Happiness of a scientist III: testing ideas

If you like rationalizing events (see Happiness of a scientist I), you would probably enjoy convincing yourself that your jobs actually fits with your initial expectations, and this will require a certain amount of rationalization and a posteriori reinterpretation of your expectations ;-)
So this is why I will try to do now.

Scientific research is about understanding phenomena (whether natural [e.g. physics], human [e.g. economics] or abstract [e.g. mathematics]). A scientific breakthrough can either be a discovery of a new phenomenon, a new insight about an existing phenomenon, or an explanation/justification/proof of an insight.
The point here is that it is all about having a better understanding of things.
What for? Well, besides the joy of understanding, the immediate consequence is the ability to positively impact the real world: e.g. build new systems or better ones, do things you could not do before, make a step forward technologically...
Of course, this is not necessarily the main motivation nor is it the main justification for scientific research, but it can be seen as a nice side-effect of understanding.

In any case, if you come up with a new idea, whether you care about applying this idea or not, the first thing you want do to is to test this idea. Testing can be done in two ways

• publishing it (so that your peers can comment on it, argue, verify, derive consequences...)
• experiment it (so that you can verify by yourself that the new understanding is indeed correct)

It is often difficult and time-consuming to design and execute experiments, so many scientists are satisfied by publishing.
But if one is given the necessary resources, experimenting can be a good substitute to publishing.

In private companies, not only do you often have resources for experimenting ideas, but also you are encouraged to do so (this is your part of your 80%) and even further, to go beyond experiments towards real-world applications (which give the ultimate validation of your ideas).

The conclusion I want to draw from all this is the following: one can be very happy doing research in a company because publishing is advantageously replaced by real-world testing of ideas.

[Note: I carefully avoided the topic of long-term research vs short-term engineering solutions... I might deal with this in a future post, although it is more comfortable not to face this at the moment ;)]

Comments

[marco]

thanks for these comments! They do make a lot of sense. As you state it in the end of your post, you do not address the issue of short-term vs long-term research. However, as scientists involved in machine learning, I believe we are already quite biased towards experimental evidence, e.g. things that work. This, in my opinion, shouldn't be interpreted as a justification to push academic research in the field towards (re)discovering tricks and hacks that seem to work now. On the contrary, I really believe originality to be super important in Academia. What I just mean is that for people working on this topic, the gap is probably a lot smaller than say, for someone involved in "pure" maths, and that our interest for data already sets us quite close to the motivations of industries.

[Patrick Haffner]

I agree that to have an idea vindicated by an experiment is often an equal or stronger source of pride and satisfaction than being accepted for publication.

But, and this is the pitfall of working for a company, one who is not encouraged to publish is often just happy whese these experiments only his colleagues have heard of. He ends up not publishing these experiment, and come to regret it later, out of seing somebody else being credited for the discovery.

This is a just punishment: not publishing experiments should be a crime...

I have been guilty of this crime many times.

[Shanker]

Hi!

Interesting blog! :) Im studying machine learning for the first time and Im stuck in what looks like an unescapable mess..
Im doing an assignment to train a program to recognize "fashion" in a certain environment. Ive collected data and I was looking for a suitable algorithm. However, I couldnt find any because my data set has only "yes" examples and most of the approaches(decision trees, etc) need both "yes" and "no"s. Would you happen to know how I can solve this problem or suggest an approach? I realize that this comment is completely off the topic and unrelated, but Im really stuck at a loose end and Id appreciate the help :)
Thanks

[christbirawan]

Usefull info's

[Papers on Research]

looks quite a great post, it's having good information for research analysis. great job

[Ithri]

Hello Olivier,

Very interesting thoughts ! I share completely your mind and I exactly ask the same questions to myself every minute of my life! Many research teams pretend actually that they are doing research, but when you join them, they just want you do more engineering than research, moreover, they want you publish with that !!
I think that there are two main reasons for that :
1- Serious research is very hard to do and is very demanding (in intellectual effort of course (often suffering), in time, in sacrificing some of your private life mainly when one idea crossed your mind and you do not want it to go like that, or to be competitive with other researchers and teams, etc.)
2- Nowadays, research is very related to money (because money for many is comfort and so it is actually !). So many research leaders care more about satisfying the terms of their research projects contracts (very often engineering solutions are sought by the funding organization(s) ...), and about publishing anything in conferences (and usually this work is done by Masters or PhD students) and offer to themselves free trips, at the expense of good quality research of course !

Though the two points are tightly linked to each other ... lol

Having said this, it is not easy to judge researchers because this suggests knowledge of all the parameters surrounding them (either research parameters or private life parameters), which is not the case of course ...

Thanks,
Ithri
PhD.