MyDigitalWorldE8U1 split PDF

1 What is Artificial Intelligence? In 2022, a Google engineer was fired for stating that the chatbot he worked on is aware and can sense like a 7 to 8-year-old child. Activity 1 : Discuss with your teacher and classmates: 1 What are your thoughts on the engineer sta...

1 What is Artificial Intelligence? In 2022, a Google engineer was fired for stating that the chatbot he worked on is aware and can sense like a 7 to 8-year-old child. Activity 1 : Discuss with your teacher and classmates: 1 What are your thoughts on the engineer statement? 2 Why do you think an employee can be fired for saying that the computer is sentient, even though we commonly refer to computers as smart? 6 In 1956, two computer scientists, Allen Newell and Herbert Simon, developed a program they named General Problem Solver. The program was designed to mimic problem-solving behavior similar to that of humans, addressing a broad range of problems using the Means-End Analysis Method. One of the main principles of this technique is associating symbols commonly used in our daily lives with their meanings to arrive at solutions. The program successfully tackled logical problems structured similarly to mathematical ones, but it struggled with solving real-world problems beyond this scope. Consequently, it fell short of addressing general problems. Do you believe that a program which associates symbols together can be considered intelligent? 7  Example: In a scenario inspired by John Searle, picture a room with two openings—one for messages, one for instructions. Inside is someone without Chinese skills, armed with a guidebook. Outside, someone sends Chinese questions through the first opening. Despite the language barrier, the person inside answers correctly using the guidebook, then send the responses back out. This demonstrates how one can respond accurately without understanding the language. Q Looking from an external viewpoint, this individual may seem proficient in Chinese, but is this genuinely accurate? 8  This scenario closely resembles interactions with smart gadgets gadgets. But do these gadgets truly understand what they say, or do they merely give the impression of understanding? Activity: 1 Form a team with two of your classmates. 2 Select your favorite search engine. 3 With your team, search for "strong AI" and "weak AI". 4 Collaboratively define both terms with no more than 3 sentences. 5 Share the definitions with your teacher and classmates. Q If programming computers to interpret symbols can make them seem intelligent, can they perform tasks without explicit programming? Can they acquire skills solely through observation? 9  In 1959, computer scientist Arthur Samuel created a game capable of learning and developing winning strategies without explicit programming. Through learning from trial and error experiences, the game succeeded in defeating its creator. Samuel coined this technique as Machine Learning. Q But how could a machine actually learn? 10 2 How Could a Machine Learn? Until 1959, the fundamental principle of computers was to provide them with a set of inputs, followed by clear and specific steps to process that input, ultimately producing an output. Input Process Output But in 1959, Arthur Samuel conceived a different approach to computer programming. He pondered whether a computer could autonomously deduce the logic required to complete a task, rather than being explicitly instructed to do so. To test this hypothesis, Samuel supplied the computer with inputs and outputs from a previously completed task, without providing any explicit instructions. The aim was for the computer to infer how to complete the task. Example: Provide the computer with the following inputs: Input Output 5 7 12 9 6 15 3 7 10 Q Do you think the computer can deduce that the equation to process the inputs is as follows? x1 + x2 = y 11  This science was termed Supervised Learning, which is a subset of a larger field known as Machine Learning. Machine Learning Unsupervised Reinforcement Supervised Learning Learning Learning Supervised Learning: This type of science relies on training the AI with extensive inputs and outputs to deduce the steps necessary to complete the task autonomously. 12  Unsupervised Learning: This form of Machine Learning involves providing the AI with a set of inputs but without corresponding outputs, allowing it to search for patterns or relationships among the provided inputs. Example: Sami and Sara use a video streaming platform that employs an unsupervised learning approach. Upon analyzing Sami's watch history (representing inputs), the unsupervised learning model discovered that most of the videos he reviewed are short and comedic. Conversely, Sara's analysis revealed that she watched videos lasting more than 20 minutes, primarily focusing on technical and science fiction genres. 13  Given that this platform employs a model capable of recommending videos to its users, do you believe the platform's recommendation model will suggest suitable recommendations for both Sami and Sara? Reinforcement Learning: Reinforcement Learning diverges from conventional input- output paradigms. Rather, it entails an agent engaging with an environment to execute actions based on the state of the environment. 14 The agent receives rewards or penalties contingent upon the efficacy of its actions, with rewards serving as points. Through iterative interactions, the agent refines its policy to develop proficient strategies in response to the environment. A notable illustration is AlphaGo. Activity 1: Utilize your search engine to gather information on AlphaGo, developed by DeepMind. NOTE: Every instance of machine learning falls under AI, but not every AI instance involves machine learning. Machine learning is just one part of AI, which is a broader and more extensive field. Activity 2: 1 Form a team with three of your classmates. 2 Choose your favorite search engine.. 3 Conduct a search with your team on AI branches. 4 Collaborate to define AI and its branches. 5 Share the definition with your teacher and classmates. 15 3 Applications Activity 1: Andrew Ng, the computer scientist and business pioneer, and one of the most influential figures in the world of AI, famously stated, "AI is the electricity of the 21st century." What are your thoughts on this statement? Q Engage in a discussion with your teacher and friends. Nearly all domains in our modern lives incorporate AI technology. Let's explore five domains as examples. 