Please create an account to participate in the Slashdot moderation system


Forgot your password?

'Data Science' Is Dead 139

Nerval's Lobster writes "If you're going to make up a cool-sounding job title for yourself, 'Data Scientist' seems to fit the bill. When you put 'Data Scientist' on your resume, recruiters perk up, don't they? Go to the Strata conference and look on the jobs board — every company wants to hire Data Scientists. Time to jump aboard that bandwagon, right? Wrong, argues Miko Matsumura in a new column. 'Not only is Data Science not a science, it's not even a good job prospect,' he writes. 'Companies continue to burn millions of dollars to collect and gamely pick through the data under respective roofs. What's the time-to-value of the average "Big Data" project? How about "Never?"' After the 'Big Data' buzz cools a bit, he argues, it will be clear to everyone that 'Data Science' is dead and the job function of 'Data Scientist' will have jumped the shark."
This discussion has been archived. No new comments can be posted.

'Data Science' Is Dead

Comments Filter:
  • data scientist (Score:5, Informative)

    by schneidafunk ( 795759 ) on Wednesday March 05, 2014 @11:47AM (#46408659)

    Call yourself a statistician or database engineer and I promise there are still jobs around. And contrary to the summary, they are highly valuable jobs.

    • indeed, and who cares if someday this alleged "fad" goes away, "get it while the gettings good. and then get out!"

      • You can't do that, unless you can figure out how to make and file TWO resumes. Different ones, I mean.

        Man, these data scientists are all pipe dreams.

        • by iggymanz ( 596061 ) on Wednesday March 05, 2014 @12:23PM (#46409071)

          no pipe dream, my employer has those people, making big money

          and what's this nonsense and misconceptions about resumes you have between your ears?

          in my career, I've held engineer, science and IT positions. I have "IT-flavored" version of my resume for when I'm seeking an IT focused type job, "engineering-flavored" one, etc. All the resumes are true, no innacuracies and all experience can be verified by contacting previous employers or talking to former coworkers or reference. So the point is of course you can have different versions of your resume with different focus on duties and skills.

        • by SQLGuru ( 980662 ) on Wednesday March 05, 2014 @12:51PM (#46409461) Journal

          You should have a different resume for each job you apply for......

          Keep a master resume with all of your details. When you apply for a job, copy the master and pare it down to the information most relevant to the job you are applying for. Then edit the result so that you look like the perfect candidate.....update project descriptions to emphasize the same buzzwords in the job listing, etc.

          If you only have one resume that you blast out to a ton of different job listings and they'll probably focus on the projects that are meaningless to their situation and weed you out more often than not.

          Oh, and BI (business intelligence) is still going strong at the company I work for......(not my area, but there's still tons of work under that label).

        • You can't do that, unless you can figure out how to make and file TWO resumes. Different ones, I mean.

          Man, these data scientists are all pipe dreams.

          Well, it is not rocket science to have more than one resume. You have one work history, but you will use more than one resume format to present it in different (but veritable) ways according to the situation.

          See, you are supposed to have multiple versions of your resume (which are true and accurate of course) according to job postings or fields of concentration. If you have a varied work experience, or you are contemplating lateral moves, this is a must.

          Consider the following situation I had to deal wi

          • and you need a college degree loaded with theory in construction that is more for high level design but not for the hands on doing the work part and HR does not like tech / trades schools that do teach the hands on doing the work part.

    • Re: (Score:2, Interesting)

      by Anonymous Coward

      As the author of the article, I'm happy to encourage people to call themselves statisticians, database engineers, etc. These roles are definitely in demand and will never go away when the bubble for "data scientists" pop.

      I'm just concerned about the recent spate of large companies trying to hire data scientists to "save" their expensive big data projects that arent producing actionable insight. Those jobs are a dead end.

    • by Anonymous Coward

      I don't know about you but I am sick and tired of DICE's attempts to
      channel and steer the employment market through astroturf postings
      to Slashdot, which they also happen to own. Most of what the talking-heads
      at DICE churn out regarding employment is simply untrue. Not 'not-the-truth'
      as that they don't know any better, but telling lies as in spreading deliberately
      misleading information, as in telling a mean-spirited lie.

