musings on music and life

April 8, 2016

The tech mindset, and why it fails in science

Filed under: Philosophy — sankirnam @ 2:05 pm

One of my very close friends (an ex-Google[X] employee), sent me this article last week; it seems that the dysfunctional management style in Verily (a subcompany of Alphabet) has come to light. My friend was astute enough to see the writing on the wall and flee before the s**t really hit the fan, and is now better for it. Derek Lowe also discusses this article, and my sentiments reflect his. An ongoing theme I’m seeing today is that executives who were successful in the tech field try to apply the same approaches or mindset to scientific research, which is quite different in nature. One example I can recall immediately is the former CEO of Intel criticizing the slow, largely unsuccessful method of pharmaceutical research, for which he was publicly lambasted. This all comes down to two things: epistemic arrogance, and a failure to realize the “domain-specificity” of one’s knowledge or expertise.

For example: those working in tech will be used to continually optimizing code to work faster. There is always some way you can look through a enormous process find something to tweak that will make the program run a couple of seconds or minutes faster. However, this same mindset cannot be bought to bear on science (I’ll give some examples from organic chemistry, since that is what I know best). I’m still baffled by the concept of the “10x” or “100x” engineer in tech, because such a situation is impossible in fundamental science. You can be the smartest guy in the institute, but you can’t make a distillation go any faster – it is bound by hard physical constants (i.e. the boiling point). Yes, you can do it under reduced pressure, but even then, if it is not that volatile, you can’t really speed it up further if you have it under high vacuum already. If the liquid boils between 30-60 °C at atmospheric pressure, then you will have to do it slowly and carefully – being 100x more intelligent than others in the room becomes irrelevant. Similarly, if you’re taking an NMR, 13C or other heteronuclear NMRs will take a long time to acquire because those nuclei relax very slowly – again, these are hard physical constants and there is no way around it. I’ve done more than my fair share of time-consuming, tedious, “grunt work” for which there was no better substitute (for example, I once had to purify and isolate a compound by vacuum distillation, but the crude isolate was obtained by extraction with about 500 mL of organic solvent, which had to be slowly and carefully transferred from a 500 mL RBF to a 25 mL RBF so that it could be distilled. Oh and did I mention that the crude was only soluble in dichloromethane, which has a notorious tendency to bump upon rotary evaporation?). There’s no real way to “optimize” what I did, which is something that experienced chemists will agree with, but those in tech will not be able to wrap their heads around. Similarly, reaction kinetics are bound by physical constants; you can’t have a rate faster than 10-9 M/s (which is basically the rate of diffusion in liquid). Most reactions with common electrophiles and nucleophiles will not be nearly that fast, but the rates can be quantified, and the limits calculated thanks to the extremely nice work done by Prof. Herbert Mayr (Ludwig-Maximilians-Universität) over the decades. His work gives ranges for reaction rates, which again are bound by hard physical parameters, which can be semi-empirically approximated (to greater and greater levels of precision as time passes) using theoretical methods.

So what’s my point in all this? Adapting the tech mindset to science is bad, because: 1. You can’t optimize everything indefinitely; 2. While the “10x” or “100x” engineer may be a real thing, there is no such thing as a “10x” or “100x” scientist; 3. “Science” is just inherently more time-consuming and expensive than just sitting at a computer and busting out code; 4. Even in “science” the fields are very different, and expertise and knowledge are domain specific (a chemist may be an expert in organic chemistry but know nothing about quantum dots, for example) 5. You need humility to realize how much we don’t know, and this is reinforced by the failures that scientists face in the lab on a daily basis. Continued success in technology has made the executives there complacent and arrogant, in my opinion.


Leave a Comment »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at

%d bloggers like this: