Computer science programs across the state are transforming in the age of artificial intelligence, offering more AI courses and minors, using more of the tools in their curriculum and soul searching how they teach students when answers to increasingly complex problems are keystrokes away.
“We are … trying to adapt to meet the needs of the industry and also leverage our strengths as well,” said Sam Xin, an Iowa State professor and chair of the computer science department. “This is a transition time, and we hope to accommodate the needs of the student and also meet the trend of the industry in the future.”
At the University of Iowa, the computer science department has been around for 60 years, said Alberto Segre, a professor and chair of UI’s department of computer science.
“A lot of things have changed, and so we undergo these upheavals periodically,” he said.
Some of those upheavals happened when the internet came around, he said. Network computing became a core issue, and there was the same issue with cloud computing. Now, AI can write code, raising concern among students about the future of entry-level jobs in the field.
“Every time new languages come out, we see the same thing in programming languages,” he said. “So this is a field that’s used to change.”
The impact of AI has universities monitoring enrollment and job prospects, with some academic leaders expecting enrollment decline and tougher job prospects. Meanwhile, computer science departments are balancing industry demand for AI skills with the need to prevent cheating.
Enrollment outlook dims
Since 2007, the computer science department at Buena Vista University has more than doubled in size, said Jason Shepherd, professor of computer science at BVU. That trend is not expected to continue.
“I actually do expect us to decline a little bit, and part of the reason for that is there is a perception, both amongst parents and high school students but also among employers, that AI will improve the efficiency of workers so that you don’t need as many humans to do the work anymore,” Shepherd said. “That’s just something that I think is a commonly held belief. I don’t know that I completely agree with that. But perception is often reality, so although we haven’t seen the decline yet, we’re expecting it at some point.”
Eugene Wallingford, head of the department of computer science at University of Northern Iowa, said its numbers are down, “but just a little bit.”
“Parents in particular, are concerned about what they’re reading in the press, about how AI, other things might affect employment,” he said. “And so we talk about that. But our student enrollment numbers … are still pretty good for us.”
In addition to the computer science degree at Iowa, there are three other undergraduate programs related to computer science: informatics, which is an applied computing degree, data science and a degree in computer science and engineering, Segre said.
“The enrollments are down a little bit,” Segre said. “Overall enrollments at the university, we had a big incoming class this year, but everybody is sort of bracing for the decline in enrollments. That’s due to demographic effects, which we probably will see fully realized in the next year or two, as the number of eligible 18-year-olds starts to drop, but our enrollments were really overwhelmingly strong about a year and a half ago.”
A year and a half ago, the University of Iowa peaked at over 1,000 students in the combined four programs, Segre said, compared to 2006, when it was 250 students.
“We quadrupled in number,” he said. “I think there has been a huge increase in interest in computer science over the last 10 years, largely in part because it’s an exciting field. It’s a wonderful field to be in, but it’s also because the salaries were really quite strong.”
Computer science graduates were starting at over $100,000 a year, when other majors were earning $60,000, he said.
“That drew a lot of people in,” Segre said. “It didn’t always draw people in for the right reason. People who are in the degree program just to get a paycheck don’t always make the best students in that degree program. … People are seduced by technology. … Now they’re worried, there’s this sense that AI is going to take over those entry-level jobs in [computer science], and so that’s led to a decline in interest, although our enrollments are still pretty strong.”
He said he recently did a census of computer science enrollment and found they have about 800 students across the four programs.
“So we had a reasonable-size drop, but the biggest drop was … in the engineering program, which was a bit of a surprise to us,” he said. “My sense from my Big Ten peers is that everybody saw a small decline, but it wasn’t an overwhelming decline. And in fact, it’s a welcome decline, because we were all kind of struggling to meet that demand, and it’s a little bit more comfortable now.”
Job outlook ‘tougher than we’ve seen it’
Wallingford said the hiring outlook for computer science graduates has “been a lot tighter” over the last year and a half, with it taking longer for students to find jobs.
“In the popular press, this gets blamed on AI a lot, but I think that there are a lot of different things going on,” he said. “So certainly, the rise of these tools, which makes other developers more productive, which might [mean] fewer developers.”
But there are other factors at play too. Big companies like Google, Amazon and Microsoft have moved toward working in the office, so there are not as many remote workers. They’ve also had layoffs.
“I think what that’s done is to increase supply out in the world so there’s more people available to hire who have been … looking for jobs,” Wallingford said. “And while those are mostly on the coast, some of those people [are] remote [and] live in different parts of the country. I think that’s affected hiring across the U.S.”
Shepherd, at Buena Vista, said the job market for computer science graduates is “tougher than I’ve seen it.”
“It used to be that a lot of our students would have jobs – when they are seniors, they would be able to find employment pretty early in their senior year, and everybody would just have a job waiting for them after college,” he said. “That has become a lot more challenging. They’re having to apply to a lot more positions. They might have to be more flexible in terms of geography, where they end up.”
Shepherd said part of the issue could be that the job market is challenging in general right now.
“I don’t actually know whether it’s any worse for people in computer science or not, but I will say that anecdotally, we noticed that students are having to put forth more effort into finding a job than they did in the past,” he said.
Infusing AI in curriculum
Colleges are in the midst of making changes — and some already have — to their computer science curriculums to account for AI.
Iowa, for example, is adding a required class in AI for all computer science majors, along with a certain number of electives that touch on AI, Segre said. There will be a newer emphasis on machine learning, more emphasis on teaching about AI earlier in a student’s academic journey, he said.
There’s also a new ethics class requirement.
“AI has become a bigger part of our curriculum and will be enshrined in the new curriculum design,” he said. “I think a lot of departments in the U.S. are undergoing a similar process.”
