Mommy, where do data scientists come from?
Now that I’m with IBM, I have huge amounts of “food for thought” almost on a daily basis
After every meeting, every conversation, every book or article I read, I can’t help but think about the challenges every industry face.
First and foremost, I would like to highlight 2 points I really love about IBM
- For IBM, Big Data analytics is among the 3 main areas to sustain its business strategy in the upcoming decade (cloud and engage systems are the other two). The commitment is huge and serious; Ginni Rometty (CEO of IBM) said data is the world's natural resource for the next century and with it, industries will be completely transform. I share this vision and I’m really happy to be part of the process.
- IBM knows exactly where AI is heading. Actually, IBM does not talk about AI, but cognitive computing. The geniuses behind Watson know it will be daunting to recreate the human brain. It will not happen in our lifetime, nor our kid’s. So what is Watson doing? Co-Evoluting along us, humans. If you read a bit, you will see that in a selected group of hospitals, Watson is helping doctors to make more accurate diagnose. Watson is also partnering with chefs and creating new dishes. At the end of the day, IBM is not trying to make a machine that thinks by itself, but a machine that will help us to make better decisions.
Coming back to our conversation.
Where does a professional that does not exist will come from?
First, we need to understand that the people skilled enough to deal with huge volumes of unstructured data is almost nonexistent (as an example, just take a look at Netflix career page).
We do have teams doing business intelligence with good old structure data. And that is tremendous valuable, but what about everything else? (call logs, mailing, social media, nfc payments, retail and a long list). Who will tap on that gold mine?
The demand for data scientist is being created as I write this, but the first problem will arise when the supply falls short. Where does this data scientist will come from? Who will come when the upper management screams: “Get me 50 data scientist NOW!”
Let’s make another pause.
I have been reading for the last 2 days the first book of The Guardian Data Team, “Facts are Sacred” and at some point, the book highlights the fact of how important it is to have journalists trained to use excel (among other tools) in order to mine huge sets of data to support the stories they write.
This professionals will have a competitive advantage among its peers. If this journalist also knows basic statistics and most importantly, knows what to ask and separate dirt from value, then we have a potential data scientist. And its bachelor is in journalism. Not in statistics, economics or computing science.
Coming back to the question we had before. Where does this data scientist will come from?
I predict that within 10 years, the software use to data mine (along other technology that will come along, such as analysis of speech or pics analysis) will be easier to use. Moore’s law will also come into play and the prices will decrease. Meaning, more professionals should be in position of analysing data. And also medium and small size companies, the true engine of worldwide economy. When then get in the data bandwagon, IBM’s vision of disrupting industries will start to unveil fully.
I also predict that only a few Universities will prepare the kids to learn how to analysis data. Let’s be honest, Universities are not even teaching basic code. How come they will start teaching data analysis to doctors, architects or psychologist? Those trailblazers will be self taught. They will attend Coursera, Udemy or our IBM Big Data University. We are talking about people that can think medium term and will get that this skill will help them tremendously.
So: easier and cheaper software + a bunch of different self train professionals SHOULD supply the increasing demand for data scientist.
This big data scenario might look familiar to the social media craze we had years ago. Now, social media is a must, but a lot of people doing it are mostly under trained and (i would say) sloppy. That could (or not) happen with data analysis.
This is the best moment in time to jump fully into the big data bus. Get familiar with json, with nosql, with Watson, SPSS, Cognos, ilog and many other tools. This is the moment to leave the sheep behind (yes, once again) and be a trailblazer.