What is the daily job routine of the machine learning scientist? I'm a master CS student in a German university now writing my thesis. I will be done in two months I have to make the very hard decision if I should continue with a PhD or find a job in the industry.
My reasons for doing a PhD:


*

*I'm a very curious person and I feel I still lack too much knowledge. I want to learn a lot and the PhD will help me for that, since I can do more good courses and read tons of papers and be an expert in data mining and machine learning. I love math but wasn't good at it in my undergrad (bad uni). Now in this German Uni I feel I developed a lot of great math skills and I want to improve that because I really really love math! (I was really really bad in math in my undergrad and during my lifetime but now I see I can do math well!)

*I will work with intellectually challenging stuff.

*I need to be honest and say that also I hate to see someone else with a higher degree than me. So if I walk into the street and see someone with a PhD, I don't have to say "oh wow this guy is smarter than me." I prefer to be on the other side. ;)
My reasons for NOT doing a PhD:


*

*I read on the internet about doing a PhD or not doing it. I found out that in the most and usual cases people with a PhD do the same kind of work of people with masters. (that was a general observation in computer science, not about ML/DM).

*I can start a career and make a lot of money in 1 or 2 years, then I can probably start my own company. 
What is not clear yet:
I still don't know what is my ultimate goal at the end. Is it to have a famous little company? Or is it to be a famous scientist? I still don't have an answer for this question yet.
To help me make a decision I want to know two things:


*

*What is it like to work as a data scientist/machine learner with a master degree in the industry? What kind of work you do? Especially when I read those ads on Amazon as a machine learning scientist, I always wonder what they do.

*The same question as before, but with a PhD. Do you do something different or the same thing as with masters?

*Am I going to deal with challenging interesting problems? Or some boring stuff?
As a slight note: I have seen a guy with a PhD in machine learning (in Germany) and is working in a company that promotes a machine learning software. As I understood most of his job is training people to use the methods and software (decision trees ..etc).
It would be great if I can get some answers of experiences related to Germany/Switzerland in some famous good companies.
 A: Before I describe my opinion of job routine, I will pick a few pieces of your post that I think are relevant (emphasis mine):


*

*I'm a very curious person

*Will work with intellectually challenging stuff

*I need to be honest and say that also I hate to see someone else with a higher degree than me (vanity)

*I can start a career and make a lot of money in 1 or 2 years

*start my own company
Based on 1 and 2, you appear to have a very romantic view of data science and research in general. Yes, you will get to work on interesting problems, but certainly 24/7 (this applies to both industry and research). 
Based on 2 and 3, you seem to consider research the pinnacle of human intellect and consider a PhD as a certification of your smarts. I do not agree, because:


*

*there are intellectually challenging problems in both academic research and industry. I think it's a strange assumption that academics face the hardest ones.

*having a PhD doesn't mean you are smart, it means you have what it takes to do good research in your field. Research is not about being smarter than someone else (though it helps). Creativity and approaching problems from a different angle are also very important qualities. If you want some kind of proof that you are smarter than the next person, take Mensa tests, not a PhD. 


In my personal opinion the smartest people are the ones that end up living a happy life with the choices they made, whether that means becoming a nuclear physicist or a carpenter. Don't make your decisions based on whether or not they grant you something to show off with.
Based on 4 and 5, it looks like you envision starting your own company at some point. Be aware that when doing startups, even technology-oriented ones, you are likely not going to spend the majority of your time with the actual technology. Marketing, business plans, management etc. etc. are all equally (if not more) important to successful startups. How do you expect a PhD to help?

