All posts by Elbert Chu

is CTO and Innovation Editor at Wired Academic. His work has appeared in: ESPN, Fast Company, People Magazine, and World Magazine. He covers people and ideas in business, design, technology, religion, sustainability, and philanthropy— especially where they intersect. He always welcomes a good lead.

LearnUp: eLearning Startup’s Potential Cure for Skills Gap and U.S. Unemployment

Image via thenextweb


Alexis Ringwald spent six months in the unemployment lines of Silicon Valley. She realized a core problem was that unemployment offices trained people with skills they thought would help them get jobs. But applicants sent resumes and never knew why they were never hired. That’s when Ringwald realized the crucial stakeholder was missing in the training process — who knew better what skills applicants need than the companies themselves?

“The skills gap is an information gap,” Ringwald said. “The current system is broken. Our education system is not adapting quickly enough to keep up with what the labor market needs.”

So Ringwald and co-founder Kenny Ma started LearnUp. Here’s what you need to know:

  • Pilot program in San Francisco Bay area, started in 2011
  • Clinton Global Initiative goal is to give 100,000 young job seekers access to online job training
  • Employers offering training: Staples, Gap, Safeway, Wholefoods, Starbucks, Banana Republic, Wells Fargo, KPMG, and Ringwald says they have a wait list.
  • Rolling out training (most run 2 hours) to partners Corporate Voices and California Community Colleges‘ 3 million students
  • Training videos free to applicants, revenue comes from employers for successful placements, and perhaps more advanced features.
  • $1.9 million in funding. VC funded by Floodgate (Mike Maples), Andreessen Horowitz, Greylock (Reid Hoffman), NEA, SV Angel (Ron Conway), TechFellows Fund, and other angels.
  • Competitor skilltesting provides more technical competency testing of potential hires.

Bottom line: It’s nice to see that LearnUp has partnered with California Community Colleges, but it also begs the question, what is the value proposition for the community colleges? If employers can directly train future employees, this is the further erosion of community colleges. For sure, LearnUp mostly meets the need of entry level positions right now. But is there any reason why the platform doesn’t scale up to include more technical or advanced positions? Starbucks and Staples both list manage positions on LearnUp.

From Forbes:

Companies like LearnUp have partnered with big box employers to develop curricula that allow someone with an undifferentiated or out-of-date degree, to learn the relevant skills necessary for a job, and to signal intent in a crowded marketplace. Students can turn what could be a loose proxy for smarts and responsibility, say for example, an Associates degree in English, into a strong signal and skills-based orientation that tells the retailer that the individual has not only the talent, but intention, to be a floor-manager at a big box retailer.

From Wired:

Here’s how LearnUp works: A prospective job applicant can look at an employer’s open jobs, each of which has a set of recommended “trainings,” which can include videos, written tutorials, magazine articles, and links to classes and certifications offered outside of LearnUp. Staples’ trainings for its manager position include a YouTube tutorial on QuickBooks, a Harvard Business School “Working Knowledge” article on “Making the Move to General Manager,” a article on “Retail Store Management Tips,” and a TED talk on the “Eight Secrets of Success.”

As the prospective job applicant completes the trainings, she builds a “skills resume” on LearnUp she can use to apply for open positions.

Video cheat sheet:
Skip to 3 minutes for how Ringwald developed the platform idea.
Skip to 19 minutes for Ringwald on LearnUp’s growth trajectory.

[ Forbes: On Digital Literacy, Wired, Forbes: On LearnUp Launch, TNW]

Infographic: Three Factors that Drive Costs for E-Learning Course Development

Custom e-learning Development Team
In case you’re more of a visual or auditory learner (vs. the Excel jockey who prefers the itemized spreadsheet of MOOC startup e-learning costs), here’s a handy infographic accompanied by a 15 minute podcast, both produced by Toronto-based Scholarix, an e-learning consultancy. We say skip the podcast, because the infographic is much more concise and easily digestible. One thing that would really help add value is some actual numbers attached to these variables, even a rough range.

