Klamka: PhD graduates increase while academic faculty positions stay flat
Jake Klamka · Jake Klamka, Insight Data Science // Data Driven #25 (Hosted by FirstMark Capital) · Mar 20, 2014 · at 3:51
Jake Klamka, founder of Insight Data Science, discusses academic career trends and why quantitative PhDs transition into technology roles at Data Driven NYC in 2014.
0:00 / 0:06exact quote · 6.2s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“The amount of PhDs, granted, keeps going up, but the amount of faculty positions has been flat.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Jake Klamka
The biggest mistake companies make is interviewing data scientists like software engineers
“Great question. And so, again, this is biased predominantly towards sort of tech companies, but I think the biggest one is actually interviewing them as if they were software engineers. And so, in particular, the fellows, when they graduate, the program go int…”
PhDs outperform master's graduates in data science because of research experience
“From what I've seen, yes. And I'm, clearly I'm biased, because I'm working with a bunch of PhDs, but I've also, but in speaking with a lot of hiring managers, the answer I've heard is yes. And the reason is the, that underlying skill set of, in data science, y…”
Industry data scientists must aim for 80% perfection, unlike in academia
“Your goal is to go get everything a hundred percent right and then publish, and if that takes a couple of years, then you take a couple years. Ah, in our industry, in, you know, in industry, clearly that's not gonna work. Ah, you wanna probably get to, like, 8…”
Test data science candidates on learning velocity instead of static skills
“Adjust your, if at all possible, try to think of ways to adjust your hiring process to test for how quickly they're able to kind of traverse this gap. Kind of get two data points basically and see how, how quickly they're moving along instead of just kind of t…”