Knight Lab – The Undergraduate Experience

View Github Repository Open presentation in a new window

TylerFisher

See all presentation from TylerFisher

Knight Lab – The Undergraduate Experience

0 0

knight-pres-2013

Presentation for Knight Foundation visit, Dec. 3

On Github TylerFisher / knight-pres-2013

Knight Lab

The Undergraduate Experience

by Tyler Fisher

Hello!

Good afternoon, my name is Tyler Fisher. I'm a senior at Northwestern studying journalism in Medill, and this is my second year as an undergraduate fellow at the Knight Lab.

Today, I'd like to talk about the work I've done here at the Lab, including SoundCite and some new, internal projects. But more importantly, I'd like to show you how Knight Lab played an important role in my development as a journalist, offering me opportunities I could not find within Medill's standard curriculum.

Finding the Lab

My hiring here is unique in that I came to the Lab with an idea. In May 2012, I was enrolled in a course called "Advanced Design and Interactive," taught by Professor Jeremy Gilbert, who is a faculty collaborator with the Lab. For the final project, he gave a prompt that simply said,

"Design the future of news."

That's a lofty prompt. I had no idea what to do. The night before we had to give product pitches to the class, I had nothing. I started thinking about what frustrated me in my previous journalism experience.

I thought I was Greil Marcus.

When I entered college, I wanted to write about music. I actually enrolled as a dual-degree student in euphonium performance and journalism. In high school and in my first years here, I wrote for various publications online that served as my first introduction to online publishing.

I found that the most difficult part of writing about music was writing about music. Putting sound into words is unnatural. Often, the writer's mental model of music does not match the reader's model. Writers, including myself, tend to rely on cheap metaphors and poor analogies to make points about music that can feel forced and out of context.

What if we had a way to embed snippets of music into our text so we could simply cite the section of music we are discussing like a quotation? Then, we could easily move onto the fun part of criticism: the critique.

Thus, my idea for SoundCite was born. Throughout the rest of the quarter, I hacked together a prototype that Professor Gilbert found promising enough to pitch to the Lab. That summer, the Lab hired me as an undergraduate fellow to develop SoundCite into a real product. For all of my junior year, I worked on SoundCite at the Lab, and we released a beta at the end of the school year. I'll come back to what I learned from the developing SoundCite with Knight Lab, but I'd like to step back and talk about my experience in Medill as a whole.

Journalism, Technology and Medill

Journalism education is deeply flawed.

The average student graduating from Medill does not know how to compete in digital journalism; they simply do not have the skills. It's not from a lack of drive, curiosity or intelligence. My peers are some of the smartest students in the country. Medill is producing the best damn newspaper reporters, magazine designers and broadcast journalists in the country.

I could have passed through Medill with no digital skills.

But shockingly few of them know how to use the web. Even fewer know how to _create_ for the web. That could have easily been me.

Web classes at Northwestern

  • Interactive News
  • Web Producing
  • Intro to Computer Programming
  • Tools and Technology of the World Wide Web
  • Human-Computer Interaction
These are the courses at Northwestern that focus on teaching web skills. Notable omissions here are Professor Gordon's course on networks and a database class Medill tries to teach, but never seems to get enough interest from students. I have taken four out of these five in some form or another. With the exception of Intro to Computer Programming, which teaches Python for the purpose of teaching the basics of programming, all of these courses teach an introduction to HTML, CSS and JavaScript. And for the average Medill student, there's a catch.

None of these classes are required.

No required class in Medill attempts to teach digital literacy holisitically. Only students with pre-existing personal interest will be exposded to the web. And for the students who are interested and want to extend their knowledge further, there is nowhere within the curriculum for them to go.
I got interested through a bit of dumb luck. Near the beginning of my freshman year, I started attending meetings for the interactive section of North by Northwestern, the campus's best online publication, on a whim. At that time, the section still operated in Flash, so I started slowly learning ActionScript. By the end of my freshman year, everyone left or graduated, and I found myself somehow editing the section.

How do I learn with no roadmap?

How do I teach without skill?

Editing the interactive section also meant teaching producers how to code. So naturally, that meant I needed to learn how to code. I began reading up on the latest technology and saw the need to leave Flash behind and adopt "HTML5" as our savior. I had no concept of what that actually meant, but I was learning all I could and trying to stay one step ahead of my producers. Meanwhile, I ran the section for a year, and we produced projects ranging from unsightly to mediocre.

Coming to Knight Lab gave me a roadmap.

While this period of my life got me interested enough in journalism and technology that I dropped my music major and decided to focus on coding full-time, I would consider the work I did largely unsuccessful. And I think that is because I had no roadmap, no guidance to point me in the right direction. I was wandering aimlessly, just trying to make some jQuery work. There was no sense of the overall picture. This is where I was when I entered Professor Gilbert's class and eventually found the Knight Lab. Through sheer luck, I interned at the Chicago Tribune News Applications Team and began working for the Lab in the summer of 2012, and finally, I found my map.

