UX at Grace Hopper

Conference journal — Orlando, October 2017 🍊

I was lucky to attend the Grace Hopper Celebration in Florida during the first week of October. A celebration of women in technologies, where Computer Science, STEM, and UX students join women from the industry for discussions, talks, career advice and interviews.

Here are my highlights and key learnings:

Who’s hiring (Internship time!)

Not all tech companies have a lot of user facing interfaces or are big enough to offer proper UX internships. Some are: Amazon, Tesla, Google, Apple, Airbnb, and (soon) Chan Zuckerberg Initiative. And most of these are recruiting now or in January — Don’t wait until late Spring, most opportunities come much earlier in the year. Here’s some advice.

Talks (What’s going on in tech?)

Biases in AI (Neelima Kumar — Oracle)

Everyone is biased. The goal is to be aware of our implicit biases and this is something we’re getting better and better at. The issue is that there is an impression that data and artificial intelligence are exempt of biases. But here’s the thing: Data is collected and labelled by humans before being used to train AI. This means AI generates biased output and amplifies human biases. There are several examples of blatantly racist or sexist AI results, sometimes cultivating class discrimination.

We know this is an issue, but i’s also a complex one: It’s hardly impossible to get “clean” data, decisions made by AI are not easily understood, the lack of diversity in the tech teams makes it difficult to catch biases earlier, taking away some of the biases will also have an impact on accuracy.

Here are some solutions:
Awareness of possible biases and design for inclusion and diversity.
Explainability of individual AI decisions and accountability

This is a problem, but it’s also an opportunity:

AI will change the world, but who will change AI? It’s in our hands.

Precision Agriculture (Jennifer Marsman — Microsoft)

Jennifer Marsman from Microsoft took us in on the pilot agriculture project her team has been working on for a few years. They are combining drones (taking photos of the fields), sensors (moisture), and weather (rain, wind, pollen) data to generate daily farming recommendations through machine learning.

They ran into important challenges and found ingenious solution to them:

Connectivity — It was costing a lot of money to send the data into the cloud and process it. And connectivity in remote areas where farms are located is often suboptimal or simply absent. The team asked for government permission to use TV white channels to sent data packets, which is almost free.

Drones capabilities— Drones are not easy to use and farmers should not be expected to be drone pilots.They created an auto pilot algorithm, and an app with a simple UI.

Drones Power Consumption and Battery Life — Even the most powerful drones don’t have more than 40 minutes of flight time on their battery. They leveraged wind’s direction for takeoff, and optimized flight paths.

Alternative Futures — How AI will shape our lives

This talk was more geared towards HCI and design, and suggested an interesting way to be critical of technology and our own designs.
We did a post-it workshop in the class, where we started with a product that would be powered by AI and surrounded it by positive and negative outcomes.

AI product: Meaningful conversations through AI.
What are the immediate positive and negative outcomes we can think of?
Negative: Isolation, addiction, reliance
Positive: Opening minds, therapy, insights and learning
From there, we went even further and for each of the post-it we had to create a positive and a negative outcome.
Ex: Therapy
Positive: offering support to people who may not get it otherwise.
Negative: Might lead to only opening up to “machines”, and not people.
Each attendant picked a post-it of their choice, and sketched out a scenario for a positive outcome, and a scenario for a negative outcome, which would involve the same character.

This exercise made us realize the complexity and the unknown behind such technologies, and how much more thought needs to be put into them so we are aware of potential futures.

Atomic Design

By HBOGO Head of Design — Atomic design is the methodology many product teams use in the industry to improve the relationship between engineering and design and to make design more efficient. It is simply a library of defined design elements to make your job easier and ensure consistency across the product and easier implementation. This is something you should expect to work with in many design jobs.

The speaker also mentioned the way the role of the designer has evolved and what companies are now looking for: the T-shaped designer (rather than just the ‘UX designer’)

Thanks for reading!