Now that these preliminaries are out of the way: my personal opinion on the job routine of a machine learning scientist. First of all: you get to work with state-of-the-art methods on big/complicated/interesting data sets with an emphasis of your choice. It is most certainly very interesting work.
... BUT
Real machine learning involves a lot of grunt work
You will not spend every working hour in a utopian world full of mathematical elegance while an army of computers does your bidding. A large portion of your time will be spent doing grunt work: database management, preparing data sets, normalizing stuff, dealing with inconsistencies, etc. etc. I spend the majority of my time doing tasks like these. They do not grow more exciting over time. If you are not passionate about your topic, you will eventually lose motivation to do these things.
If you have taken machine learning classes you typically get nicely labeled data sets without inconsistencies, no missing data, where everything is as it should be. This is not real life machine learning. You will spend most of your time on trying to get to the point where you are ready to run your favorite algorithm. 
Expectation management in collaborations
If you want to do interdisciplinary projects, you will have to learn how to work with people that know little to nothing about what you do (this is true for any specialization). In machine learning that often implies one of two scenarios:


*

*Your collaborators have seen too much TV and think that you can solve everything, with a fancy algorithm and lots of cool visualizations.

*Your collaborators don't understand the techniques you use and as such don't see the benefits or potential applications.

A: 
•What is it like to work as a data scientist/machine learner with a
  master degree in the industry? What kind of work you do? Especially
  when I read those ads on Amazon as a machine learning scientist, I
  always wonder what they do.

The business problems do not really change depending on your degree, so you would look at the same or similar things. If you work in a big organisation, you work on the company's large datasets. This can usually be product/client data or operational data ( chemical process data, financial markets data, website traffic data, etc.). The generic end goal is to leverage the data to save money or make money for the company. 

•The same question as before, but with a PhD. Do you do something
  different or the same thing as with masters?

The answer is as above, you would do pretty much the same things. However, in the reseach / quantiative analysis / or a similar technical department of a large international corporation, if you have a PhD, you have an edge over someone with an MSc. in terms of career progression. PhD teaches (or is supposed to teach) you to be an independent researcher, so with a doctorate, the company usually 'values' your labour (inquisitive skills and diligence) a bit more. BUT I would strongly advise against doing a PhD, just for the sake of (potentially) faster career progression. Doing a PhD is a hard and -especially towards the end- painful process, you would have to like (ideally love) your subject and also in my opinion have a potential interest to remain in academia (which is proxy to reveal your affinity towards research and the partiuclar topic) in order to make it bearable.
Also bear in mind that going back to industry with a PhD, you will be lagged in the career ladder and may end up being channeled into a technically oriented support role (which pays less compared to those people that earn real money for the company) - which may not be your primary objetive. Finally, if you are working in a small scale company, in your own company, the edge of having a PhD virtually disappears in terms of career progression or salary. 

•Am I going to deal with challenging interesting problems? Or some
  boring stuff?

I guess there is no generic answer to this. ML is cross-disciplinary. If you work as an analyst, you would usually look at data and try to build models, if you are on the development side, you end up dealing with the knitty-gritties of implementation. If you are client-facing, you may have to do a lot of hand holding and training of clients (but likely earn more money). Usually, the answer to your question depends on personal preference and also how much flexibility your employer provides. 
A: Alex, I can't comment specifically on Germany or Switzerland, but I do work for an international company with a staff of over 100,000 people from all different countries. Most of these people have at least graduate level degrees, many have Masters and PhDs and, except for the HR and Admin staff most of us are expert in one or more different scientific domains. I have more than 30 years experience, have worked as a skilled scientific / technical specialist, a manager, a Project manager and eventually returned to a purely scientific role that I enjoy. I have also been involved with hiring staff and perhaps some of my observations that follow may be of value to you.


*

*Most new graduates really don't know exactly what they want and it usually takes a few years to find out. In most cases their workplace experience turns out to be quite different compared to what they had expected for a range of reasons. Some workplaces are exciting while some are dull, boring and "workplace politics", bad bosses, etc can sometimes be big problems. A higher degree may or may not help at all with any of these issues.

*Most employers want people who can "do the job" and be productive as soon as possible. Higher degrees may or may not matter, depending on the employer. In some situations the door is closed UNLESS you have a PhD. In other situations, the door may be closed BECAUSE you have a PhD and the employer wants someone "less theoretical and with more practical experience".