      DICE is not a platform for you and me to find lucrative jobs. Instead it is the
      other way

      • But don't forget, tell Dice to junk the Beta UI for Slashdot! It is a social media scam, dsigned to simplify the interface into a blog so it is easier to data mine. So FUCK Beta, and FUCK analytics, and FUCK business data, and if needed, FUCK data scientists.
    • by Sir_Sri ( 199544 )

      Indeed, the job of doing serious data analysis is not new, 'data science' is really only a temporarily independent field while we try and sort out some of the technical problems that arise from working with large datasets. Once solutions for large datasets become mainstream, reliable and agreed upon (and inexpensively vendor bought) we'll go back to just having scientists who are specialized in whatever area.

      And that's perfectly alright.

      • Only if you have testable hypothesis. The "Science" part is dubious for exactly the same reason the "Science" in economics is dubious. The "Science" in Computer Science is a little more substancial, but it is closer to mathematics than any emperical science.
    • Call yourself a statistician or database engineer and I promise there are still jobs around. And contrary to the summary, they are highly valuable jobs.

      Is a data scientist really a pseudonym for SEO analyst? (SEO = Search Every Option or Search Engine Organization or that other one, ending in Optimisation)

  • Maybe (Score:5, Insightful)

    by stephencrane ( 771345 ) on Wednesday March 05, 2014 @11:47AM (#46408661)
    But this general domain in the realm of contemporary giant data sets is the basic science research of our times. To say that 'data scientist' roles are dead in the near future based on a ROI analysis is to suggest that all these huge data sets aren't likely to pay off for a corp in the near future. And that doesn't sound right at all.
    • by gnick ( 1211984 )

      I hold a Master's in EE specializing in Information Theory. That seems to sound like a "data scientist", but I've never met anyone that presented themselves as one.

      Maybe I should modify my resume to include my years of experience as a "nocturnal ergonomics specialist" or "cinematic purveyor".

      • Lot's of people are up in arms here thinking im against data, queries or even information in general. LOL

        I think guys like you are going to be in serious demand, and yes data is piling up hugely. The winning organizations will have serious, smart and well trained people who know how to manage it.

        And yes, I think "nocturnal ergonomics specialist" is the career of the future =)

    • To say that 'data scientist' roles are dead in the near future based on a ROI analysis is to suggest that all these huge data sets aren't likely to pay off for a corp in the near future. And that doesn't sound right at all.

      I think what's really going on here is that lots of organizations have jumped on the Big Data bandwagon expecting that it will be easy, and hired lots of people who don't really know what they're doing (because they also saw an opportunity). There's lots of value in large corpuses of unstructured data, but teasing it out requires more than just a desire and some computing resources. As the field matures and builds up a well-known set of techniques which can be packaged up and on which people can easily be t

      • But what may be closer to the mark is that many companies go into this expecting a quick return and discover that it takes more effort than that, and so they move on to some other fad of business. It is the short leash they are on held by investors who are clamoring for instant ROI that drives this, not a real scientific curiosity. So do you think that counting clicks or doing RegEx matches in blog posts translates to pay back? Or is it just the same old confidence game that has always existed, you find a f
    • I'm not against big data sets.

      I've just noted that a lot of big data projects hyped by vendors are misguided, and that there will be very large and visible failures coming up. this is just a bubble and the "data science" guys will be in trouble.

      The long term prognosis for large scale data analysis is good, and machine learning will probably yield good results. Also the pendulum will swing back to structured data.

  • by Anonymous Coward on Wednesday March 05, 2014 @11:49AM (#46408679)

    That's a very strong claim, I'll need to consult my Data Scientist to see if it actually fits the data.

  • How appropriate (Score:5, Insightful)

    by Micklat ( 986895 ) on Wednesday March 05, 2014 @11:50AM (#46408687)

    No data has been cited during the creation of that blog post.

    Opinion is fine, but when the observations are so weeping, just a little bit of substantiation is nice to have.

