Skills for Data Science and Machine Learning: The Video

Here we look at the skills and employment opportunities that are surfacing as we move into an era of advanced analytics and ubiquitous AI.

The videos are a mixture of seminars and discussions avaialable under standard YouTube license, with landing page descriptions added. Minor edits have been made where necessary.

What Kaggle has Learned from Almost a Million Data Scientists (O’Reilly

Published on May 25, 2017

A talk by Anthony Goldbloom from Kaggle.  Kaggle is a community of almost a million data scientists, who have built more than two million machine-learning models while participating in Kaggle competitions. Data scientists come to Kaggle to learn, collaborate, and develop the state of the art in machine learning. Anthony Goldbloom shares lessons learned from top performers in the Kaggle community and explores the types of machine-learning techniques typically used, some of the tricks he’s seen, and pitfalls to avoid. Along the way, Anthony discusses work habits and skills that help data scientists succeed.

Andrew Ng: Artificial Intelligence is the New Electricity (Stanford Graduate School of Business)

Published on Feb 2, 2017

On Wednesday, January 25, 2017, Baidu chief scientist, Coursera co-founder, and Stanford adjunct professor Andrew Ng spoke at the Stanford MSx Future Forum. The Future Forum is a discussion series that explores the trends that are changing the future. During his talk, Professor Ng discussed how artificial intelligence (AI) is transforming industry after industry.

What Are the Top Skills Needed to Be a Data Scientist? (SAS Software)

Published on Oct 26, 2015

Dr. Goutam Chakraborty of Oklahoma State University outlines the top skills you need to be successful as a data scientist. From SAS Academy for Data Science

Top Data Scientist D J Patil’s Tips to Build a Career in Data (FactorDaily)

Published on Jul 20, 2017

D J Patil, who along with top data scientist Jeff Hammerbacher coined the term “data scientist” tells you what it takes to be a data scientist During the Obama administration, Patil became the first ever chief data scientist to be appointed by the US government.




Guest Blog: Machines Are Taking All The Jobs? What Decision-Makers Say And Do

Original blog link:

Machines Are Taking All The Jobs? What Decision-Makers Say And Do

by Gil Press

A new PwC survey provides fresh and illuminating data on the burning questions of the day: Are machines going to take over our jobs? And how much do we rely (or over-rely) today on machines, automation, and algorithms?

Experts are confident that machines are going to replace many workers. A much-quoted report from Oxford University has estimated that “about 47% of total US employment is at risk” for being fully automated. The machine threat to employment is even greater in developing economies—a recent report from Oxford estimates that 77% of jobs in China and 69% of jobs in India are “at high risk of automation.”

But maybe estimating the type of jobs that the machines are going to replace is the wrong approach. Tom Davenport, who just published a book on strategies for coping with automation, Only Humans Need Apply: Winners and Losers in the Age of Smart Machines (co-authored with Julia Kirby), told the Wall Street Journal recently: “Computers don’t tend to replace whole jobs; they replace specific tasks.”

The McKinsey Global Institute (MGI) agrees: “…a focus on occupations is misleading. Very few occupations will be automated in their entirety in the near or medium term. Rather, certain activities are more likely to be automated, requiring entire business processes to be transformed, and jobs performed by people to be redefined.”

MGI estimates that 45% of work-related tasks can be automated. This finding does not bode well for knowledge workers who were sure their cognitive skills could not be automated and that they will always outrun the machines. Even CEOs, according to MGI, spend over 20% of their time on activities that can be automated with current technology.

What has been missing in this discussion is data on how much we rely (or don’t) on machines today, rather than estimates based on experts’ assessments of how automation-prone are various occupations and activities. Specifically, has the era of big data and increasingly sophisticated algorithms changed the nature of business decision-making? What is the extent by which business executives rely on machines today when they make strategic decisions?

A new PwC survey of more than 2,100 business decision-makers across more than 10 countries and 15 industries sheds new light on these questions. It frames the discussion as follows: “Executives who once relied firmly on their intuition and experience are now face-to-face with machines that can learn from massive amounts of data and inform decisions like never before.”

59% of the decision-makers PwC surveyed say that the analysis they require relies primarily on human judgment rather than machine algorithms. That means that 41% already tend to rely more on algorithms than their own experience, judgment, and intuition. “We are not talking about pricing a seat on an airline,” says (via email) Dan DiFilippo, Global & US Data and Analytics Leader at PwC. “We are talking about big, strategic decisions that almost certainly involve some combination of human and machine, but clearly we see a significant involvement of the machine.”

The most interesting findings are about the type of decisions that tend to be assisted by machine algorithms and the ones that rely more on human judgement. In the chart above, “respondents who answered closest to zero are nearest to the survey’s overall average reliance on analysis from machine algorithms and human judgment. The farther away from the center point, the greater reliance on either mind or machine,” says PwC.

“Shrinking existing business” was deemed by survey respondents as the type of decision that relies most on human judgement and “Investment in IT” as the one relying most on algorithms. “Investment in IT,” says DiFilippo, “can cover many areas including shop floor automation, CRM systems, HR systems, risk management systems, etc., all of which have varying degrees of machine algorithms and can be assessed by machine algorithms.”

The breakdown of results by country offers a striking juxtaposition of China and Japan with the former as the country/region relying more than others on machine algorithms and the latter as the country/region second only to Central and Eastern Europe in its reliance on human judgement. One would think that China and Japan will have similar attitudes toward and use of algorithms in decision-making but this is apparently not the case. It’s possible, however, that the results are due to different interpretations of the survey questions. Says DiFlippo: “We don’t have a precise answer or explanation for this—we are still working to gather more on this front.”

Finally, the breakdown of results by industry shows that different economic sectors differ in the degree by which decision makers rely on their own judgement vs. relying on machine algorithms. Conclude DiFilippo: “Involving the machine can help reduce/eliminate bias (at the individual, department or organization level), add more accuracy and/or more computing power to crank through a high volume of scenarios that human can’t do (or can’t do in a timely manner), and importantly—and the data supports this—there is a sense that the machine can help de-risk the strategic decision… we see that those who had a high degree of machine algorithms felt a high degree of managed and known risks.”

So should we search for the right mix of minds and machines in the context of a specific decision or should we succumb to a universal McAfee’s Law and agree that “as the amount of data goes up, the importance of human judgment should go down”? What’s your experience with trusting machine algorithms rather than your own judgment?

Gil Press

Gil Press is Managing Partner at gPress, a marketing, publishing, research and education consultancy. Previously, he held senior marketing and research management positions at NORC, DEC and EMC. Hewas Senior Director, Thought Leadership Marketing at EMC, where he launched the Big Data conversation with the “How Much Information?” study (2000 with UC Berkeley) and the Digital Universe study (2007 with IDC). He blogs at and Twitter: @GilPress

Guest Blog: Only Humans Need Apply Is A Must-Read On AI For Facebook Executives

Original blog link:

Only Humans Need Apply Is A Must-Read On AI For Facebook Executives

by Gil Press

Under pressure to remove alleged human bias from its “Trending Topics” section, in August. Facebook fired the editors who were selecting and writing headlines for the stories, explaining that this “will make the product more automated.” The results of trusting algorithms more than humans have continued to make headlines ever since with the Trending “product” promoting a fake news story about Fox News’ Meghan Kelly, a conspiracy article claiming the 9/11 twin towers collapsed because of “controlled demolition,” and Apple AAPL +0.78%’s Tim Cook announcing that Siri will physically come out of the phone and do all the household chores (a story from an Indian satirical website, Faking News, that was Trending’s top story on the day of the iPhone 7 launch event), to mention just a few of the more embarrassing machine failures.

Silicon Valley has never displayed much love for fallible humans, but has shown a lot of confidence in the continuous improvement and now, self-improvement, of machines. Do humans still have an important role to play in our automated lives which are increasingly controlled by sophisticated algorithms and seemingly smarter machines?

In Only Humans Need Apply: Winners and Losers in the Age of Smart Machines, knowledge work and analytics expert Tom Davenport and Julia Kirby, a contributing editor for the Harvard Business Review, offer optimistic, upbeat and practical answers to this much-debated question. “The upside potential of the advancing technology is the promise of augmentation—in which humans and computers combine their strengths to achieve more favorable outcomes than either could do alone,” they write.

There is not much difference, contend Davenport and Kirby, between technologies of automation and technologies of augmentation. The difference lies in the goals and attitudes behind the application of these technologies. Automation is unidirectional and focuses “primarily or exclusively on cost reduction” via the elimination of human labor. In contrast, “augmentation approaches tend to be more likely to achieve value and innovation” and they are bidirectional, making “humans more capable of what they are good at” and “machines even better at what they do.”

It is a shortsighted (and short-term) strategy for companies to favor automation over augmentation: “If the goal is to provide truly exceptional or differentiated products and services at scale, only an augmentation arrangement can accomplish that,” write Davenport and Kirby. They advocate a “workplace that combine sophisticated machines and humans in partnerships of mutual augmentation” and mutual benefit.

Gil Press

Gil Press is Managing Partner at gPress, a marketing, publishing, research and education consultancy. Previously, he held senior marketing and research management positions at NORC, DEC and EMC. Hewas Senior Director, Thought Leadership Marketing at EMC, where he launched the Big Data conversation with the “How Much Information?” study (2000 with UC Berkeley) and the Digital Universe study (2007 with IDC). He blogs at and Twitter: @GilPress

Tremors in the Workplace: the Dawn of a New HR

Human Resources technology has come a long way in the past several years. Without a lot of fanfare, it has been subject to many of the trends that are active throughout today’s enterprise IT. There is change based around use of analytics, around mobility in new roles; in social media usage, and in use of cloud solutions to democratize access and act as a centralization point for mobility.

HR activities and systems are shifting from maintaining records to active management of skills deployment and development, and enabling continuous monitoring of job performance. There are key changes afoot as we move into this new era; they include the end of rigid performance reviews; new opportunities for training, such as MOOC’s and inexpensive online courses; and, overall, much greater opportunity for employee input and engagement in innovation and performance improvement.

Traditional HR software is being challenged by startups from outside of HR. Of particular interest is the intersection with analytics. A significant area of development is in employment engagement and feedback. Here we see companies such as TINYPulse, Glint, CultureIQ, Culture Amp and others. Other vendors such as Humanyze are bringing location monitoring to this sector. Korn Ferry provides a recruitment tool that generates profiles for open executive positions based on benchmark data from people who have performed in that role. In social media, LinkedIn offers a tool providing insight into social activity of candidates to understand who they may know, and how they are interacting.

Opportunities are appearing for

  • Mining of social media to determine suitability of candidates;
  • Analysis of needed positions and job requirements gained from mining internal data;
  • Process change based on real-time monitoring of employee actions;
  • Creating a more dynamic and plastic environment in which HR technology is a component of resilience and agility
  • Permitting companies to respond swiftly and optimize outcomes across all activities.

As HR technology continues to evolve there will be a significant impact on the workplace. Within an environment of rapidly shifting skills and job descriptions, there is a need for greater engagement with employees and for more attention to the specifics of everyday roles.

There are also downsides to this technology. Employees may rebel against over-monitoring and over-direction, for example. On the whole, however, next generation employees are likely to be more comfortable with the trade-offs between privacy and opportunity. Still, there is a risk of reducing flexibility by forcing employees to think constantly about their action. Answers might come from gamification to motivate and ease control issues.

Assuming, that we have learned the lessons of early 20th century “Scientific Management,” there is great promise in these technologies creating a better integration between employees and the evolving IT and robotics environment.

As we move into an era of continuous AI and Big Data analysis, it is important to understand that behavioral expectations for employees will change. These changes will enable us to think faster, react more flexibly, and perform more comfortably within a technology environment. But the effects upon culture and thought processes are yet to be determined.