It’s Time for Social Media Algorithms to Grow Up

When I was growing up in Edmonton, Saturday morning car rides were filled with the sound of CKUA Radio, from the University of Alberta. Opening with Ethel Merman belting out “There’s No Business Like Show Business” came the weekly musical. Camelot, Kiss Me Kate, Brigadoon – all accompanied by my mom Esther z”l singing along with every one. Musicals made mom very happy, she had a good voice and a great memory for lyrics.

Years later my appreciation for CKUA came from an afternoon show by the legendary Bill Coull (yes, pronounced “Cool”). Bill had the most varied and wonderfully-connected radio show I had ever heard, at least until FM started experimenting, and CBC’s Alan McFee’s Eclectic Circus came to late-night radio. Bill Coull’s show is legendary for a particular reason – he refused to follow a formula. His show made connections that I had never heard of, or imagined. Bill would play a classical piece then show how that theme had been lifted and put into a piece of prog rock, then play a jazz trio who played in that same musical mode, either making connections or contrasts. For me Bill defined the word eclectic.

Eclectic is the opposite of what most of today’s algorithms value.

What is an algorithm? It’s just a computer program. The algorithms I’m referring to are of a particular type – the ones that try to determine what you should read, or listen to, or watch next. These predictor algorithms are all around us in the online world. Facebook builds a timeline for you that try to keep you on-platform, where they can sell more advertising space to entice you. Twitter tries to do the same. So does Spotify, and Netflix is happy to tell you about how they do it too.

What these predictors are missing are serendipity, and non-trivial connections, and eclecticism. They are no Bill Coull. They are no Roger Ebert. Their scope of deciding what you might like next is based on popularity and the assumption that your taste is just like everybody else, that if you like that, then people like you also liked this.

Thankfully I have a variety of friends (and children) who know me well enough to suggest things that break up my own or algorithmic silos. I am grateful for that. And I am worried. I’m worried about the people who don’t have a diversity of friends to challenge their thinking, the people who want to only hear about and believe the people who think just like they do. Because the decisions about who should run our countries, who we should trust, whose businesses we should support and which we should boycott, and whether some acts of law enforcement officers are fair or racist, are being affected by algorithms that value polarization over diversity. And that’s a problem for all us.

Facebook is the most powerful opinion-shaper for many people these days, and their algorithmic choice “For instance, the Pages and Groups that people identified as most meaningful were often the ones that they’d followed for a long time, the ones they engaged with often, and the ones that had a lot of posts and activity.” What that means in practice tends to be the posts you are likely to agree with, and posts that are highly polarized.

At Thoughtexchange we’ve built a platform that brings people together over decisions that matter – like priorities for building new schools, how to make company cultures more equitable, and how we balance safety and connection as we forge this path through a pandemic.

One of they key elements that makes Thoughtexchange work is a diversity algorithm. The algorithm that selects which thought you should consider next is based on randomness, fairness, and diversity. We find that providing a diverse set of thoughts to consider and rate in a safe space without knowing whose thought it is, and without having to defend your choice to people who might disagree with you, helps connect a community to one another, helps them increase empathy for alternative points of view, and in the end informs decisions that are representative of the group’s highest shared values.

Recently Yvonna Cázares, Director of Community Engagement in Oakland California posted this on LinkedIn after running several city-wide exchanges and being exuberant about the engagement and insight Thoughtexchange provided adding, “Thanks for making democracy possible in this important moment!!”

Until now we haven’t able to make decisions with the active participation and discussion of thousands or tens of thousands of people in a timely, fair manner. Maybe now that we can we should replace the polarization and animosity that seems to accompany majority-based voting. Imagine democracy created by conversation and active engagement to find the common ground to move ahead together.

It might be time to re-think democracy.

Apollo at 50

It’s been 50 years since Apollo 11 launched, and its impact on my life has been profound. The Apollo mission was the most audacious engineering project since the Tower of Babel, and it inspired me in many ways.

The Apollo program is a powerful lesson about bringing together political will, the corresponding budget, and a lot of math and engineering. While the engineering was state-of-the-art for 1969 it is still so amazing that I meet people who think the challenge was so formidable that it was impossible, and was actually just a film set directed by Stanley Kubrick.

The history of designing the Apollo spacecraft is amazing in itself, and there is a beauty in the flow of necessity. The key decision was made to send 3 astronauts, to have 1 in orbit around the moon while the other two descended in the Lunar Module to land on the moon, then to launch and rejoin the orbiting command module. Once you know the mass of the astronauts, their living supplies, the weight of the spacecraft and fuel, the size of the Saturn V was “just math,” calculations that lead to the required velocity and power needed, creating the largest machine every built.

As a software engineer I have a great appreciation for the complexity of building a real-time computer in the 1960s when computers were the size of rooms, and the invention of the transistor was less then twenty years old.

Margaret Hamilton is one of my new-found computer engineering idols, who not only coined the phrase “software engineering” but also showed us by example the difference between computing science and software engineering. Science uses hypothesis and experiment and failure teaches as much or more than success. Engineering is about building things that don’t kill people even when something goes wrong.

Even though the Apollo Guidance Computer was state-of-the-art at the time, it had quite limited capability, something more like a scientific calculator that took input from sensors and a crude user interface called the DSKY. Margaret realized that it was possible for the computer to become overloaded if the sensors send data more often than they should have (an event that actually happened by mistake) and that by creating a computer task whose job it was to self-monitor, could decide to drop low-priority tasks in favour of the critical ones, which included monitoring the landing radar and calculating distance and velocities.

I may have only been 11 years old when the Eagle landed and became Tranquility base, but seeing that people could use science and technology to make a dream come true has been an inspiration for me ever since.

We might feel the same when people set foot on Mars.

[foogallery id=”480″]

metabadge reunion

We convened a metabadge reunion event, shared a toast to our recollections of one of the best teams we each had ever worked with – high expectations, an adaptive process (including estimating, and unit tests!) and an amazing outcome of a voice-based pda at least a decade ahead of its time.

metabadge team at Seasons in the Park. April 23, 2017

The Demo Doctor

Interaction design is about creating experience for users of technology that enhance their lives. The best interaction design, like other design objectives, tend to disappear and get out of the way of the user trying to accomplish their goal. At its very best, it anticipates use cases the user hasn’t discovered or articulated yet, but when they try it, they feel supported and encouraged.

That’s what makes people love the products they use everyday. They’re designed for humans, by other humans, and the care comes through.

For over 20 years, working in areas as diverse as telephone switching systems, voicemail, satellite imaging, 3d computer graphics, and digital prepress, I’ve combined designing much-loved user experiences with a kind of wholistic approach to Agile product management. The result – exceptionally profitable products that users love to use (and their managers happy to pay a premium for!)

As the Demo Doctor I have successfully demo’d and sold millions of dollars of products, by working the features and story down to its wabi-sabi roots. The elevator pitch is so 90’s – now you have to demo your product in that narrow time window in a way that shows both your accomplishments so far, and the promise of upcoming capabilities. I can help you do that.