in AI, we generally only deal with discrete values.

Also, note that all these sets should be finite except for the domain set. Each variable in the variable set can have different domains. For example, consider the Sudoku problem again. Suppose that a row, column and block already have 3, 5 and 7 filled in. Then the domain for all the variables in that row, column and block will be {1, 2, 4, 6, 8, 9}.

10) Develop a planning problem to achieve some goal.

As there are different kinds of goals, be sure to state what kind of goal you are satisfying.
Provide at least5 states, and
Provide at least5 actions
1. feel free to have more state and actions as necessary to satisfy your goal).
Explain the planner approach that you would use; and
provide a search space diagram to describe it.

Answer :- Artificial Intelligence is a critical technology in the future. Whether it is intelligent robots or self-driving cars or smart cities, they will all use different aspects of Artificial Intelligence!!! But planning is very important for designing some such AI project. So much so that Planning is a critical part of Artificial Intelligence which deals with the actions and domains of a particular problem. The reasoning side of acting is known as preparation.

All we humans do is with a specific purpose in mind, and all our actions are directed towards our purpose. Planning is also performed with Artificial Intelligence in a similar way. Reaching a precise destination, for example, involves preparation. It is not the only necessity in planning to find the right way, but the activities to be performed at a given time and why they are performed are still very significant.

It’s also why scheduling is known to be the rationale for acting. In other words, preparation is all about agreeing on the actions to be carried out by the Artificial Intelligence System and the running of the system on its own in domain-independent contexts.

Part a)

The goals of AI is learning, listening, perception.

Part b)

Forward State Space Planning (FSSP) and Backward State Space Planning (BSSP) at the basic level.

Part c)

Types of AI Actions

Simple Reflex Actions.

Model-based reflex actions

Goal-based actions

Utility-based actions.

Learning actions

Part d)

Planning, by definition, is to “devise detailed methods for doing, arranging and making something”. For different things different approaches should be adopted. For instance, planning an engineering structure such as a bridge is very different from planning a watershed complex. Some useful approaches employed in watershed planning are explained as follows: 1. Bottom-up approach 2. Iterative approach 3. Flexible approach

Part e)

Fig 3.Space search

Source:- https://www.computing.dcu.ie/~humphrys/Notes/AI/statespace.htm

BONUS: Does (or when does) admissibility imply mono-tonicity of a heuristic?

The admissibility property of a heuristic means that the cost to reach the goal is never overestimated (i.e. it’s optimistic). Admissibility: A search algorithm is admissible if it is guaranteed to find a minimal path to a solution whenever such a solution exists.

Monotonicity: This property asks if an algorithm is locally admissible—that is, it always underestimates the cost between any two states in the search space.

Discuss the merits and faults of Turing’s criteria for computer software being “intelligent“.

Describe your own criteria for computer software to be considered “intelligent.”