  • FUD (Score:5, Funny)

    by aBaldrich ( 1692238 ) on Wednesday March 05, 2014 @11:50AM (#46408691)
    How to prevent more people from flocking into your field:

    1) Write a Slashdot article
    2) ???
    3) Profit!
    • Re:FUD (Score:5, Informative)

      by aBaldrich ( 1692238 ) on Wednesday March 05, 2014 @11:59AM (#46408793)
      According to this guy, Mathematics is not a science because you don't conduct experiments. The key error is this:

      Science creates knowledge via controlled experiments

      Which is false. Science checks hypotheses and tries to prove them, or makes repeated experiments that show the failure to disprove them.

      • Science creates knowledge via controlled experiments

        I think astronomers will be very surprised to learn that they aren't scientists.

      • by ZahrGnosis ( 66741 ) on Wednesday March 05, 2014 @12:28PM (#46409145) Homepage

        "Science" lacks a robust definition, but clearly the OP's definition is overly simplistic and narrow. Stephen Hawking has a lecture somewhere (found it: where he talks about the idea of the "positivist" approach defined on the ability to predict outcomes, and I like to apply that definition to Science (Hawking doesn't, directly, but it's sort of an underlying theme). That is, Science becomes the observation and experimentation required to form predictions or cause changes in predicted outcomes.

        So Social Science can be a science in so far as it actually informs usefully on how people will behave or provides useful ways to affect and improve the behavior or state of society's future. Computer science is a science insofar as it is required to make computers function as expected (as predicted) -- if you want something to perform faster, you must do the research and experimentation to cause the outcome to be faster. Even archaeology can be a science by this definition in that discoveries are added to a general model of the past that predicted all sorts of things -- ancient society's behavior, glaciation, geological events... "predict" may be a stretch there (except when archaeological finds help predict the future), but in this case the method of building a model of how the world worked based on observation to describe and generalize behavior (of the earth, of ancient religions, or what have you) is a form of prediction; it's just after the fact.

        Data Science is very much science in this form; the job of a data scientist is almost universally to predict what the data will say about the future given what it has said in the past. This is invaluable to businesses and while the name may fall into disfavor, in the same way "actuary" which means something very similar already has, the abuse in this article is unwarranted, unfounded, and inaccurate. I will only agree that many who sport the "Data Science" moniker may not actually be doing science by any definition, but that's the individual's fault, not the concept's.

        • thanks for your thoughtful post. I agree actually that the word "Science" is a bit elusive in definition, and that certainly just because a field of inquiry is not a science by one definition does not mean it's without value.

          Also, I accept that there are multiple definitions of the word and some may include fields like Computer Science and Social Science, which of course are all legitimate fields of inquiry.

        • I think your post makes good points mostly, but I have to ask, since when did the phrase "actuary" fall in to disfavor, and when was anyone going to tell me?

          A P&C actuary

          • Thanks fort he nice replies... "disfavor" was probably strong, in terms of "Actuary", but I meant it more in terms of not having a negative connotation, just not having a rock-star or very popular buzz-word sort of connotation that "Data Scientist" seems to have now. I grew up wanting to be an actuary, curiously -- my father was one -- I got the math degree, I just wandered off into database work, but I do actually see our jobs as very similar. I think actuaries still come in as some of the "best" jobs an

      • Trying to prove a hypothesis is fraught with danger--witness the use of complexity as a "proof" of intelligent design. Failure to disprove is about the best that science can do while maintaining its objectivity. That's not to say that working scientists don't get attached to their ideas and try to "prove" them, but they're not supposed to. If the idea is sound, you can hammer on it all you want, it'll stay standing.
      • Re:FUD (Score:4, Funny)

        by tomhath ( 637240 ) on Wednesday March 05, 2014 @01:50PM (#46410195)

        According to this guy, Mathematics is not a science because you don't conduct experiments.

        He's obviously wrong. Try this experiment - it proves addition:

        # python

        # print 1+1

        # 2

      • According to this guy, Mathematics is not a science because you don't conduct experiments. The key error is this:

        Science creates knowledge via controlled experiments

        Which is false. Science checks hypotheses and tries to prove them, or makes repeated experiments that show the failure to disprove them.