Three months of development for one hour of online coursework. In terms of numbers, that’s one point that came out of listening to the podcast, which was a helpful landmark. That estimate was for a middle of the road course on the three variables:

  1. Graphics/Multimedia
  2. Level of Interactivity
  3. Instructional Design


Custom elearning Costs Determining Factors


How edX Will Scale MOOCs with a Machine Learning Robot Grader


Bottom line: Machine learning will enable larger scale evaluation of student performance. Still in infancy, but will develop further. Question is where robo-grading will fit in guiding students to desired outcomes. That is, a student who just wants to audit and see if they have a ‘good enough’ understanding vs. someone seeking employment and/or certification for professional development.

On the positive: instant feedback, reduced person resources.

On the negative: assessment quality not as good, critiques not as nuanced

In practice: Requires training of 100 tests graded by humans to train the computer how to grade.

edRadar take: This is just the beginning. No one would have thought a few years ago that a computer could write believable news stories, but that’s exactly what Narrative Science is doing right now with Quill. And you can say “sayonara” those warm fuzzy days when humans were still beating computers at chess.

Next: Machine learning will enable grading algorithms to become smarter and more robust. We believe this will enable scale for online learning, so student outcomes could be tiered. Perhaps premium students pay for human critique, and edit process. On the other hand, corporate development and human resources would likely want the evaluation of domain experts to demonstrate expertise for raises or increased skill sets.


Dud: Five Reasons Why the Rosetta Stone Acquisition of Livemocha is a Fail



Rosetta Stone, arguably the largest language learning company gobbled up five year old Livemocha for $8.5 million in cash earlier this week. That’s a pittance compared to the reportedly $19 million Livemocha has raised in the past. There are a few reasons why it seems Livemocha went out like a garage sale.

LingoLive, a platform for Spanish learning, has a great round up of five reasons why nobody is a winner in this deal.

  1. Language learning is all about mobile, and Livemocha has no mobile app.
  2. Different business models: Livemocha is freemium, Rosetta is premium only. It will be interesting to see how Rosetta reconciles this massive difference.
  3. Rosetta Stone is trying to buy online cred with users to overcome a narrow sliver of a lucrative audience.
  4. Livemocha doesn’t help learners sort through the language learning noise.
  5. Livemocha and Rosetta don’t fix learner’s main pain point. [Insert LingoLive plug for newsletter] We signed up but didn’t receive it. But we suspect it has to do with live conversation, judging from LingoLive’s offering.

Still, that said, Rosetta [RST] investors seemed to like the move, which ran the stock up four percent on the day. But there was hardly any volume (no conviction from the herd), and on the week the stock closed roughly flat at $15.40.

Meanwhile, Arlington, VA-based Rosetta Stone also announced they would cut 16 percent of its employees and  shutter 56  remaining airport kiosks. Desperate times. But the broader course correction by management is making investors happy with the stock up over 50 percent YTD.

And online language learning is buzzing. Just last week, we saw Babbel‘s language learning platform score $10 million in Series B funding, with a grab for PlaySay.  And it’s starting to become a crowded room with Duolingo, Voxy, MindSnacks, Busuu, and LivingLanguage just to name a few. Perhaps they’ll have better exits when they reach their destinations.

[TechCrunch, edSurge, LingoLive, Fool, Skift]

Insider: 10 Things to Know About UPenn’s MOOC U. Strategy



Here’s an insider’s look at UPenn’s strategy and outlook on MOOCs and higher ed in a broader perspective. It’s a good long read well worth your time. No time? Here’s what you need to know:

1 Currently has 19 courses hosted on Coursera.

2 Keeps 6-15 percent of revenues from Coursera, plus 20 percent of profits.

3 Per course costs roughly $50,000 upfront investment of stipends for profs, TAs, and video production.

4 Double the effort to develop the first iteration of a course.

5 Revenue from course content flows to professor treated as intellectual property/patent. Additional revenue from licensing courses to other schools. (See #8)

6 Prof. Al Filreis pushing the format to make MOOCs more interactive with the goal of the feeling of a community of learners “doing it together.” (Modern and Contemporary American Poetry)

7 Oddly, the director of open courses there thinks of MOOCs as textbooks.

8 Both University of Washington (for hybrid classes) and Antioch University already grant college credit for UPenn Coursera classes.