16 Medical Care Sector: AI has revolutionized healthcare by enhancing diagnostics, personalized medicine, drug discovery, and robotic surgeries. Using AI algorithms, medical data is carefully examined to detect diseases such as tumors early on. Additionally, AI supports doctors in selecting optimal treatments for each patient by leveraging their health records. 17 Activity 2: Do you believe AI can be leveraged to improve medical care in your city? Financial Sector: In the financial sector, AI is utilized for marketing, trading, fraud detection, and providing customer service. AI systems are adept at analyzing market trends, forecasting stock movements, automating business, and efficiently managing electronic wallets. Additionally, AI strengthens security measures by identifying abnormal transaction patterns, thereby mitigating potential fraud risks. 18  Transportation Sector: AI is a cornerstone of the transportation sector, driving advancements such as autonomous vehicles, including self-driving cars and drones, as well as enhancing other transportation systems. These systems process data from a multitude of sensors and cameras to navigate safely, detect obstacles, and make real-time decisions. AI also enhances public transportation routes and traffic management by optimizing routes for efficiency and enhancing overall traffic flow. Q Do you think your town is ready for Self-driving cars? 19 Activity 3: 1 Consider the transportation system in your city, and discuss with your teacher and classmates how you can enhance it using artificial intelligence. 2 Mention 3 suggestions at least. Trade Sector: AI improves customer experiences in retail and e-commerce by offering personalized recommendations, managing inventory effectively, and deploying chatbots for customer service. Through machine learning algorithms, customer data is analyzed to predict buying patterns and offer tailored product suggestions, leading to increased sales and better engagement with customers. 20 Integrating AI with Internet of Things devices in smart homes: The development of Internet of Things devices has been remarkable, especially when combined with AI technologies. These advancements allow these devices to collect data during their daily operations, analyze it, learn from it, and enhance the services they provide. Activity 4: 1 Have you ever used an Internet of Things device for smart homes? 2 Did this device incorporate AI technologies? 3 Mention three AI devices. 21 3 Generative AI Activity 1: 1 Browse the following link: https://thispersondoesnotexist.com/ 2 Reload the site multiple times and observe how the faces change. 3 Did you recognize any of those faces? In fact, none of these faces belong to real individuals! Instead, they were generated by AI. 4 How do you think AI accomplished this? 22 Thus far, our exploration has revealed the diverse capabilities of computers in handling inputs, including processing, analyzing for insights, and deriving efficient methodologies. But how can a computer autonomously generate data without any prior input? Activity 2: Create fictional faces using AI. 1 Browse the following link: https://this-person-does-not-exist.com/en 2 Select the gender, age, and ethnicity of the person, then click Refresh Image. 3 Did the outcome meet your expectations? 23 Generative AI: Generative AI marks a significant leap in artificial intelligence, empowering machines not only to analyze and comprehend data but also to generate fresh and unique content across various mediums like text, images, music, and more. This form of AI operates by assimilating patterns, styles, and structures from vast datasets, subsequently leveraging this knowledge to produce novel content akin to the learned styles or content. Typically, this is accomplished through machine learning models. 24 Generative AI main features: 1 Data-driven learning: Generative AI learns by studying datasets, enabling it to understand patterns, techniques, and structures within the data. This could be anything from written human language to visual artistic patterns. 2 Content creation: Unlike traditional artificial intelligence, which is mainly used for identifying or categorizing data, generative AI is employed to produce fresh content. This might involve crafting texts in unique styles, generating entirely new images, composing music, or even making deepfake videos. 25 3 Applications: Generative AI spans diverse applications, including art, entertainment, design, literature, and more. For instance, it can generate lifelike images and artworks, compose poems or stories, or even produce models for scientific investigations. 4 Variability and unpredictability: Generative AI models can produce a wide array of outcomes based on their input and training. Sometimes, these results can be surprising and varied. While this can lead to creative and innovative results, it can also occasionally yield unexpected or less desirable outcomes. 5 Ethical Considerations: The utilization of generative AI prompts significant ethical questions, particularly regarding issues of originality, copyright, and the possibility of misuse in generating misleading or harmful content. Types of Generative AI: 1 Text-to-Text Generation: involves producing textual content in response to a text input or request. Check out the video link below to explore various methods for generating textual content from text inputs: 2 Text-to-Image Generation: involves creating images based on textual input or requests. Click the video link below to discover various methods for generating images from text inputs: 26 3 Image-to-Image Generation: involves creating an image based on another image. Click the video link below to explore various methods for generating images from existing images: 4 Text-to-Video Generation: involves creating video content based on textual input or requests. Click the video link below to discover various methods for generating videos from text inputs: 5 Text-to-Voice Generation: involves creating audio content based on textual input or requests. Click the video link below to explore various methods for generating audio clips from text inputs: 27 28