Hiring Machine Learning Professors Fast to Catch-up Short-Term Will Make You Fall Behind Long-Term

From current observations, I would like to suggest a new law of hiring professors:

Hiring professors fast to catch-up short-term will make you fall behind long-term.

The reason why I’m writing this are the large amounts of money being made available to German universities to hire new machine learning professors (think “1000 professor program” and the like). This money is being provided by the country of Germany as well as individual states, and it is substantial. The intention is to catch-up to the U.S. and China, who are perceived as leading in machine learning techniques and artificial intelligence.

Sadly, I doubt that these good intentions are likely to have the desired effect, at least not long-term.

The supply for professors specializing in machine learning is small; industry is often more attractive and definitely pays more. This supply is generally, and now thanks to new public funding, far outstripped by the demand for such professors. As a consequence, too many people not that qualified will be hired, just to make sure the funds are pulled down from the government and don’t go to waste. Short-term, this may boost machine learning research and education at universities.

Long-term, it will hurt. Professors in Germany get their job for life with a president of a university having little say over what they could or should do. Ten years from now, the lower quality of professors hired too fast will start having its impact, with lower quality of research and teaching. Because you can’t get rid of under-performing professors, German universities will be stuck with them for the next thirty years.

At the same time, the money dedicated to new machine learning professors will not be available for other topics, hence it will starve computer science research and education more broadly.

Predicting the future is hard and I don’t expect a university president to accept under-performing professors without a fight. Once these consequences take hold, I expect more performance agreements and professor evaluations, something not (yet) common in German universities. Whether this is a good thing is debatable. The German basic law (constitution) guarantees freedom or research and teaching at universities on the same level as free speech. The fight will be on.

One Reply to “Hiring Machine Learning Professors Fast to Catch-up Short-Term Will Make You Fall Behind Long-Term”

  1. There should be more Professors on Quantum computing appointed . That’s the area every IT company is struggling . Every one is making big claims on their Qbits but nobody knows the truth . At least when there is a massive clientele of academia present on the font . Some truth can be deciphered , as it is going to be the benchmark for the technologies AI , cloud computing , block chain for the upcoming years .

Leave a Reply