        It's not just in mathematics that this is false. Controlled experiments are one of doing science. By the definition of the author of the original article, combing through existing genomic data to identify SNPs possibly associated with disease isn't science. By his definition, identifying comets or asteroids or planetoids by examining collections of astronomical images taken by others isn't science. By his definition, the work of theoretical physicists who do not perform experiments isn't science. By his def

  • by sega_sai ( 2124128 ) on Wednesday March 05, 2014 @11:51AM (#46408699)
    The author of this piece clearly have never done actual science, as confirmed by his resume, and his opinions on what science is and that somehow some observational sciences are "soft" are very questionable at best.
  • shouldn't there be a link to an article or a more in-depth argument presented than just "b/c i think so"? Perhaps, say, explain who the hell Miko Matsumura is, or provide greater context?

    I get it though, nobody reads the articles on slashdot... :/

  • 100% disagree (Score:5, Insightful)

    by netsavior ( 627338 ) on Wednesday March 05, 2014 @11:53AM (#46408719)
    In my career I have worked for boring banks and boring monolithic enterprise software giants.

    If there is one thing I know for certain it is that big enterprise will ALWAYS have a huge appetite for quantification of data. It almost doesn't matter if it actually does anything for you, executives at giant corporations have to DO SOMETHING have to REVIEW SOMETHING. Large scale data aggregation and reporting (one of the many things that go by "big data") might not have sciency uses, but any time a V level can provide a C level with "something" that says "We are doing stuff" there will be a huge market for it.

    Basically what I am saying is, even if "Big Data" is nothing but a placebo, like say "HR Training", "Wellness programs", "performance reviews" or "teambuilding" it is a permanent fixture in the big, boring, high paying, stable job providing corporate world.
    • I agree with you, but I think there's merit even in those quantifications that turn out useless. This sort of science, at least the way you describe it being done, is really shooting in the dark. Sometimes you find an interesting and meaningful correlation or analysis, but more often than not you're juggling and sorting numbers to no useful end. People will always pay to keep taking shots.
    • by Anrego ( 830717 ) *

      100% agree (with your post).

      The argument made in article is ridiculous, but even if we grant it and companies suddenly lose interest in their data, the skills used to analyze it can easily be rebadged and applied in other fields. Not like all the large scale infrastructure, data management, and algorithm skills are only specifically applicable in the "big data" realm, that's just the most profitable place to apply them at the moment.

      IT isn't a job where you learn a skill then make money from it forever. You

    • by Anonymous Coward

      Yep, managers and bosses love widget counting even if your business has little to do with widget production. You'll always have a job if you can give your boss some table or graph to wave around in meetings, meaningless or not; kept me employed for 20 years.

    • fully agree, data and big data will always have a home, and corporate behavior is very much how you describe.

      Just thinking that the pendulum is going too far from structured data and that people are getting a bit ahead of themselves with respect to vendor hype in this area.

  • by Anonymous Coward

    Seriously. Who is this guy and why does his opinion matter? Either he has credentials or he backs up what he says with some evidence. Neither seems to be the case.

  • TFA is BS (Score:5, Interesting)

    by Sarten-X ( 1102295 ) on Wednesday March 05, 2014 @12:02PM (#46408829) Homepage

    Unfortunately, unless this is structured data, you will be subjected to the data equivalent of dumpster diving. But surfacing insight from a rotting pile of enterprise data is a ghastly process—at best.

    Sounds like this Miko Matsumura has no idea how successful Big Data projects actually work.

    To refine his analogy, unstructured data is much like processing recyclables. Everything that might possibly be good gets thrown into a large bin, and several sorting processes run to extract individual relevant (though messy) pieces. While those pieces alone aren't pure enough to be useful, there's enough meaningful information in them that statistical analysis can separate the good from the bad, and that's where the insight comes from.

