ARTIFICIAL INTELLIGENCE
Overview
Artificial Intelligence (AI) has a significant potential to contribute to global economic activity. AI is expected to create 13 trillion dollars annually by the end of 2030 and boost global gross domestic product (GDP) by about 1.2 percent a year (McKinsey Global institute). AI, as we know it today is not new, it has been around since 1970. However, it did not show any sign of success before the 2000s.
As AI becomes more and more widely used, the real leaders of the next revolution in AI will be the enlightened group who understands business and, at the same time, knows enough technology to develop a consistent and achievable AI vision for their organization.
Figure 1.1 | The AI leader is a person who combines domain-specific expertise and AI knowledge to solve industry-specific challenges.
Data is an essential element of an impactful AI. Before the surge in the usage of computers, all the recorded data was in the form of physical documents and files. The internet has made collecting data much easier.
Another important ingredient for AI is computing power. Learning from data does not happen by itself. We need fast processors to have a functional AI module.
To give you a sense of how different the situation is in comparison to the early 2000s, we look at Alex Waibel’s work. He is one of the pioneers of AI in speech recognition and among the first employees of Facebook’s AI department. He was asked how machine learning (ML) work has changed in the last 20 years. He responded that the most powerful computers they were using in the early 2000s were as big as an apartment, cost a few million, and he needed to rent them to develop AI. Today, he has much more computing power on his personal computer for a few thousand dollars. Even your phone is probably more capable than top computers researchers used 20 years ago.
The abundance of easy-to-access data and cheap computing power has created the perfect environment for ML to bloom. Indeed, many (most) applications of AI today depend heavily on ML, such as the Siri voice virtual assistant, Google Translate, self-driving cars, and many more.
Almost all the progress we are seeing in AI today are modules that are only capable of one task. Artificial narrow intelligence (ANI) refers to AI modules and applications that do one thing, such as a smart speaker, a self-driving car, or a web search. These types of AI can only perform one specific task, but they are incredibly valuable when you find the appropriate task.
However, AI also is referred to as Artificial General Intelligence (AGI), which is the ultimate goal of AI. The goal of AGI is for an AI to be able to learn and do anything a human can, or even reach superintelligence and do even more things than any human can.
ChatGPT represents a major advance in self-learning AI, a step toward AGI. It is the next big thing in AI, attracting more than a hundred million users. However this large language model (LLM) only gives the illusion of understanding when it provides an answer to a question. It simply manipulates words and symbols.
Today we can see the significant process in ANI and almost no progress to AGI. Both types of AI are worth analyzing, and unfortunately, the rapid progress in ANI, has caused people to think that there is a lot of progress in AGI, which is leading to some irrational fears that AI can take over humanity, which is not the case. Nevertheless, there are concerns over the lack of regulation and ensuring the responsible use of AI, which will be covered in this short course.
For clarity, in this course when we mention AI, we are referring to ANI (Artificial Narrow Intelligence), which will be the focus.
Learning Outcomes
By completing the lessons in this micro credential, you will be able to:
- Understand what Artificial Intelligence (AI) and Machine Learning (ML) is and can do.
- Describe different applications of AI for non-profit organizations.
- Discuss and analyze case studies that show how AI can be used to build applications that bring considerable value to organizations, government, and communities.
- Engage AI technology responsibly, understanding the risks of AI and the need for adequate safety measures.
Instructor
Farnaz Tehranchi is an assistant professor at the School of Engineering Design and Innovation (SEDI). She received her Ph.D. degree in Computer Science and Engineering from Penn State in 2020. After a post-doc at CMU, she joined Penn State in the Fall of 2021. Dr. Tehranchi’s research interests are in artificial intelligence and computational cognitive modeling, based methods for developing predictive, explorable, and exploitable models. Tehranchi also has an affiliation with the Industrial and Manufacturing Engineering Department and the Center for Social Data Analytics (C-SoDA) to support research and education campus-wide in Social Data Analytics. Tehranchi manages the Human-Centered Artificial Intelligence research lab (HCAI). Her research group has particularly been focused on computational cognitive models (human-like AI models, simulated eyes and hands models), human-computer interaction, data-driven decision-making, and machine learning. Her thesis work won an award at MathPsych/ICCM in 2019.
Graduate Student Support
Amirreza Bagherzadehkhorasani is an industrial engineering Ph.D. student at Penn State. He earned his Bachelor’s in Iran University of Science and Technology. Amir has worked in different areas of differential games, optimization, and deep learning applications. He joined the HCAI lab in Spring 2021. He is currently working on how humans learn and how they make decisions. In his free time, he enjoys playing Tennis and doing Jiu-Jitsu. He also enjoys listening to Joe Rogen and Lax Freedman’s podcasts. Especially the episodes with Elon Musk!