Developing SoundCite

For the first few months of SoundCite's development, I worked with Professor Gilbert when I could and we would report back to Miranda every week or two. Most of that time was spent fleshing out what SoundCite was actually going to be rather than focusing on functionality. This was my first real product, so learning how to think in a product state of mind was new to me.

Initially, collaboration was hard.

The Knight Lab was still in the Ford Engineering Design Center, where all of the staff had their own offices, and it was hard to work collaboratively there. I didn't exactly feel like a part of the Lab yet.

Then, two key things happened.

We hired Joe Germuska. We moved to Fisk Hall.

Joe!

Near the end of November, we hired Joe Germuska to be our Chief Nerd. Knowing Joe from my time at Tribapps, I was ecstatic. Having Joe to organize the myriad engineering efforts at the Lab has been invaluable.

An open, collaborative space felt like home.

Shortly after Joe's hire, we moved to our current space in Fisk, which put the development team and any student fellows in the large open space. It is amazing how much that changed my experience at the Lab. Now, I probably spend 15-20 hours per week in this space working on Lab projects or doing my own work.

April 2013: Alpha release

June 2013: Beta release

Beginning in winter quarter in January, Joe became my first point of contact on SoundCite, and development took off. I had a better sense of the Lab's resources and what to use it for. I felt comfortable working in the Lab. In April, we released the developer alpha of SoundCite, and in June, the beta. I completed my first real-world product.

SoundCite was my first product.

I'm still parsing what I learned.

I'm still parsing all of the things I learned in the process. I'll get into some of the takeaways from the development process, but first, let me demo SoundCite quickly for you.

What is a journalism product?

For SoundCite, we needed to create a tool that was as friction-free as possible. Thus, we chose SoundCloud as an audio host. Anyone can host files on SoundCloud, and similarly, any file from SoundCloud can be used for a SoundCite clip.

What do journalists need?

  • Universally accessible
  • No coding required
  • Free to use
Again, SoundCite was the first product I designed and released to the world, so it was the first time I had grappled with making a product for journalists. That requires thinking about _journalists_ as your users. What are their needs?

SoundCite's Answers

  • SoundCloud
  • Generated embed codes
  • Easy-to-use clip creator
We also wanted the clip creation process to require as little coding knowledge as possible. Most of the development time on SoundCite was spent on the clip creator, not on the clip display. Its current iteration allows users to use the SoundCloud player to set start and end points on the clip and receive an embed code to paste for each clip. SoundCite is a user-friendly solution that requires little technical knowledge. Its use of SoundCloud as an audio backend allows anyone to use the product and ensures the reliability of the audio actually working. We think this fits the requirements for journalists, and the results confirm this. Newsrooms from The Washington Post to Education Week to WBEZ have successfully used SoundCite with little technical support from the Lab.

Multimedia is the next frontier of web storytelling, long and short.

In developing SoundCite, I considered a lot of what makes multimedia powerful. At the most basic level, all multimedia gives storytellers the power to show rather than tell. With my initial inspiration for SoundCite — music criticism — I wanted to show the user the snippet of music rather than struggle to tell it.

Longform storytelling is beginning to standardize multimedia.

On the web, long-form storytelling has exploded in the past year. The phenomenon has extended beyond traditional news outlets. Publications from Mashable to SB Nation to Pitchfork have developed their interpretations of what long-form stories on the web can look like. They feature cover-sized images, looping videos, and complex animations. The emphasis is on the multimedia, using the web to show rather than tell.

SoundCite works with any story.

SoundCite is a tool that works in any type of story, providing the "show, don't tell" aspects of recent long-form storytelling on the web to any form of web publishing. [The Washington Post](http://www.washingtonpost.com/blogs/the-fix/wp/2013/06/26/this-tweetstorm-was-planned-in-advance/) used it in a fairly standard recap story of Wendy Davis's 13-hour filibuster in June 2013. What could have been just another story about Wendy Davis in the aftermath of her national rise to fame became something more, something that provided her words in context of the whole story. The future of storytelling on the web is bringing the powerful tools we have standardized in longform storytelling and making them accessible to all of our stories. SoundCite, an inline tool that can be used in any system, is a step towards that goal.

Audio is not a first-class citizen on the web.

SoundCite still doesn't work on mobile.

SoundCite still doesn't work on mobile. Without getting too deep into technical details, SoundCloud's API is built on a JavaScript library called SoundManager2, which enables native, HTML5 audio when possible while providing a Flash fallback when not possible. Due to how SoundCloud passes SoundManager2, the Flash fallback is always used in SoundCite's case.

Without a universally acceptable audio codec, browser suppport will remain weak.

This is because native audio support in modern browsers is still a work in progress. While support for the mp3 codec has reached most modern browsers, its use is a [mess of patents](http://en.wikipedia.org/wiki/MP3#Licensing_and_patent_issues) regarding the encoding and decoding of the codec. Thus, browsers like Mozilla Firefox (on Apple devices) and Opera have not yet supported the mp3. Given the lack of a cross-platform audio codec, passing audio through Flash is currently the only viable solution. Until audio becomes a widely-supported native element of the web, using audio properly on the web will be extremely difficult.