    With a typical RDBMS, insight is readily apparent. A hypothesis that 75% of a user's purchases were widgets is simple to verify. In a non-relational database, as is often used in Big Data projects, that would be an inefficient computation (though it can be done). Rather, those databases are more aligned to produce a whole list of correlations between user demographics and purchasing habits, showing for example that users who buy widgets have often already bought foo bars. The "Data Scientist" didn't have to ever look specifically at statistics for widgets or foo bars, but the correlation is presented in a nice and accessible form, gleaned from millions or billions of independent data points.

    Miko Matsumura is a Vice President at Hazelcast, an open source in-memory data grid company.

    This is a SlashBI article written by executives for executives, with little basis in fact. Lovely.

  • Buzzwords (Score:5, Insightful)

    by SpankiMonki ( 3493987 ) on Wednesday March 05, 2014 @12:09PM (#46408901)
    Since "Data Science" is dead, do we go back to using the old buzzwords? Or do we have to wait until some marketing MBA whiz-kid comes up with a sexy new word for "Analyst"?
    • hardly dead, analytics and data analysis are still huge and growing field with high salaries for those with formal training. my employer has department of such people.

  • by wcrowe ( 94389 ) on Wednesday March 05, 2014 @12:12PM (#46408937)

    The term reminds me of "Computer Scientist". I remember a TV commercial from the 80s for a digital watch that mimicked analog watches. The announcer would declare that the watch had been designed by "computer scientists" while an actor was displayed wearing a lab coat and looking at the watch under a microscope. The first time I saw it I was afflicted with fits of laughter.

    • Lets get rid off the "Dexter Lab Coat" stereotype.

      Many people, kids or adults doesn't really like to use "lab coats" & prefer to use the clothing they use on the school or street.

      • Although there's noting like a good serious apron if you're working working on stuff. Lab work? Painting? Engine rebuild? Cleaning the basement? Grilling up a mess of bacon? Wear an apron.

      • Actually not the reason people where lab coats is to stop thier clothes getting dirty/ruined by chemicals etc also there is a status thing going on as blue collar workers will probably be in overalls a white coat marks you out as a "professional" why do you think Abby wears a lab coat in NCIS - part of its is to mark her out as senior.
      • You do realize that there are practical reasons to wear a lab coat and it's not just a fashion statement?

  • Sometimes I think slashdot should really do a better job of filtering these types of things, or at least highlight that this is an opinion and the person writing it has no clue what they are really talking about. I work in the BI space and do everything from Analysis, Architecture, Dashboards, Reporting, ETL, and any other job that fits into that space. We do have a data scientist here and he does nothing close to what this article talks about. In fact, I would argue that if you do the types of jobs this
    • Data Scientist, may sound Statistics to some of us, and Data Base related to others. Maybe you are right, and we are wrong.

      Could you provide a link to a web site, where Data Science could be described ?

      I do believe that concepts like Data Statistics, Relational Algebra, SQL, could be consider part of a Data Science diploma, but, none of this concepts should be consider individually Data Science, by themselves.

      Maybe, a new proposal for a Data Science diploma, could be created from this post.


  • by Daniel Hoffmann ( 2902427 ) on Wednesday March 05, 2014 @12:19PM (#46409021)

    90% of what a data science expert do is what people like to call data-juijitso (data reconfiguration). Which basically means getting data out of your RMDBs, SAP, Twitter, Facebook, random text (.csv, etc) file dumps, random Excel/Word Files and legacy databases and into some place you can actually generate conclusions from (like inside a HDFS Hadoop cluster). Plus during this process you need to normalize all your data so you can apply the same algorithm no matter where the data came from.

    All this means is that you will spend countless hours trying to connect to the client legacy stuff and then countless hours trying to get the data out (without impacting production systems!), so you can then spend countless hours formatting this data around to be able to spend countless hours trying to get this data into your Big Data(tm) solution so you can finally run some algorithms and create results. Now multiply all that by the number of different kinds of databases the client has and you get the idea.

    As an IT professional you really do not want to work in this field. No organization keep its data in a clean uniform way, data scientist is like an IT janitor.

  • Um, I'm a calibration scientist. My job is to pick through data and look for errors, which I then correct. I'm a scientist, not an engineer, because the data and its errors are from real physical processes. (The data I work with comes from multispectral satellite instruments.)

