Gen Z is debating whether college is still worth it—and it’s a conversation even GV’s CEO David Krane has had at the dinner table.

Krane’s oldest son, who’s halfway through college, is among the Gen Zers questioning the value of a college degree, the executive told Fortune during the Brainstorm Tech conference in Park City, Utah, last week.

“He’s at a liberal arts university on the East Coast in the United States, and he’s asking himself real questions—even bluntly: ‘Is this education a scam?’” Krane told the audience.

Krane, who was employee number 84 at Google, explained that his son spent the entire summer working in AI with large language models and coding assistants.

“He spent the last month of his summer spending a lot of time with peers at Ivy League schools and other notable undergraduate programs who are wrestling with exactly that tension,” Krane said.

Krane said Gen Zers are asking themselves: “Is this education worth it? Do I want to take on additional debt? Can I go build now? Because there’s certainly people that will back me and give me a little bit of capital to begin to figure those questions out.”

The value of a college degree

Hiring professionals are asking similar questions.

Only 41% of junior U.S. professionals say a college degree is necessary for career success, according to a new LinkedIn Workforce Confidence survey. Director-level professionals skew higher, with 47% saying degrees are essential. The survey spans fields including real estate, financial services, and education, saying a skills-first hiring approach would, on average, broaden talent pools in the U.S. by nearly 16 times.

Gen Zers looking for work in tech, like Krane’s son, could see higher return on investing in skills than the average. The tech and media industry, for example, stands to gain more from a skills-first hiring approach, with potential to broaden the talent pool by 24 times, according to the survey.

However, the computer science field for recent graduates has been significantly impacted by AI, resulting in fewer job opportunities and even making it challenging for high-skilled job seekers to secure an interview.

Experts tell Fortune that as companies adopt AI, workers who utilize the technology daily often don’t have an education in coding.

“Based on the data we see helping 10,000 companies like Boeing, Adobe, and Pfizer use AI internally to build software, the most successful employees using AI are actually not necessarily coders or people with CS degrees,” David Hsu, CEO and founder of Retool, a business software builder, told Fortune. “Instead, they are just increasingly people in roles like sales, operations, and finance who learn to use the powerful AI-powered coding tools that exist today to create powerful apps.”

Degrees are evolving

Some experts argue that education is even more paramount in the age of AI.

“Anyone can prompt an AI tool to generate a block of code, but knowing whether that code is efficient, secure, and ethical requires human insight,” Dana Stephenson, CEO at Riipen, a platform that connects students with experiential learning through real work projects from industry partners, told Fortune. “A student might prompt an AI to build a simple app, but without training in software architecture, debugging, and responsible data use, they risk creating something fragile or even harmful.”

Anant Agarwal, Chief Academic Officer at 2U, an educational technology company that contracts with nonprofit colleges and universities to develop online programs, told Fortune that as AI develops rapidly, the skills it demands are appearing in jobs across industries more than ever.

“In this environment, learning deeply and building real expertise is more important than ever because the AI roles and applications are in the context of these other fields,” Agarwal, who is also an electrical engineering and computer science professor at MIT, said. “Degrees also future-proof your career by preparing you for the next big technology, whatever it might be.”

When it comes to coding LLMs or building agentic AI applications, degrees are still needed to provide a rigorous foundation in math and computer science necessary to truly understand how the systems work, Agarwal added.

“At the same time, the four-year degree is evolving,” Agarwal said. “People are increasingly combining certifications and online programs to create flexible, personalized learning paths. Degrees still matter, but how we earn them is becoming more adaptable, modular, and directly tied to the skills the market actually wants.”