Shortage of AI Faculty Exist and The Reason is Very Complicated

Uber wanted to set up a new Pittsburgh facility for self-driving cars. In 2015, the company approached robotics scientists and researchers at nearby Carnegie Mellon University. Soon after, lured 40 workers from the center with bonuses and doubled salaries. These high-profile stories have helped to reinforce the belief that artificial intelligence specialists are leaving academia for industry. According to the report this month, by the Center for Security and Emerging Technology, the lack of AI professors in U.S. universities does not reflect a depressed job market. AI experts are still interested in academic careers. However, university hiring for AI faculty has not kept up with student demand. Although big tech has helped to fill the gap, experts warn against being too cautious as the incentive structure of the industry is different from that of academia.

Academics have always produced a steady stream of entrepreneurs, developers, and engineers that have fueled an AI innovation environment. This stream has been directly linked to AI faculty teaching capacities. While student enrollment in computer sciences programs has increased dramatically over the past decade, universities still haven’t hired enough computer science teachers to meet this demand. (The researchers used student interest in computer science as a proxy to student demand for AI since the latter is difficultly quantifiable.

According to Jack Corrigan (one of the report’s authors), universities have tightened admission requirements for computer science programs in an apparent response to student demand. While a growing number of computer science Ph.D. holders have expressed an interest in academic careers in the meantime, universities have not responded to this demand with an increase in faculty positions. Universities are generally successful in hiring AI faculty, contrary to industry poaching.

For their part, technology companies have offered alternative routes to AI education and training to meet student demand.

A Google Health Advisory member and WHO founding member John Nosta said in a statement that giant tech companies are now becoming at the forefront of innovation. He also quoted that Google does not always require to have degrees for their education. It’s not the universities that are generating the excitement, but innovative companies like SpaceX, Amazon, and Apple. These businesses are disrupting traditional education models.

Google isn’t the only company that has removed the requirement for certain positions to have a college degree. IBM and Apple also removed the requirement. This effort may help diversify the talent pool by allowing access to people who have not had as many opportunities in their early years.

A former IBM executive said that there are many AI jobs and a single barrier has to be removed. Instead of a degree, she explained that IBM screens for “propensity to learn” and offers training. It is believed that the tech industry is the pure training playground for AI professionals.

Dan Rockmore, Dartmouth College computer science professor, said, “I’d be reluctant to call the tech companies universities.” They are only interested in a specific set of skills, which are not universal. However, I believe they are creating a new type of technical school.

Rockmore agreed that AI curricula at universities are not always relevant to market needs. However, he cautioned against relying upon tech companies for AI education.

“This will be an extremely narrowly trained group, whose products are going to have extraordinary ramifications on how we interact and where we govern,” he said. They build technocratic solutions’ without any education or a clear and thoughtful outlook on its implications.

Others highlight the historical interplay between higher education, the tech industry, and other factors. Take computer graphics as an example. This is the suggestion of Cherri Pancake, a former president of the Association for Computing Machinery (ACM), and professor of computer science at Oregon State University.

“The vast majority of graphic practitioners today are not specialists.” She also noted that artificial intelligence and machine learning, a highly-demanded subfield of the AI industry, is no different. For decades, academics have worked to make machine learning cost-effective. Pancake stated, “Now, everybody is crying for that specialty but the real need for [machine learning] is for people who are able to apply it in practical settings.”
This requires a new kind of education, one that is focused on safe use. She said that although universities recognize the need for new curricula, it can take time to create them.

Some feel frustrated by the slow pace of academe. Some are frustrated by the slow pace of academe. They noted that more than 7500 first-year University of Washington students applied for admission, but there are likely only 550 new undergraduates.

Jim Hendler, director, Institute for Data, Artificial Intelligence and Computation, Rensselaer Polytechnic Institute, and chair, ACM global technology policy committee, is concerned about the future generation of AI Faculty scientists in the U.S. and elsewhere.

ACM sees this issue not only in the U.S. but all over the globe,” he stated. “Our curriculum committees are examining not only college education but also whether there could be new programs in K-12 and especially high school/pre-college educational programs.

Although Corrigan acknowledges that the industry plays a part in developing AI talent he urges policymakers and academic leaders not to forget about the importance of universities.

The incentive structure of private companies is very different from that of universities. If we are to create a tech workforce that is just, fair, and socially beneficial, we need to consider the incentives driving each actor’s behavior.