    If I can't call myself a 'Data scientist' on a resume, what term should I use? Approximately zero jobs are available for a 'Calibration scientist'.

    • As an addendum, the day I put 'engineer' on my resume is the day my career is over. My degree is in theoretical physics. I have zero engineering background or training. I'm a scientist, and I can't compete with engineers for engineering jobs, nor do I want to. I've spent decades keeping the word 'engineer' off my job title and resume despite stupid managers trying to tack it on.

      • I once had to fight to keep 'software engineer' off my job title.

        I kept explaining that I was not an engineer, that engineering was a specific regulated profession, and that to call me an engineer would be illegal and incorrect.

        It took a long time for our Personnel Management Engineers (HR) and the VP (who thought it sounded cool) to understand why it had to be otherwise.

        In some cases, companies like it because it sounds cool. But they have no idea that you can't simply call yourself an engineer any more t

        • by geekoid ( 135745 )

          IN some sites it's a crime to be called engineer if you don't have a PE Cert.... or drive a train.

          As it should be,

        • In what country in both the USA and the UK (The main English speaking country's) its not illegal to call your self an engineer if you don't have a Ceng or pE
    • Its a problem with the computer sciences in general. They keep cannibalizing terms and changing he meaning, like engineer and scientist.

      I am an aerospace engineer that specialized in fluid mechanics. Guess what? I know more about computer programming, linux, unix, memory architectures, compiling environments, programming language, etc. than most computer "scientists/engineers".

      But I need to know about all those things because I take the theory and discretize it for solution on whatever architecture

      • by mikael ( 484 )

        It does get confusing especially with job adverts... I see titles like "software consultant", "freelancer", "programmer" (for visualisation work), "scientific programmer" (for parallel processing research into fluid dynamics), member of technical staff, test engineer, compiler engineer, software engineer, senior software engineer, principal engineer, architect, as well as data scientist (with "Big Data", R, Java/Hadoop and Reduction).

        The main different between a programmer and an engineer was that the engin

      • by geekoid ( 135745 )

        really? Most ?


        "computer science degrees are simply collections of network certs."
        No, they aren't

        I don't think you know what a computer scientist is or does.

  • There are several subjects disscused in this post.

    * First, wheter there is a "Data Science" career. Since, I discover how math can be applied to handle data, when learnt Relational Algebra, & Data Normalization, in Collegue, I realized there could be a "Data Science" school diploma. And include OLAP, Key-Value, NoSQL, Statistics data handling, with their respective math theorical support.

    * Second, like any other career, there is a diference in how is taught in school, & how is applied in business.


    • this is a good and rational perspective. I think if there is a formalization of the concept of Data Science including academic certification, maybe it could become more credible.

  • by Anonymous Coward

    Big Data: the belief that as the size of a pile of shit increases, the probability of finding a pony approaches one.

    • Y'know, once your pile covers a large enough area of land, there's bound to be a pony in there somewhere.

  • Synergized leverage of 5,000 mile view with a stakeholder buy-in increases both capacity and bandwidth of core clientele and creates engagement with the low-hanging fruit.


    Shut up, and code.
  • UA, a Baltimore company hosts data science meetups. Why? Because UA is data science driven. All company decisions are made based on data. So it seems that the OP is complete BS, because it is effectively creating results, and those results are highly successful for a major corporation.

  • OK, I wasn't going after "data science" specifically, but ad algos and how my twitter feed has become a cesspool of mental masturbation for ad algorithm people, which is the "killer app" for "big data":

    1. PhD invents fantastic ad algo. 2. Guy sees ad on iPhone, takes EBT card there. 1. PhD applies for EBT card. #OneTwoPunch []

    Wow, I can't believe I just typed that many buzzwords in a Slashdot post; but at least I had a reason and put most of them in quotes... dammit. "algos". Anyway, I wonder if ever

  • I'd think that'd be a red flag on a resume. (Sorry autodidacts.)

    • by geekoid ( 135745 )

      Why? Calling your self a Doctor without one would be, but not scientist.

