Subscribe to RichardOnData here: / @richardondata
In this video, I discuss an idea that has been circulating for a while: will data science die in 5 years? (Or 10 years?)
There are probably three main reasons which are causing this idea to circulate around. They are:
1) Automation. Tools like AutoML from Google and Azure Machine Learning Studio are getting more powerful.
2) Hype theory. There is speculation we have hit a peak of inflated expectations and are descending to a "trough of disillusionment".
3) Companies failing to become data-driven cultures and achieve their intended ROI from analytics.
It is important when we have this discussion though, to keep a few things in mind.
1) Applied statistics and probability theory have been around since the 17th and 18th centuries. Hence people have been using these to make business decisions for decades. Data science was a term that picked up steam around 2010 due to the skyrocketing of computational power.
2) Data is being collected and generated at exponential rates. That's just an incredible tool that businesses will not stop leveraging.
3) Certain organizations DO use data science and AI to great success, and they provide blueprints for organizational success to others.
However there is currently some misleading expectations for data scientists. This is because of academia and the media excessively hyping ML and AI excessively. Additionally, organizations occasionally fail to support their data scientists with proper infrastructure, organization, or data. This can lead also to unrealistic and misleading expectations for data scientists. These problems are related.
I think we will see a couple solutions in the future. People will get a bit more specific with "Data Science", leading I think to more specialization. Ultimately successful companies will learn from others -- not all, but some will. Meanwhile, the rise of AI will enable automation of menial tasks and processes, allowing data science workers in this decade to solve the next, newest, greatest problems.
Links:
"Data Science is Dead" - Alexs Thompson: / data-science-dead-alex...
"Radical Change is Coming to Data Science Jobs" - Nate Oostendorp: www.forbes.com...
"Google's AutoML: Cutting Through the Hype" - Rachel Thomas: www.fast.ai/20...
"Data science is different now." - Vicki Boykis: veekaybee.githu...
"Still Searching for ROI in Big Data Analytics? You're Not Alone!" - Salman Khan: www.kdnuggets....
"Companies Are Failing in Their Efforts to Become Data-Driven" - Randy Bean and Thomas H. Davenport: hbr.org/2019/0...
"A Very Short History of Data Science" - Gil Press: www.forbes.com...
"The AI Hierarchy of Needs" - @mrogati: hackernoon.com...
"Why There Will Be No Data Science Jobs Titles By 2029" - Noah Gift: www.forbes.com...
Негізгі бет Will Data Science Die in 5 Years??
Пікірлер: 127