On Monday, Google’s AI research division, Google Brain, announced its newest research program dubbed as People + AI Research initiative or PAIR.
According to Google.ai, “PAIR is devoted to advancing the research and design of people-centric AI systems.” The people behind PAIR also added that they aim to understand the full spectrum of human interaction with machine intelligence.
PAIR initiatives’ primary goal is to do fundamental research, invent new technology, and create frameworks that will be beneficial in humanizing artificial intelligence. The program also focuses on the relationship between users and technology, the application that AI enables, how to make it widely inclusive, and everything that touches the “human side” of AI.#GoogleBrain just launched #PAIRInitiatives to further improve #AI research.Click To Tweet
The Three Areas of the PAIR Initiative
A July 10th blog post on the Google Blog stated that PAIR’s research covers three areas based on different user needs: engineers and researchers, domain experts, and everyday users.
With engineers and researchers, PAIR initiative wants to explore how can they make it easier for these people to build and understand machine learning systems. Also, what educational materials and practical tools will they need for their research works.
For domain experts, PAIR wants to determine how AI can augment professionals in their work. The initiative will tackle possible ways to support doctors, designers, technicians, and other working professionals as they integrate AI into their jobs.
Lastly, PAIR initiative wants to ensure that machine learning is inclusive so everyone can benefit from breakthroughs in AI. They want to explore the possibility of democratizing the technology behind AI to help open up new AI applications.
As part of Google Brain’s effort to further improve AI research through PAIR initiative, the company developed a set of best practices that their teams use to design experiences like machine learning.
Human-Centered Machine Learning
Jess Holbrook, a Google UX Manager and Staff UX Researcher, and Josh Lovejoy, a Senior Google Interaction Designer, published on Medium a post about the ‘7 steps to stay focused on the user when designing with machine learning.’
The post, titled Human-Centered Machine Learning, gives researchers and designers some pointers on how to effectively design ML-driven products. As per Holbrook and Lovejoy, the pointers were drawn from their work experiences with Google AI and UX teams. Below are the seven pointers as stated in the Medium post.
- Don’t expect Machine Learning to figure out what problems to solve. We still need to define that. You still need to do all that hard work you’ve always done to find human needs.
- Ask yourself if ML will address the problem in a unique way. Once you’ve identified the need or needs you want to address, you’ll want to assess whether ML can solve these needs in unique ways. There are plenty of legitimate problems that don’t require ML solutions. Address the following uses cases:
- Fake it with personal examples and wizards. When doing user research with early mockups, have participants bring in some of their own data—e.g. personal photos, their own contact lists, music or movie recommendations they’ve received—to the sessions.
- Weight the costs of false positives and false negatives. Your ML system will make mistakes. It’s important to understand what these errors look like and how they might affect the user’s experience of the product.
- Plan for co-learning and adaptation. While ML systems are trained on existing data sets, they will adapt with new inputs in ways we often can’t predict before they happen. So we need to adapt our user research and feedback strategies accordingly. This means planning ahead in the product cycle for longitudinal, high-touch, as well as broad-reach research together.
- Teach your algorithm using the right labels. Labels are an essential aspect of machine learning. There are people whose job is to look at tons of content and label it, answering questions like ‘is there a cat in this photo?’ And once enough photos have been labeled as ‘cat’ or ‘not cat,’ you’ve got a data set you can use to train a model to be able to recognize cats.
- Extend your UX family, ML is a creative process. Machine learning is a much more creative and expressive engineering process than we’re generally accustomed to. Training a model can be slow-going, and the tools for visualization aren’t great yet, so engineers end up needing to use their imaginations frequently when tuning an algorithm. Your job is to help them make great user-centered choices all along the way.
The set of practices presented by Google and the launch of PAIR initiative is proof that even the most advanced and wealthy tech giant today has a lot to learn about designing ‘humanized’ machine learning artificial intelligence. However, PAIR initiative is still a good move to better understand the future applications of AI in people’s lives.