As AI has become more popular in the last decade, computer science programs have gradually adapted to incorporate more of it in their courses and offerings, Iowa State’s Xin said.
“Many new courses in computer science programs have been developed to meet the needs of the student and also meet the needs of the industry,” Xin said.
He said AI-related courses are popular among computer science students as well as those in other majors.
“It’s not just [attracting] students from computing fields, actually lots of students from, for instance, engineering, and even from other majors as well,” Xin said. “They are all interested and took those AI courses over here at Iowa State.”
ISU offers multiple AI minors, Xin said.
“We also [are] trying to develop a graduate certificate, probably called responsible AI, so you can see that academic programs, including certificate minors, have been really growing fast for AI programs,” he said.
In 2014, Buena Vista decided to include different interest tracks in the computer science major. One was AI and robotics, Shepherd said.
“It gives students some additional coursework, and in some cases, some co-curricular activities to help them dig a lot deeper into AI and robotics,” he said. “We do a lot of really interesting data manipulation and prediction using artificial intelligence and machine learning, but our students also learn how to apply that directly to robotics.”
Buena Vista also has a robotics club that builds autonomous robots and has competed in contests across the Midwest, Shepherd said.
Drake University has an AI major that is in its sixth year. A lot of students that major in AI are also majoring in computer science, said Chris Porter, director of Drake’s Artificial Intelligence Program and co-chair of the department of math and computer science.
“But only in the last three years [have] the AI tools emerged that can actually help out with these things and do that work, and that’s forced us to make some adjustments,” Porter said. “So we’re seeing upper-level courses in software engineering starting to introduce agentic coding.”
Among other adjustments: Professors came up with a new grading template for students who decided to use ChatGPT to help with coding.
“You need to say that and explain what’s going on, like add an additional layer of responsibility to the assignment,” Porter said.
He said Drake also regularly evaluates the major for needed structural changes.
“Right now, Drake has [a] bachelor of science and a bachelor of arts in computer science,” he said. “We’re changing the major so that the bachelor of science is a bit more technical. … We’re adding more math requirements, adding some more theoretical courses, adding courses that have some more hardware, systems components to them, because that stuff’s harder to automate. So the [bachelor of science], I think, would make for [a] computer science degree that’s more robot-proof.”
Professors are also having to adapt teaching methods to fend off cheating.
Despite honor codes and consequences of zero credit or being reported to the college, Segre said cheating is common and AI makes it easier.
“What you have to do is you have to change the way you evaluate students,” Segre said. He said instructors are now giving oral exams, which has been the European tradition for centuries. Students come in one at a time and are asked a question that they work out on a blackboard.
“They can’t really use AI to do the problem for them,” Segre said.
Other professors resort to pen and paper or paper and pencil testing under control conditions where no electronics are allowed, “which we do in our larger courses,” Segre said.
Wallingford said GitHub, an open-source platform where anyone – large companies, individual programmers – can share the software they create, has facilitated cheating.
“Our students have been able to go download programs or libraries of code that do something very similar to or exactly the same thing that they’re trying to accomplish, which is a challenge for us,” he said.
It’s difficult to write a programming assignment or task for a team of student programmers that an AI tool can’t generate code and “finish the task for them,” Wallingford said.
“That’s been the new thing for us,” he said. “The challenge is we want students to be able to use those tools effectively, since it seems clear that they will become a part of their professional practice in the future. A lot of companies are using them already, and probably won’t change that in the future, but we want students to be able to, after you’ve used a tool to generate software, start to be able to read the software … and test the software to make sure that it’s actually doing what they want it to do.”
The challenge for educators is to ensure they can teach students basic skills to write code and read it and understand it and then be able to use that knowledge in conjunction with the tools that generate code quickly for them, he said.
“We have upper division courses right now that allow students to use these tools under selective guidelines,” he said.
‘You still need humans in the loop’
With generative AI, tasks that would have taken Shepherd weeks to do might take a day or two, he said.
“It stands to make individual people more efficient and more productive,” he said. “However, you still have to have humans in the loop, because generative AI makes mistakes. It’s really giving you a really good guess at what it thinks you need.”
For example, a student who has never programmed before may ask ChatGPT to create a software application, then they’ll get that code, he said.
“They have no way to know whether that actually is going to work or not,” he said. “And so there still have to be humans in the workflow, in the pipeline, to take that code and then adjust it to make it do what actually needs to be done.”
Shepherd said he believes there will be a reduced need in the industry for people who program.
“If you have a lot of programmers that then become way more efficient, you don’t need as many of them,” he said. “I think that’s generally true. But in talking with folks in industry, I think the degree to which they think they’re going to be able to shrink their workforce is very overestimated. I’ve heard of companies talking that they want to reduce their programming staff down to like 10% or even half of what it is today and that doesn’t sound right to me.”
Segre said although there may be huge enthusiasm for AI, it’s not clear yet how large AI’s impact will be on the economy.
“It’s probably going to be big, but nobody really knows how big, and it could still be a dud,” he said. “Nobody really knows, so we’re in this funny period where people are unsure of how to proceed and what’s going to happen, and I think that’s also feeding some reluctance to over-hire at the junior level. But there’s still a lot of good jobs out there. There’s just not the $200,000 starting salaries for bachelor’s degrees.”
Nicholas Haisler, who graduated from Drake in December opted for a graduate school instead of looking for a job. He will study machine learning in Japan beginning in March.
Haisler said he wants to learn more about AI. His major has become more valuable as tools like ChatGPT have become more popular, he said.
“I think my understanding of AI and where it’s applicable has broadened a lot. … If I were to join a company right now without the AI major, I think I would struggle to see where I can actually use it and where it’s valuable and help guide the company correctly,” he said.
