An algorithms tour

A year ago, I never would have thought I would be relying on a staff of eighty data scientists to dress me. But, I do.

A year ago, in preparation for attending a conference, for example, I would’ve been trying on a stack of clothes in a dressing room and relying on a total stranger to tell me if what I had on looked good and fit well.

Today, many of the outfits I wear are personally chosen for me and delivered in a box –- 5 pieces at a time –- based on my dimensions, my style preferences, and collaborative filtering algorithms. There are actually 85 data points captured.

The clothing box service I subscribe to uses these algorithms not just on my preferences but also for warehouse assignment, cost calculation, and inventory. Trained neural networks are used to describe pictures on my Pinterest board. Natural language processing is used to score items based on my textual feedback. After all the calculations, a request gets routed to a human stylist and pick the five things they will send me and write me a short note. I can contact my personal stylist if I want to. It’s very personalized. It’s unified with other areas of my online presence. Its mobile app is simple – I just have to click keep or return.

THIS is thinking differently about the shopping experience.  Check out the tech here.

If you want to flip HR on its head, you have to think differently about the experience we provide workers. You need to think differently about the learning experience.

How we ended up in an era of intentional ignorance

Here’s a six-month old essay about online media that resonated with me for a couple of reasons. Firstly, because I was an early-ish analyst blogger in the learning space (2006) and blogged the same way the writer did back then (quick commentary vs in-depth analysis) and secondly, it’s a sad and accurate story of how we got to where we are today – from WordPress-powered “Hello World” reflective blogs to memes and to trolls on Reddit – basically, a whole bunch of garbage. Perhaps I’m just part of the noise. Definitely.

“…instead of this promised blossoming of the modern mind, instead of education in an instant, we snapped our fingers and got entertainment, a medium without prerequisites, perplexity, and exposition.”

Big Wide Logic | Internet Stone Soup | July 10, 2017

If you didn’t see that job ad, maybe you’re too old (age discrimination on digital platforms)

PROPUBLICA and The New York Times co-published an article about several companies that are recruiting applicants for jobs using targeted ads that exclude particular age groups. I’m not talking about millennials. At a time where companies look to remove unconscious hiring bias by utilizing talent acquisition software, this is alarming.  An employment lawyer calls it blatantly unlawful. Facebook defends it as an accepted industry practice. And it’s not just them. We’ll have to wait and see how the legal challenges pan out. Meantime, you want to remove your age from social media platforms. You’ll lose some cake emojis once a year but oh well.

PROPUBLICA | Dozens of Companies Are Using Facebook to Exclude Older Workers From Job Ads | December 20, 2017

What does it take to become a Data Scientist?

Fun facts: Data Scientist is listed as the #1 job to have in the US in 2017, according to Glassdoor. The median base salary is $110,000.  There will be about 2,270,000 data science positions in the US by 2020. There is a need for female data scientists!

This article breaks down the data from 1,001 LinkedIn profiles of people working as Data Scientists and gives you a game plan for getting your foot in the door. Note that 40% of data scientists have attached online courses to their resume so that might be a good place to start since it’s such a new field. It’s also a good to know how to code in R or Python.

SmartDataCollective | Here’s the Data on How You Can Become One of 2.7 Million Data Scientists by 2020 | December 21, 2017

Creation spaces are a key vehicle to accelerate learning

John Hagel writes about coming together in creation spaces – “a cell, a small group of people, typically somewhere between 3 to 15 people, who come together on a very frequent basis and share a common goal to have an increasing impact in some domain.” These small groups, he notes, connect to expanding networks with no limit to scaling potential. The cells, he goes on to say, embody the levels of a knowledge pyramid: skills, knowledge, capabilities and passion. On the power of passion: The best place to start is by cultivating capabilities, especially curiosity and emotional and social intelligence. “Those capabilities will help you to explore an expanding array of domains until you finally find the domain that draws out your passion.”

Edge Perspectives with John Hagel | The Hidden Dimension of the Learning Pyramid | December 20, 2017

Gallup’s State of the Global Workplace

Gallup has released its State of the Global Workplace report. You can access the executive summary by supplying your contact information. The full report is available for purchase. Here are a few interesting points:

  • Businesses that orient performance management systems around basic human needs for psychological engagement, such as positive workplace relationships, frequent recognition, ongoing performance conversations and opportunities for personal development, get the most out of their employees
  • Higher levels of autonomy promote the development and implementation of new ideas as employees feel empowered to pursue entrepreneurial goals that benefit the organization — that is, to be “intrapreneurs”
  • At 31%, employees in the U.S./Canada are more likely to be engaged than are those in any other region, though there remains ample room for improvement
  • As employers in the U.S./Canada focus on organic growth and competition for talent, they aim to engage employees who feel optimistic about their ability to find a new job and who are seeking better career and development opportunities. To retain their workforces, organizations in North America need to consider how they are creating a future that employees want to be a part of, and how they can help employees feel innovative, fulfilled and optimized.

Gallup News | Good Jobs, Great Workplaces Change the World | December 19, 2017

The shift in how software and hardware is being built

I was pulled into this article when I read “a gigantic shift in computing is about to dawn upon us.” The gist – the US and China are investing heavily in designing high-powered AI chips to handle the linear algebra computations used in AI and this represents a fundamental change in how we build software and hardware. There are some examples of how these chips are being used – for face recognition, map street view, chatbots, and self-driving cars to name a few. Of course data is needed to train AI algorithms and that has come from private companies. Top AI talent will also come from those companies too ( I wish I paid attention to algebra in school). Worth a read is the link to China’s strategy and agenda for “intelligentization.” All that said, in my opinion anyway, the most interesting aspect of all of this is going to be ethics: robots rights, threats to privacy, discrimination, moral considerations, etc.

O’Reilly On Our Radar | The artificial intelligence computing stack | December 20, 2017

Thinking of implementing AI technologies?

Here’s a great (kinda wonky) article by Ben Lorica on O’Reilly’s “On Our Radar.” Ben shares slides from a recent presentation, offering an overview of the state of adoption of AI and suggestions to companies interested in implementing AI technologies. Ben also shares a sketch of a typical tech stack for intelligent applications.

Notable is a recent survey of 3,000 executives, managers, and analysts conducted by MIT Sloan Management Review that suggests low adoption (54% have not started adopting AI technologies). The author recommends the following:

Educate yourself on the current state of tools and technologies, identifying use cases in your domain or industry starting with small pilot projects.

Maybe start with this article.

O’Reilly On Our Radar | The state of AI adoption | December 18, 2017

Is competency more important than credentials? A warning against degree inflation

This US-focused HBS article discusses “degree inflation” – the practice of employers demanding bachelor’s degrees for job that don’t require them.

Because the pool of graduates is limited, the author explains, this practice can cause a misalignment between supply and demand, especially for middle-skills positions.  Complicating matters are automated hiring tools often exclude applicants without college degrees.

There are some great stats based on a review of 26 million job postings and a survey of 600 HR executives:

  • Only about one-third of the US population have earned a four-year degree
  • For typical middle skills job titles, sixty-seven of job postings required a bachelor’s degree or higher; yet just 16% of workers in those jobs held such a degree

The author suggests organizations invest in work-based learning opportunities, co-op programs, or paid apprenticeships. The article gives one example from a healthcare company that reviewed all job descriptions to identify skills associated with each position. The author closes with this:

“Competency is more important than credentials. Degree inflation is not just hurting individual workers; it undermines American competitiveness. American companies can’t let that happen.”

Harvard Business School, Working Knowledge | Why Employers Must Stop Requiring College Degrees For Middle-Skill Jobs | December 18, 2017

AI: Improving Workflow Through Worker Behaviors

The last paragraph of this article by Gigaom offers good advice about AI: look for solution providers that are using AI to help improve workflows based on workers’ behaviors. In my experience, this often means looking beyond solutions that are ready-made for learning. 

Gigaom also identifies four “tectonic shifts” taking place that are driven by the need to provide “greater personalization and efficiency in how we use technology.” The first shift listed is the changing behavior and expectations of Millennials and Gen Z employees. Blah, blah, blah… enough generational stuff! (It’s always been a pet peeve.) Newsflash: The desire for a “consumer-level technologies” that allow for a “quick, efficient, and intuitive” experience is something most workers expect (note I didn’t say employees) regardless of age. Even crotchety old Gen X’ers like me want an online experience that helps us accomplish our work quickly, efficiently, and more intuitively.

The second shift listed is digital transformation, defined in the article as a modernization of business activities, processes and models to become completely digitized. Again, not really a tectonic shift as much as evolution – although a very difficult shift to make.  

The third shift is about the trend toward technology procurement and a change in how decisions are made – at the unit or team-level vs. the C-level. I see this more as learning becomes more about the work.

The fourth shift is cloud-based technology training challenges associated with rapid change and how AI may help solve this challenge. For me that means thinking about training differently. AI will be big deal in 2018.

Gigaom | How Artificial Intelligence Will Personalize How We Work | December 14, 2017

Past posts have been archived


If you’ve arrived at this site by following a link and find it missing, please note that many dated, past posts have been archived. I started this blog in February of 2007 and wrote on the subject of corporate learning and development. The blog has been mostly inactive for the past few years because my work-related writing is published elsewhere and on Twitter. If you have any inquiries into past posts, please contact me.