9 Worries for community colleges, which are “institutions whose primary benefit to students is education.”

10 Banking on institutional prestige for “signal of quality,” networking, social environment to transition to adulthood. The idea that sum is worth more than it’s parts.

Key quotes:

+ Unbundling: “I think down the line you’re going to have instructors who re-mix content that’s prepared by potentially one, two, or three different instructors—or more—and create trajectories that make sense,” said Daphne Koller, founder of Coursera, sketching a future in which the professor becomes a sort of DJ. “That is what it’s about to be an instructor, I think. It’s a curation process.”

+ Profit vs. Ed? “My biggest fear, frankly, is not a fear connected to Penn at all,” says UPenn’s chief academic officer, Vincent Price, “It’s a fear that thinking right now about higher education, and especially public higher education, is driven by logics of efficiencies, concerns about the spiraling costs of education, et cetera. And that, too rapidly, these [MOOCs] will be seen as ways of bending the cost curve. And that efficiencies, real or imagined, will become a device for withdrawal of support from high-quality education, and replacement of that experience with something that’s perhaps adequate, but not outstanding. I’m very, very concerned with the misuse of these technologies in a way that is viewed as a cheap way out.

h/t @Kimie_B [The Pennsylvania Gazette]


Here’s how DARPA replaces teachers with artificial intelligence

Wall•E Silke Gerstenkorn via Compfight

What you need to know.
If you see DARPA coming for you, don’t run if they say they’re here for “Education Dominance.” They just want to teach you a thing or two. Pay attention: here’s what DARPA is doing in digital learning — artificial intelligence algorithm that’s here to teach you. Meet your next teacher, called the Digital Tutor.

From edSurge— Any report on digital learning technologies would be incomplete without a discussion of artificial intelligence, and the example offered from the Navy’s Information Technology Specialist training program is particularly potent. The Defense Advanced Research Projects Agency, or DARPA, went to work on a significant problem the Navy faced: it needed to train highly effective IT specialists who could solve complicated technology problems aboard ships deployed around the world.

To get IT Specialists to adequate levels of training previously required highly experience instructors, a significant amount of classroom time, and then several additional years of on-the-job training. The process was slow, expensive, and dependent on limited human resources.

So DARPA created a small ideal training program with elite instructors and a small cohort of trainees. The program underwent rapid-cycle testing and iteration (design-based research) until its new graduates could outperform experts with years of experience. DARPA then built a Digital Tutor, or DT, program that could mimic the abilities of the elite trainers in the live-instruction pilot program.


Op/Ed: Will the Real Internet Teaching Stars Please Stand Up?

TED x PARIS previewTED x PARIS preview (Photo credit: PhOtOnQuAnTiQuE)

This is a comment I started in response to Kevin Convey’s meditation on MOOCdom and Jeff Jarvis‘s question in response, “Does education become a hits business like show business?” It got a bit out of control, so I’m just making it a post.

Early indications show that online edu has already in some ways demonstrated the ability to amplify and reward rock star teachers, and why not? Could the internet inspire a generation of kids to aspire to be the Will Smiths of education? I think so. Continue reading

Number Crunch: Why K-12 EMO Virtual Schools Give Students a Good Reason to Cry

My EMO virtual school sucks. photo: flickr user nirazilla


File this report in this week’s must read. The main take away from authors of new research: Most education management organization [EMO] K-12 online schools do not meet federal No Child Left Behind performance standards defined as adequate yearly progress. But always follow the cash to put numbers in context.

EMOs are companies like K12, Inc. and Connections Academy that manage and largely account for the huge growth in charter schools. The term EMO was hatched by Wall Street because these education companies operated much like health management organizations [HMO].

We covered this National Education Policy Center research earlier this week, but thought it was important to additionally note a closer look at the report’s source itself.

