Brain Drain of AI Researchers: Academia vs Industry

Onikle Inc.
Published in
5 min readMar 22, 2021


Image by Atsutaka Odaira

Artificial-intelligence (AI) researchers at all stages of their academic careers are drawn to industry by higher wages and benefits such as access to large data sets and computing resources. Businesses are looking for AI experts to work on projects ranging from modelling risk in finance to developing crop-harvesting robots. On the other hand, Universities have long been a source of gifted leaders for industries, but a mass departure of academic researchers with expertise in artificial intelligence has raised important questions for the future. This blog will look into why this might be happening.

According to a report released in 2018 by the Canadian software provider Element AI, the number of PhD graduates on LinkedIn who claim to have AI expertise has increased by 66% in that year. The report’s authors discovered, based on a sample of 4,500 researchers, that the United States remains a major hub of AI training and jobs. China, the United Kingdom, and Germany are among the other hotspots. However, according to Yoan Mantha, market intelligence lead at Element AI, demand is outstripping supply. According to the industry, there are approximately 144,000 AI-related job openings in the United States, but only about 26,000 developers and professionals are looking for work. According to a Deloitte 2020 survey on the state of AI in business, all enterprises polled noted an AI talent gap. AI developers, programmers, journalists, and data scientists were the most in demand.

Craig Wills, a computer scientist at Worcester Polytechnic Institute in Massachusetts, conducted a 2018 study in which he examined ads for computer-science faculty positions with start dates in 2019 at 409 institutions, mostly in the United States. According to him, the proportion of advertisements seeking experts in AI, data mining, and machine learning has nearly doubled since 2015. Wills discovered that 42 percent of searches for computer-science faculty members by top US graduate schools did not result in the desired number of new hires in a study of 176 institutions released earlier that year.

Anecdote from a professor at Imperial College London illustrates the magnitude of the issue. The professor was perplexed by the absence of one of her students as the new academic year began at the university. He’d been working in her lab for three years and only needed one more to finish his research. However, he had stopped coming in. Later on, the professor found out that he had left for a six-figure salary at Apple. She says, “He was offered such a huge amount of money that he simply stopped everything and left” and “It’s five times the salary I can offer. It’s unbelievable. We cannot compete.”

This is not a one-time occurrence. Across the world, talented computer scientists are being enticed away from academia by attractive private-sector offers. According to a Guardian poll of Britain’s top research universities, tech companies are rapidly recruiting AI professionals, fueling a brain drain that has already impacted research and teaching. One university executive expressed concern about a “missing generation” of scholars who would usually educate students and be the driving force behind research projects.

The reason for this phenomenon can be explained with basic economic. The demand for experts has outpaced supply, causing prices to skyrocket. As a result, sector incentive packages for top researchers are extremely high. Furthermore, supply has proved to be somewhat inelastic; being a leading AI specialist requires years of hard work and outstanding skill. Experts are also drawn to industry by diverse data sets, abundant computing resources, and the opportunity to affect millions, if not billions, of people through commercial products.

The challenge for academia comes when professors leave to work for corporations, either in large numbers, such as when Uber recruited 40 people from a Carnegie Mellon robotics lab, or separately, as with notable AI researchers Daphne Koller, Andrew Moore, Andrew Ng, Fernando Pereira, Sebastian Thrun, and others. Facebook has vigorously recruiting professors from universities like New York University, Carnegie Mellon University, and the University of California, Berkeley in recent years.

The consequences of brain drain may extend far beyond academia. The majority of top AI researchers moved to a few corporations, implying that their talents and knowledge were not shared with the rest of society. This is a problem because only the diffusion of innovation, rather than its concentration in a few companies, can mitigate the dramatic disruptions and negative effects that AI may cause.

So, you might think, what is in it for AI researchers in academia. First, scholarly research places a greater emphasis on novelty, while business research places a greater emphasis on utility. In machine learning, this is sometimes combined with academia working on rapid prototyping for smaller datasets, while business research spends more in making things work. The second major distinction between academia and industry is that individual vs teamwork. In academia, the funding structure (and, to a lesser extent, the credit structure) often discourages cooperation among multiple faculty-level researchers. Typically, people in industry, where people are willing to collaborate. This enables the incorporation of more people’s knowledge and the development of more sophisticated systems.

Despite what appears to be a conflict of interest between academia and business, recent developments indicate that the two worlds are moving toward collaborations and partnerships to complement each other. Big Tech Companies such as Google, Facebook, Microsoft, and Amazon are increasing their support for academic research organizations by providing more funds, data, and compute resources.

Now many university professors are now being offered senior researcher positions at industry also allowing them to keep teaching at universities. For example, Facebook allows professors to spend approximately 20% of their time at colleges, which encourages recruitment for Facebook and enables professors to teach classes and communicate with students. This formula can be seen with companies too.

Although many people prefer industry to academia, many, if not most, notable AI accomplishments have their origins in years of academic research. Clearly, academia requires new blood to keep the wheel of innovation turning.At this rate there would be not enough AI researchers for both industry and academia.

If you are interested in our blogs, please register your email address in the following link to get notifications through our newsletter:

Written by Wanonno Iqtyider



Onikle Inc.

Parent compant of NapAnt. NapAnt is a service that helps improve development efficiency, man-hour estimation accuracy, and unleashes the potential of your team.