    • You know that in astronomy several amateurs have made major discoverys? And a Scientist is just an engineer who lacks thumbs :-)
  • by rockmuelle ( 575982 ) on Wednesday March 05, 2014 @01:48PM (#46410175)

    I've been working with big data since before it was a term and currently run a scientific software company that touches on many aspects of "data science". Many of my colleagues also work in the field. I've seen many fads come and go. Data Science as a profession is one of those.

    Most people who call themselves data scientists are really just doing "big data" processing using tools such as Hadoop. They are delivering results to managers who have jumped on the big data band wagon and, not knowing any better, have asked for these skills. In 99% of the cases, the processing is simply haphazardly looking for patterns or running basic statistics on data that really isn't that big. However, there is a lot of low hanging fruit in data that hasn't been analyzed before and most practitioners who've suddenly become data analysis experts are rewarded for trivial findings. A tiny bit of statistics, programming, and data presentation skills go a long way.

    Compare this to the Web Masters of the late 1990s. The Web was new and managers knew that they needed Web sites. HTML and CGI were techie things but also fairly easy to learn. A group of people quickly figured out that they could be very important to a company by doing very little work and created the position of Web Master. A tiny bit of programming, sys admin, and design skills went a long way.

    Web Masters disappeared when IT departments realized that you actually needed real software developers, real designers, and real sys admins to run a corporate Web site. Sure, the bar is still low, but expertise beyond a 'For Dummies' book is still needed. And, few people can be experts in each area, hence the need for teams.

    Real data science has actually been around for a long time. Statisticians and data analysts have been performing this role for decades and have built up a lot of rigor around it. It a tough skill set to develop, but a very useful one to have. "Big Data" distracted people a bit and let the current generation of data scientists jump in and pretend everything was new and we could throw out the old methods. As the field evolves, data science will necessarily transition back to the experts (statisticians) and become a team effort that includes people skilled in programming, IT, and the target domain (analysts).

    That said, there's good money to be made right now, so if you have Web Master on your resume, you might as well be a data scientist while you can. ;)


    • Funny how this simple "webmaster" job seems so flipping hard. I have just started a new gig at an agency and after only 3 days I have looked at two major companies websites and found such major howlers that I wonder how there development/webmasters can mange to walk and chew gum at the same time.
  • Because they're expanding Data Science.

  • There's no question that there's a certain amount of hype around 'big data' / 'data science' at the moment, and with that comes a lot of "me too!" people. If the argument is that there's a lot of people just jumping on the bandwagon saying they can do 'data science' then I'll give the author that... but the suggestion that 'data science is dead' seems a bit hyperbolic to say the least.

    I've worked with a lot of top-notch people that would likely be labeled as 'data scientists' and I can tell you that:
  • Our current-generation workhorse instruments here at the telescope [] spit out tens of gigabytes per night as it is. The new camera we've been commissioning produces something like two gigabytes per exposure. And oh, yes, that data has to be archived, reduced, analyzed, etc., using things like IRAF or IDL. (Not my job.)

  • What's the difference between a data scientist and a statistician? A data scientist is a statistician who lives in San Francisco. Credit: []
  • Data Science is not "dead", but I've blogged my response: The End of Data Science as We Know It []
  • Data Science will be around for a while yet, but the truth is most organizations that hire expensive data scientists will collect data, analyze it, and proceed to ignore any recommendations and proceed on gut instinct, or overcorrect retroactively, rather than making decisions based on the analyzed data. Sometimes this works, sometimes it doesn't. Depends on whose gut instincts are being followed.
  • A year ago I argued that the whole "Big Data" thing was just a buzzword to make DBAs feel better about themselves. Basically big data is when you use statistics 101 and some halfway decent modern computers to do what you should have been doing all along. Some Big Data sales blowhards would use terms like ML which usually turned out to be just statistics 201.

    But after saying this I had a bunch of people jump all over me screaming that I didn't have a clue. About the only intelligent comment was someone as

e-credibility: the non-guaranteeable likelihood that the electronic data you're seeing is genuine rather than somebody's made-up crap. - Karl Lehenbauer