My ATDTK 2020 Session Recaps

I had the great opportunity to attend and facilitate two Playground sessions at #ATD TechKnowledge 2020 Conference in San Jose, CA. ATD changed the format of #ATDTK offering fewer but more focused sessions and piloting a "silent disco" format that I loved.

Silent Disco

Every attendee was given a small radio device with headphones at registration check-in. Super Sessions and Playground sessions (these were interactive, hands-on sessions) were held in open areas allowing attendees to move around from session to session without much effort. Each session was assigned a channel. Attendees tuned into the session channel to listen to the speaker.

For the Super Sessions, three speakers presented at the same time on the same stage with different topics and slides - complete visual overload, but I was able to tune to adapt and focus on the speaker I was interested in hearing. When you walked into the big ballroom where the Playground was located, it was eerily silent even though it was jammed packed with attendees all super engaged in their sessions. As an attendee, it felt like the speaker was speaking directly to me; I was not distracted by environmental noises. I also discovered that I could take a break in the chat lounge one floor beneath the Playground and still channel surf the sessions and listen in. It made for a great attendee experience. KUDOS to #ATD for their willingness to experiment.

Machine Learning

Given the new format that ATD was experimenting with, my first session was a 30-min Machine Learning session in the Playground on the Disrupt stage. I had to find the sweet spot for the right amount of content and activities to drive my session objectives. Given that this is a relatively new topic that Learning Professionals are just starting to dip there toe into, I decided to go light on the slides and high on the activities. I was a little nervous about timing and having everyone complete an activity, but it worked out very nicely. The key was to keep my activities simple, relevant, and engaging. Here are the activities that I used:

1. Play Rock, Paper, Scissors (+ hands my version of paper - lol). I was initially going to have attendees play a few rounds of tic tac toe. However, after seeing the room setup and experiencing the silent disco format, I swapped my activity to Rock, Paper, Scissors. A last minute decision minutes before my presentation - talk about on the fly. I asked participants to pair up with their neighbors and play a few rounds while also tracking the number of rounds played, which play each made, who won each round, any patterns they individually spotted. As they started playing, I threw a wrench in the game, asking them to go faster, faster, faster. I stopped the game after a minute and asked questions about the data they were supposed to collect while they played. I used this game to create a mental model of how a machine learning model is trained. Overall participant feedback: it was fun and made its point.
2. Play a round of QuickDraw. After sharing brief information about the different ways a machine learning model is trained, we then played a game of QuickDraw to add new data to a neural network to help it recognize doodles. Participants accessed the Google website and played one round. We spent about 2 minutes on this activity. This activity allowed participants to participate in training a machine learning model through gameplay and framed the discussion simply. Overall participant feedback: can we keep playing?

1. Psychological Digital Footprint. This was by far my most popular activity. Prior to my session, I uploaded my LinkedIn and Twitter data (dating back to 2008 to present - a significant amount of data) to a web app called Apply Magic Sauce that uses a machine learning engine to creates a psychological profile based on your social media data. I also had my friend Ann Rollins upload her data to generate a similar report. I shared our results - it turns out I'm the epitome of masculinity and Ann is really NOT as smart as the average bear. This generated some chuckles and also interesting questions about data ownership and bias algorithms. My goal with this activity was two-fold, provide a real-world view of the amount of data needed to train a machine learning model and illustrate the bias that can be injected into a machine learning model resulting in skewed predictions and personalization - Overall participant feedback: I can't wait to discover what my digital footprint says about my psychological state.

Paper Prototyping: Learner Journey

In my last session was a 45-min Playground session on the Build stage. My goal was to get participants to roll-up their sleeves and take a macro view of their learner's journey - not tied to a learning pathway, but the overall learner experience. Being that this was a crazy early session - 8 am, I wanted to kickstart the participant's brains right out the gate with meaningful activity. I provided everyone with a 14" x 17" blank sheet and some post-it notes and started the session by framing what we were going to be doing. I wanted them to explore the learner experience from the time a course was assigned to them, not focusing on what they do in the course but how they learn about the course, get to the course, take the course, and what happens after.

Task 1 - Write Out the Process

In this activity, participants were asked to write down the process that a learner goes through to take a course and what happens after they take the course. I asked guiding questions and prompted participants on what they should capture on a post-it note. We decided that the learner would get an email stating a 30-min course had been assigned to them. Participants were prompted to write out one sticky note with the word email and an image if they wanted to draw something. I then asked, "What next? What action does the learner have to take?" They all agreed the learner had to click a link in the email. I prompted participants to write out one sticky note the step Click Link. I continued asking, "What next? and What action does the learner have to take?" At the end of this activity, we quickly debriefed, and some interesting insights were shared like:

"That's a lot of steps for a 30-minute course"

"I discovered that I've been making a lot of assumptions about my learner"

"I don't think I really know my audience"

Task 2 - 5 Map, Identify the Problem, Spitball Solutions, Test it out

We moved on to a fun exercise of creating a "learner journey" that maps out the different learner interactions we had just written out to the emotional impact of these interactions. We used a smiley, meh, and bummer face to map out the experiences. This method helps to provide insight into the learner experience, which by the way, is not limited to the fantastic course you just created and assigned.

We then reviewed our maps and individually selected one step in the process that we each would identify as a problem that, if fixed, would have a significant impact on the learner experience (i.e., taking a leaner experience from let's say 30% satisfaction to 50% satisfaction).

We then created a Storyboard where our problem was the main topic. Participants were then asked to spitball potential solutions (ideas) that would fix the problem - writing one idea per post-it note and adding it to the storyboard. The final selected solution would have to address the problem identified directly.

I had a lot of fun facilitating this session; we had some great conversations. Overall participant feedback: I had never thought of looking at things at a broader scale.

Takeaways

#ATDTK forced me to get focused on each of my sessions; I really liked the Playground session format. My participants wore their headsets and were really engaged, which was different and fun. Once I discovered that people might be channel surfing and not actually be sitting in my session, but still listening in. I changed my verbal presentation to address those who may not be in the room. Saying things like: "You are on Channel #, listening in on XYZ session, I'll make sure to verbally summarize things that are on the screen and the activities so you can follow along." "Grab some paper and a pencil; you can use your conference binder...."

I would love to see this format used again with some minor tweaks. Again, Kudos to #ATD for experimenting