Human-centered AI: The ethics of designing new technology.

Mindful Studio
4 min readOct 16, 2023

--

What is Human-Centered AI? (Published by Clickworker)

Human-centered AI is at the intersection of humanity and technology, intending to empower and assist humans while adapting to their changing requirements. This isn’t about handing over the role of human-centered design to a machine. It’s creating a tool for designing a harmonious experience between people and technology. Designers should uphold specific ethical standards, continuously challenging assumptions and biases, but it’s a significant challenge because bias is hard-wired. By contrast, AI ethics are hard-wired principles such as fairness, transparency, and privacy — meaning AI systems should avoid discrimination, provide information about their operations, protect users’ data, and consider society at large, which is very difficult for humans to uphold because it’s hard to imagine someone else’s experience.

Design Ethics Vs. Dark Side UX (Published by UX Planet)

A recent Netflix documentary tells the story of “Big Vape” following the lives of James Monsees and Adam Bowen, the founders of Juul electric cigarettes. Monsees and Bowen graduated from the prestigious Master’s program in Product Design at Stanford University. The d school at Stanford was the brainchild of David Kelley, famous for defining the guiding principles of design thinking; as a founder of IDEO, he is responsible for IDEO.edu, which makes design thinking and human-centered systems design accessible to everyone online. Like Agile, design thinking methodology believes in the utopian view of all the good technology can do for the world. And yet, Monsees and Bowen developed the prototype for an electronic cigarette even though lung cancer isn’t the only harmful effect of tobacco addiction. If the d school had a lab setup for students to use digital twins as predictive models, it’s possible that they could have helped analyze future outcomes by applying human-centered AI. The impact on society could have been predicted if ethics hadn’t been left in the hands of two college students. Despite “concerns” from faculty members, Monsees and Bowen’s final project was accepted, and they managed to attract major Silicon Valley angel investors. The brand was anti-big tobacco on a mission to save lives by providing a safe alternative to carcinogenic cigarettes. As the documentary shows, Juul succeeded in its mission to design a superior product to any competitor in the tobacco industry and grew beyond anyone’s imagination. Using the same playbook as big tobacco Juul advertisements were intriguing to underage consumers who began vaping and finding ways to buy counterfeit pods. Juul settled its biggest lawsuit payout this year, $462 million was awarded to seven states and territories as restitution for its contribution to the vaping epidemic.

Juul is an example that demonstrates how design in isolation leaves too much room for devastating outcomes. Monsees and Bowen never considered what might happen if people who weren’t like them used their product. Social media is an incredible landscape for more examples of similar linear thinking. Mark Zuckerberg has been brought before Congress and the Senate multiple times to testify that Meta is not responsible for eroding our democracy and harming young people. However, Frances Haugen, a former Facebook data scientist, testified before a Senate subcommittee in 2021, explaining how Facebook prioritizes profit over the well-being of children and all users. She shared research she and her team conducted as Meta employees, proving young teenage girls were most at risk for self-harm because of the content they consume through Instagram. When products are designed for limited use cases and personas, it’s an example of linear thinking. Systems thinking acknowledges that unintended users (human and nonhuman) will always appear. Examples like Juul and Facebook demonstrate that the designer’s role must change — and one of the most significant contributions will be human-centered AI design principles.

Frances Haugen, who testified before the Senate on Tuesday, is not the first former Facebook employee to warn of how the platform can be used to create conflict. New York Magazine Photograph by Jabin Botsford / Getty

Designers need to evolve into human-centered systems thinkers because the broader implications of technology matter. Juul demonstrates design thinking’s brilliant methodology to solve complex problems, but clever design can create problems if the big picture isn’t considered. Facebook and Juul believe they aren’t responsible for the misuse of their products because their products were designed for people like them. Juul was intended for adult smokers seeking an alternative to a long-term, preexisting habit. Juul wasn’t made for nonsmokers and underage smokers, so Juul wasn’t culpable for unconsidered groups of people. The focus was on a narrow goal based on an ideological mission to design for themselves and people like them. It’s essential to know that the technology isn’t to blame. Designers are accountable for the impact of their designs on society, and human-centered AI and systems thinking is the best way to start genuinely improving the world.

Human-centricity in AI governance: A systemic approach (Frontiers in Artificial Intelligence)

--

--

Mindful Studio

A collection of short articles exploring Industry 4.0 Design