We agree the industry needs to be scrutinized and the most effective schools identified. But to be fair, know that some of the largest funders of the National Education Policy Center are  Teacher’s unions— groups decidedly unfriendly towards online education. The Los Angeles Times wrote a critical piece about teachers and did some digging:

“…a major source of funds for the policy center is the Great Lakes Center for Education Research and Practice, a foundation set up by the National Education Association and six major Midwestern teacher unions affiliates. The NEA was one of the teacher union groups that backed an unsuccessful call for a boycott of The Times when “Grading the Teachers” was first published.

Another difficulty with this research is the adequate yearly progress [AYP] itself. Although researchers culled most of the data from state reports, sometimes the companies themselves updated the numbers themselves. The main problem is that standards and metrics swing widely from state to state.

Researchers note the AYP is a “crude indicator of whether schools are meeting state standards.” The problem is that EMOs that target lower income and low performance districts already have harder times hitting their adequate yearly progress marks. For comparison, an estimated 52 percent of all U.S. public and charter schools made adequate yearly progress.

Edweek has a good review of the current AYP landscape.

Beyond just online schools, for-profit EMOs have 394,000 students, over double the number from five years ago. Non-profit EMOs account for almost an equal number of students. Over 56 percent of non-profit EMO students made AYP compared to 48 percent in the for-profit group. Together, they amount to roughly 780,000 students across the nation. Performance in the over all group is roughly inline with traditional schools.


Read the full National Education Policy Center’s report.

via Students of Virtual Schools Are Lagging in Proficiency –

Five Promising Education Startups to Watch in 2012

Keep your eyes on these startups. Photo: Marie Linner, flickr user GORE-TEX® Products

With the new year under way, we thought this would be a good time to highlight some promising education startups.

General Assembly, a community of entrepreneurs, and Startl, a learning technology accelerator, hosted demos and Q&A with five education startups in a packed room of roughly 130 people earlier this week. The session was the seventh in the Digital Learning Series held at General Assembly in New York City.

Startups Codecademy, EasyBib, tutorspree, littleBits and Study Edge presented. Here’s the rundown of the companies in the order they presented and why you should pay attention:

Codecademy, Zach Sims
Codecademy teaches non-techies how to code in javascript with an uber-simple interface in snack-sized chunks. Sims said Codecademy started with teaching javascript because it was the low-hanging fruit and easy to implement inside the browser environment.

Codecademy recently launched Code Year, a one year program in 2012 to capture the New Year’s resolution crowd resolving to nerdify. It’s creating a huge boost to their already massive email list. So far, nearly 170,000 people will receive a lesson each day in their email. Since we all know how most New Year’s resolutions work out, we’ll check on retention rates later this year.

Why You Should Pay Attention:
For their next act, Sims said, “We’re building a platform so people can build their own curriculum.” This could prove useful for teachers and professors that want to leverage the Codecademy groove. In fact, they already have their first user generated course written by one of their investors at Union Square Ventures.

We think this platform can translate well for STEM teachers that want to initiate a new generation of programmers.

Codecademy is adding engineers to broaden from javascript to Ruby and Python in the near future.

EasyBib, Neal Taparia and Darshan Somashekar

Taparia and Someshekar started EasyBib because compiling bibliographies for school papers was such a pain. Their product aims to take the guess work out of how to reference research materials like websites versus books.

Technically, EasyBib is not a startup. Work began 10 years ago, but Taparia said they only recently realized the potential of their platform and went full time last year.

The EasyBib team says their mission is to teach information literacy. We see this as an effective tool to teach students to identify credible sources.

EasyBib’s business model is selling the product to schools. So far, customers include NYU, Ohio State and Emory.

Why You Should Pay Attention:
They already helped 30 million students produce over 450 million citations said Taparia. And they continue to grow 30-40 percent each year.

An interesting fruit of their product is the social effect. Students can research to find new related sources other users have generated with what Taparia called their beta “social research engine.”

tutorspree, Aaron Harris

We recently covered tutorspree, so we we’ll report what’s new here.

Tutorspree is focused on cornering the market on tutors. Harris wrote to us via email, “The first step has to be to create the best base of tutors around. The technology layered on top is just a tool used by tutors and students.” So we found it very interesting that in addition to the thousands of tutors they have captured already, Harris said tutorspree has a wait list of over 8,000 tutors who want in.