We need more open source tools for journalists.

I struggle to see a future where every newsroom has developers working for them and solving problems for them in the newsroom. Naturally, this is our ideal to strive towards, but realistically, we need groups like the Knight Lab developing open source tools like SoundCite that any journalist can use. The success of our products, especially Timeline, have shown that there is a desire for these kinds of tools.

What's next?

After SoundCite was released, the job I had been hired to do at the Knight Lab was more or less complete. Without a major development project, the opportunity to do different kind of work for the Lab emerged. The Knight Lab provided me a roadmap, showed me where the possibilities were in the intersection of journalism and technology. But it seemed we weren't bringing many more students like me into the Lab.

Finding talent at Northwestern should be easier.

In the 2012-13 school year, we had ten undergraduate fellows. Six of them graduated with the Class of 2013. We needed to find new talent, and if none existed, train new talent. But there were no obvious places to look. Northwestern still lacked a unified community of webmakers.

The Knight Lab must become Northwestern's community of webmakers.

With no community to turn to, the answer became obvious: the Knight Lab had to become the community for journalists looking to become digitally literate. For the remainder of my time at Northwestern, that is what I am working on with the Knight Lab.

Community-Building Initiatives

  • learn.knightlab.com
  • Open Lab Hours
  • Brown Bag Lunches
This quarter, we developed three community-building initiatives that have been extremely successful in developing talent and generating interest around and about the Knight Lab.

learn.knightlab.com

At the core of our initiatives is learn.knightlab.com, a site that gathers previously existing online tutorials and resources and presents them in a guided curriculum aiming to make journalists digitally literate. The site is inspired by my own experience learning how to code. Until I came to the Knight Lab, I was largely a self-taught programmer. I worked through online tutorial after online tutorial, trying to make sense of the web. The hardest part about navigating all of these resources was knowing where to go next after I learned something. Learn.knightlab.com provides context and a path. Show the site here.

We have the technology.

What's different and important about this site is that it recognizes that we already have incredible resources for teaching digital literacy. From Codecademy to Code School, developers better than me have invested their lives in teaching others to code. Those resources should be used.
Another unique aspect is the focus on journalists. I was inspired by Professor Gordon's venn diagram outlining the different members needed to form a news apps team. These roles are defined by specific skill sets. Similarly, learn.knightlab.com should provide paths for people to follow based on their interests.

Three Categories of Skill

  • Production
  • Data
  • DevOps
I have three loose categories: production, data, and devops. While these skills often overlap for tasks like data visualization, they are an easy way to break down particular skills. Production covers front-end design and development, including HTML, CSS, JavaScript and version control. Data covers everything from basic Excel to managing data with Python. DevOps includes back-end programming with frameworks like Django and server-side management.

But first, digital literacy for all.

Still, an aspiring digital journalist should have some knowledge before embarking on a particular path. That's why the first five lessons do not mention these paths at all. Instead, the initial lessons are aimed at all beginners so that everyone who follows the site has the same base level of digital literacy. This includes basic HTML, CSS and JS as well as a solid understanding of how the Internet works.

Open Lab Hours

Complementing the learning site are our Open Lab Hours. Every Wednesday night from 7-10 p.m., the Knight Lab opens its doors to Northwestern students who wants to work on web projects. Some students come with their own work while others come to look for guidance on where to begin.

A physical space for makers and learners.

If the Knight Lab wants to become the community for web makers, it needs to open up its space and allow people to feel comfortable with us. One of the best parts of the Knight Lab is the friendships I have developed with our past and present student fellows. The Open Hours try to extend that community feeling to all interested in journalism and technology, whether they are fellows or not. The Open Hours are a place for students to ask questions, discuss important topics or just feel a part of something instead of feeling alone.

Student-run, student-owned.

Having the hours run by students allows the lab hours to not feel less like a classroom and more like hanging out with your friends. These are carefree environments where no one is unwelcome and no question is too stupid.

Brown Bag Lunches

Finally, the Knight Lab hosts brown bag lunches on Thursdays to gather our staff, faculty, fellows, and anyone else interested and discuss anything that falls under our umbrella.

Using our staff expertise

The discussions this quarter have been diverse. One week, Professor Gordon led us in a thought exercise about how to cover crime data for a metro newspaper, using his experience at the Miami Herald as an example. Another week, Joe facilitated a discussion about new business models for journalism. At an early lunch where no one from outside the Lab showed up, we spent the time discussing the learning site and where it should go. These lunches all center around the expertise of our staff and their ability to lead discussions in a productive way.

Knight Lab changed my life.

Knight Lab changed my life. Through my work here, I found a purpose in journalism. I am committed to solving journalism's toughest problems and building a community of digital journalists that can make our craft indispensable.

Every journalism school needs a Knight Lab.

I shouldn't be alone in this feeling. Every journalism school needs a Knight Lab, and every journalism student should be exposed to the innovative thinking that the Knight Lab cultivates.

Questions?