*A PhD does not necessarily mean faster promotions or even much difference in salary and may or may not make any difference to the sort of position that you can obtain. Generally when I have been interviewing candidates, I have been most interested in finding people with relevant work-related experience. A PhD might be a final deciding factor in securing a position, IF the candidate's thesis topic is specifically relevant.   

*People tend to change jobs more often now than they used to in the past. Your age divided by 2*pi is not a bad rule of thumb for a good number of years to stay in a job before you start going around in circles. Some people work for a while and then return to higher studies. Some people (like me) start on a PhD and then get an "offer too good to refuse" and leave the PhD to go and work. Am I sorry I did that? NO, not at all, and if I were starting over again I would do a PhD in a completely different topic anyway.

*The best suggestion that I can give you is to do what you most enjoy doing and see how it unfolds. No-one else can tell you what will be best for you. Sometimes you just have to try something and, if it doesn't work out, then learn as much as you can from it and move on to something else.  As Rodin said: Nothing is ever a waste of time if you use the experience wisely.    
A: Or you can try to join some research group where statisticians and machine learners are not an everyday appearance. For example infestation and disease spreading, botany or ecology, social insect or maybe social sciences? 
I can´t give you exact examples, but if you are a good statistician/ML at a place where there are only few of them, than people and different research proposals will find you. The point is, that you will be really in demand without too much effort from your side.
If you like that idea, than try to search for machine learning problems outside your current topics (industry), and maybe you will find the way how to find your "challenging interesting problems" and "work with intellectually challenging stuff". 
A: I agree with the other answers. I would just emphasize that one common way (at least in the US) for people like you who hesitate between continuing with a PhD or doing the industry after their undergrad degrees is to apply for PhD, then take a leave (one year or more) if things aren't as great as they expected or simply want to explore industry. It is generally easier to apply for PhD right after undergrad: you haven't forgotten yet the habit to cram exams (GRE), professors who are going to write recommendation letters for you still remember you well, etc.
Also, in your comparison between PhD and industry, amongst the opportunities you have, you might want to compare the access to interesting datasets, computer cluster availability, software engineering skills of the place and how many people are assigned for each project.
Lastly, you can find a lot of intellectually challenging stuff in the industry as well, e.g. check out IBM/Google/Microsoft/Nuance/Facebook/etc. research department (just like you can find a lot of intellectually unchallenging stuff academia). E.g. the folks behind SVM were working at AT&T, IBM Watson is at IBM, Google Translate is one of the best machine translation system, Nuance and Google have the top voice recognition system, and those are very far from isolated examples. In fact I've always wondered who among industry and academia contribute the most toward machine learning research (I had asked the same question regarding the database research on Quora: Has database research been mostly driven by the industry over the last decade?).
A: To get a PhD, you have to advance the state of human knowledge. You don't just have to learn more stuff. You have to produce something original. This is a long, slow, and painful process, and not everyone succeeds at it. So you should do a PhD only if you think you have a new, creative, contribution to the field in you.
If you just want to learn the field and apply the field, take your Masters at most, and then spend the rest of your life learning while you apply. Read things. Take the occasional workshop. If at some point you are infected with the urge to do something truly original, take a (long) break from career and try to get that PhD then.
A: When you choose the /famous little company/ route, you have the freedom to establish a research department in your company.
Here, you can get annoyingly creative, as in, unrestrained... explore all your childhood fantasies, intellectually challenging stuff... you set the pace... you will be /the man/.
You don't have to sit at University Labs to write a /Killer/ research paper. 
That notwithstanding, while at it, you can always coordinate with relevant research departments back at Univ. see...? zwei vögel mit eines stein :-)

...someone else with a higher degree...

Well, vanity, in moderation, motivates us to seek the best that there can be.
Good luck.
yb