Another interesting number Harris provided is tutorspree tutors on average charge $40 an hour. Based on tutorspree estimates, school teachers earn $20 an hour. Tutorspree recently released a report based on data collected from 2011.

Why You Should Pay Attention:
Part of the brave new world of education is how technology extends access to students who couldn’t otherwise afford tutors or an education. Bharani Rajakumar of, an education data analytics startup, asked Harris how tutorspree will help make learning more accessible to people.
Harris said they eventually plan to work with nonprofits and the government. “The plan is to provide a product for them to track performance,” said Harris. “The government spends $700 million tutoring low income students and right now nobody has any idea how they are doing.

We’re curious to see how tutorspree will help provide assessment, considering their existing review metrics of tutors are rather subjective. Harris did say they are constantly changing the questions they use to elicit reviews.

littleBits, Ayah Bdeir and Paul Rothman

littleBits (re)intro from ayah bdeir on Vimeo.

LittleBits designs and makes little smart circuit boards that connect with magnets. Each circuit board contains its own logic for modules like: light sensors, LED arrays, pressure sensors, fans, buzzers, knobs and switches. It’s Lego meets electrical engineering.

The slightly larger than postage stamp-sized modules snap together with a satisfying click when aligned correctly. LittleBits has designed it so the magnets repel each other when little engineers push the wrong sides together. MOMA liked the design so much they snagged a set for their own collection.

LittleBits was born in 2008 out of Bdeir’s frustration with soldering and intricate work needed to build engineered components. Bdeir, a certified engineering geek (read: masters from MIT Media Lab), wanted to make an easy way for people to play engineer.

The company launched their first production run of 250 starter sets in September 2011 said Bdeir. Now they’re half way into production of an order for 3,000.

For kids, we know play is the best kind of learning. Just what STEM education needs: a little bit of “geeky fun.” Bdeir said littleBits has plans to provide a platform for classes to teach electronics in schools.

Why You Should Pay Attention:
“We’re having informal discussions with teachers and afterschool teachers to have curriculum to incorporate for the platform,” said Bdeir.

LittleBits will be at the New York Hall of Science for teachers in the NY area this month. Bdeir hopes to put lessons teachers generate online, and more people can add to them. An interesting point Bdeir made is littleBits has experienced equal interest from boys and girls.

In a software driven education startup space, it’s refreshing to see something kids (and even adults) can touch and tinker around with.

In the meantime, the project is all open source. Bdeir said, “I’m a strong believer [in open source] to be able to have people build on each other and build smarter things.”

Study Edge, Ethan Fieldman

This is a company flying under the radar. We’re going to devote an entire post to Study Edge, so stay tuned. But in the meantime, let’s just say Study Edge is Khan Academy except in a sustainable (and profitable says Fieldman) business delivered via Facebook.

Why You Should Pay Attention:
Study Edge doesn’t have 30,000,000 viewers like Khan Academy. But guess what, each of Study Edge’s thousands of college students pay a monthly subscription fee ranging from $25 to $75 each month. What’s exciting is that Study Edge’s model is a mechanism to reward superstar teachers. Fieldman’s team plans to expand capacity beyond their initial market, where they’ve been “perfecting” the product.

Here’s a sample lesson:

Embedly Powered

What startups are on your must-watch list of education enablers? Let us know in the comments!

How a STEM Startup’s Virtual Labs Help Students Find Their Inner Mad Scientist

Late Nite Labs

Late Nite Labs provides interactive chemistry and biology labs online for budding mad scientists in high school and college. Originally based in New Jersey, the startup moved to new digs in New York City recently. Late Nite Labs is prepared to expand their own 10 person lab staff with new funding.

What is it?

The labs are plain fun. Students can mix whatever they want in any quantities they like. What happens when you add nine milliliters of yeast to 15 milliliters of glucose with a few grams of lead in a beaker and put the mixture in a dry ice bath? Not much. But that’s kind of the point too. These labs allow students to explore and make mistakes. Underlying algorithms and equations just crunch the numbers, and results appear in the data just like they would in the lab— right or wrong. Online lab manuals provide step-by-step instructions when students are ready to get their assignment completed.

