Computer Science

(From Eagle Drive magazine, December 2023) 
 
Can machines think? Pioneering computer scientist Alan Turing posed this question in 1950, and the concept continues to captivate students and researchers of artificial intelligence and machine learning. 
 
Now AI can answer the question for itself: 
 
“Machines, like computers, can process and analyze data, make decisions based on programmed algorithms and perform complex tasks, but they don’t ‘think’ in the way humans do.” – ChatGPT  
 
ChatGPT, a language model developed by Open AI, also “explained” that machines can simulate human-like behavior and intelligence, but they “lack emotions, self-awareness and true understanding, which are essential aspects of human thinking and consciousness.” The tool has quickly become a common way for people to interact with AI in the 2020s. Users download the application on their smartphone or computer, enter questions or information requests and receive human-like responses. The “GPT” stands for generative pre-trained transformer.   
 
ChatGPT also has uses in academia. At Tennessee Tech, computer science professor Doug Talbert, Ph.D., said it can be a problem-solving tool. For example, when his research students have questions about coding and debugging, he may encourage them to try it.  
 
“Maybe it gives them a starting point on how to think about it,” he said. “They would be googling it anyway… so why not ask ChatGPT?”  
 
As artificial intelligence continues getting “smarter,” faculty and students alike are considering the opportunities, applications and implications.  
 
Department of Computer Science Chair Gerald Gannod, Ph.D., said students face a future where GPT-based software development tools will be the norm.  
 
“As such, we need to provide instruction on how to best employ these tools in everyday software engineering,” he said. “Moreso than any other major, computer scientists have the responsibility to know how to use tools like ChatGPT effectively while still leading the way in creating new technologies that impact our industries and communities.”  
 
One way he addressed the innovation was offering a special topics course in the summer of 2023 called “Software Engineering with ChatGPT” in which students explored benefits and challenges related to using an AI assistant to create software.  

“They were able to gain a deeper understanding of why the concepts they learn in our courses – especially algorithm design, test-driven development and software development – are a necessary foundation for fully understanding how to reap the maximum benefit of using GPT-based software engineering assistants,” he said. 
      
ChatGPT has also caused faculty to consider a crucial question: When is it acceptable to allow students to use generative AI in a course?  
 
“We recognize that the technology has the potential to transform our discipline and what we can expect students and graduates of our program to be able to do both while in school and when they move forward in their careers,” Gannod said. “As such, we have to up our game by modifying how we assess student work. This includes making sure we are testing student knowledge that is beyond the capability of what an AI can produce.”     
 
Talbert and fellow computer science professor William Eberle, Ph.D., agreed that it’s vital to incorporate the tool into computer science education. 
 
“If students don’t know how to use it to be more productive more quickly, they’re going to be at a disadvantage in the marketplace,” Talbert said.  
 
Eberle added, “You can look at it as making life easier. If it can start a process that would have taken longer to figure out – and do it well enough – you can do more things faster.”  

But students are also faced with an ethical challenge – avoiding AI when instructed not to use it.  
 
“Without foundational knowledge of software development, students will be unprepared to step into the workplace,” Gannod said. “A cultural change is needed that emphasizes intrinsic-driven learning (wanting to learn) rather than externally motivated learning (driven by grades).”  

With rapid advances in the field of AI, students are discovering that it’s an exciting time to be a computer science major. For example, Tech’s Association for Computing Machinery student chapter presented an artificial intelligence and machine learning demonstration to inspire fellow students. Brett Billingsley, club president, talked about how AI and ML are used in the real world and offered resources, while Garrett Hayes, vice president, demonstrated a facial recognition program he created.   
 
Students are also gaining AI experience through research opportunities alongside professors like Muhammad Ismail, Ph.D., who serves as principal investigator of numerous grants from the National Science Foundation and Qatar Foundation and leads NSF’s CyberCorps Scholarship for Service grant program at Tech.   
 
Ph.D. candidate Umair Mughal has been working with Ismail on a project to defend drones against cyberattacks.  
 
“Hackers can attack the autopilot system, so we are developing intrusion detection systems using machine learning techniques,” he said. “We launch cyberattacks on our testbed that consist of drones and hacking tools. Then we collect benign and malicious datasets and develop AI-based intrusion detection systems.”  
 
Some of Ismail’s other projects include researching the development of AI-based strategies for intrusion detection in cyber-physical critical infrastructure as well as the operation of 5G and beyond wireless networks.  
 
AI research and learning opportunities will continue to grow beyond Bruner and Prescott halls (which house most computer science classrooms) with Tech’s new MInDS (Machine Intelligence and Data Science) center and Ashraf Islam Engineering Building, the university’s first smart building, which will house the Cybersecurity Education, Research and Outreach Center when complete in 2024.  
    
It’s a lot to think about – even for a machine.