How is the development of data scientists?

I feel that the career development direction of data scientist (DS) is most related to two factors, one is the background (or the doctoral field of undergraduate course), and the other is the opportunities encountered in the career process.

First of all, for the vast majority of people, if they choose a career direction after graduation, there are three types of people who will become or can become DS:

Fields are naturally related to ds, such as CS, Stats, machine learning, partial CS, computerized informatics, etc.

There is no natural connection between fields and DS, but the academic background is very specialized. These fields often rely on data and programming, such as computational biology, neuroscience and mechanical engineers. ;

This field can be transferred to DS, especially from the level of business/social impact, such as political science, global health, etc.

The biggest difference between 2 and 3 is that programming and data often run through 2' s daily work and study, while 3 often knows the development direction of things and tends to look for data to support his point of view. The biggest difference between 1 and the other two points is that the degree of professional subdivision is poor, and only the method is learned, and then the context is specifically learned in the factual scene.

The reason why I talk about my background instead of my development direction is because my background basically determines the direction, and this decision is active and passive (market and industry needs).

For 1, most people will go to tech or fintech (some fintech may prefer some majors, such as physics and mathematics) because:

They have no special training, which is what tech and fintech like.

Going to other fields can only be a support department, and there are relatively few opportunities to contact the forefront, and career development is easy to face bottlenecks.

For 2, most people will not go to technology or financial technology, but will eventually go to companies in their own fields:

Because it is a special training, if you just go to tech to be an ordinary DS, then what you have learned for so many years will be in vain;

Technically speaking, the position at the same level competing with ML, Stats and CS is still inferior;

Now many giant companies have specialized decentralized research institutions, such as Google's Verily (or Alphabet is more suitable), and there are many opportunities that are more suitable for them;

Most importantly, these highly professional companies basically do not recruit people with pure technical background to do core positions (such as the R&D department of pharmaceutical companies).

For those who are more interested in business in 3, 1 and 2, most of them will go to tech, vendor/service company, big4 or consulting. In short, the positions they choose tend to be analytical or BI:

Tech field is more inclusive, and some DS positions are not very technical.

Nowadays, in order to adapt to the trend of big data, many consulting and big4 have many DS positions.

These positions often require people with strong understanding and communication skills, and technology is not necessary. On the contrary, people who know a little technology and have strong communication skills are most needed (of course, there are also some positions that must require PHD because of the client)

It doesn't mean that people who do business without technology are low. In fact, after 10 years, many people who take the management route come out of this track (a small manager/director who manages two people can't be regarded as the management route at all, and the management route refers to VP and above, such as EVP and CEO).