“It’s immersive, like a first person shooter game,” said Harris Goodman,  chief development officer for Late Nite Labs. But instead of machine guns, students are armed with beakers and test tubes. One chem lab for college students figures out the alcohol content in vodka. How’s that for grabbing the attention of hungover students?

Good Economics for Schools, Better for Students?

We like this idea for student STEM engagement and particularly on the school admin side. Virtual labs are an economical replacement to reduce costs like glassware, instruments and space for chem and bio labs— not to mention insurance, lab certifications and safety issues. It also helps that Late Nite Labs is free for schools and professors to use.

Goodman said getting schools to fit new products into their already thin yearly budgets was onerous. Instead, students foot the $50 per class per semester bill for Late Nite Labs. Compare that to $20-50 per lab manual per class, plus $80-200 for equipment and materials labs fees.

For students— no more goggle rings around the eyes, clean up or nasty chemical burns. In Late Nite Labs, students just drag their mess of glass and instruments into a yellow “biohazard” bin and— poof, clean up is done. Plus, the lab is accessible anywhere in the world with an internet connection, at any hour. There’s also 24/7 support for students and teachers— with real live technically trained people from West Chester, NY.

Lost in Virtualization?

Still, we wonder if playing with chemistry and biology sets online misses some of tactile joy in what fascinates young scientists. Will young scientists miss their calling if they don’t experience the clinks of glass, the heat of Bunsen burners, squeezing the pipette just right or the acrid aroma of acid?

Two small test tubes held in spring clamps

Goodman easily acknowledges their shortcomings.

“We can’t teach a student exactly how to hold and squeeze a pipette and there’s a lot of chemical reactions that cause smells. This would be hard for someone who wants to make cologne or perfume,” said Goodman.

Good for Learning?

Smell and touch may be lost in virtualization, but what won’t get lost is the critical thinking about the world with scientific experiments. Goodman says hundreds of professors from 150 universities, colleges and high schools are using their labs. Schools include Arizona State, Penn State and Iowa State among others. That translates to 20,000 students who have gone through virtual labs.

Another benefit will be wider exposure for students to STEM who might not otherwise experience a wet lab because of cost or space constraints.

Late Nite Labs’ ambitious goal is to be a STEM platform that scales and is customizable.

Professors can provide entire assignments, grade, communicate and analyze student performance through its own LMS back bone.

Late Nite Labs provides 600 different chemicals and materials, from orange juice and hydrochloric acid to e.coli and staphylococcus. Professors can use these to build custom labs and curriculum for students.

“We give professors the ability to build a real platform to generate their own curriculum,” said Goodman. Professors report better performance by students who used Late Nite Labs.

Cash Money

Late Nite Labs started in 2006 and have been growing low-key through friends and family support. Labs told TechCrunch they recently raised a $1.1 million seed round from angel investors:

Harold Levy, the former Chancellor of New York City’s public school system and EVP of Kaplan, as well as Don Burton, former head of Business Development for Disney Education.


What’s Next

They plan to get into high schools, but the barrier to entry is much higher than colleges where any professor can use whatever he wants.

“The textbook industry is much more convulted in the high schools, you need a huge salesforce to dominate. There are state by state adoption requirements. It’s very confusing and requires custom probing,” said Goodman.

In the next week, Goodman tells us the team will be rolling-out a new upgraded chemistry lab. And Late Nite Labs is working on bringing students inside forensics labs CSI-style and outside the lab— to study ecology of fish populations in lakes. They will also roll-out video embedding for online lab manuals from YouTube or professor generated content.

What Goodman’s really excited about is the idea of collaboration.

Goodman said, “Web 2.0 is all about social interaction. What would it be like to have a digital lab partner online? Collaboration and working together in lab from anywhere.”

visit: Late Nite Labs, but be warned, its kind of addictive— in a good way, the way science should be.


What do you think of virtual labs? Let us know in the comments.


related: Late Nite Labs Raises $1.1 Million To Bring Virtual Science Education To Campuses | TechCrunch.