BERLIN—With curved ceilings and pocked cement walls, the colossal Funkhaus Berlin vibrates with history. Built in 1951 by Bauhaus architect Franz Ehrlich for East Germany’s vast radio broadcasting company (dubbed “the BBC of the DDR”), it’s now a meeting space. On this day, developers are mingling at an event sponsored by Oracle, sharing ideas about modern software platforms, languages, and design choices.
As software “eats the world,” engineers aren’t just worrying about orchestrating systems that can run seamlessly across multiple geographies, they’re also puzzling over a dilemma: Just because you canbuild it, does that mean you should build it?
That was the question Dublin-based data scientist Brendan Tierney emphasized in his talk at Oracle Code Berlin entitled “Real-World Machine Learning Projects: From Detecting Fraud to Finding Cancer.” For example, he helped create an app to let drivers monitor their alertness via mobile phone camera, face detection, and machine learning. While the app could warn drivers to wake up or stop doing something distracting, it might also be used by insurance companies, Tierney noted—which could create privacy concerns.
Similarly, a women-in-technology panel discussion focused on the problem of bias and ethics in artificial intelligence. “I find it crucial to add diversity to the teams who are working on all that stuff,” said Berlin-based data analyst Maren Heltsche, board member of Digital Media Women, Germany’s largest network of women in digital business. “Often teams don’t even think about ethics, but when they do there is no dedicated role working on it.”
Software Engineering Ethics and History
“This time reminds me of the ’70s and ’80s, when I was in school,” said Alexis Noguer, head of sales for German 3D metal printing company Rolf Lenk and founder and CEO of an international technology camp for girls called the Robotics Institute. Noguer compared our current era, at the cusp of “open world” computational intelligence, to Swiss philosopher Friedrich Dürrenmatt’s 1962 play The Physicists, which debated whether the engineers who created atomic bombs were ethical. That kind of philosophical inquiry “was very pervasive in my childhood. In Germany, because of Europe’s history, I think ethics has been high in the debate. For that reason, I am optimistic that AI will be kept in check. I think writers and philosophers will have a big role to play.”
Perched on chairs in the minimalist remains of Funkhaus, which was once wiretapped by the secret police concerned with the activities of its creative personnel, members of the panel pondered different values among Europe, the US, and Asia when it comes to AI. Ronika Lewis, Cambridge, Massachusetts-based CEO and founder of RLI Consulting, an Oracle Gold-level partner implementing applications that increasingly use machine learning, brought up the importance of reaching beyond ethics boards and corporate responsibility to shape policy. Jeremy Heimans’ New Power Movement, she said, shows the modern tactics available for driving social change using crowd-sourced, peer-driven, open methods.
Falling Behind the US and China
Increasingly, European developers’ excitement about opportunities such as AI-driven applications, or planet-scale software systems made possible via containerized microservices, is tempered by a concern that Europe is not playing as big of a role as it should in designing tomorrow’s world.
“How come all the successful internet 2.0 companies are in California, and none are in Europe or in Germany?” Noguer asked.
That sense of chasing opportunity was just as palpable in Rome, where the Oracle Code worldwide tour had touched down the previous week. In a brutalist compound intended by Mussolini to house the 1942 World’s Fair (where the stunning travertine marble Palazzo della Civiltà Italiana stands sentinel), a whimsical Lego city occupied a table. Oracle engineer and Rome native Gabriele Provinciali stood nearby, demonstrating the features of a smart city project, called Proxima City. It’s a software development kit for connected, intelligent municipal services ranging from garbage collection to smart lighting to optimized parking.
“All the team is from Rome. We dreamt about how we could transform Rome to a certain degree,” he said, speaking somewhat wistfully about the municipal services in “Nordic countries.” The model SDK has been replicated by teams in a dozen cities, showing how Oracle Cloud services for the Internet of Things, big data analytics, deep learning, and digital experience can fix life’s little annoyances. “Everything we implement is what we would like to see in Rome.”
The Diversity Imperative
The Oracle Code tour has visited more than two dozen cities worldwide, and wherever the Code team goes, they find developers eager to discuss more inclusiveness across the technology landscape. Berlin attendees explored the essential role of diversity in AI. In Rome, Micaela Romanini, founder of the “Women in games Italia” association, vice director of Vigamus Foundation and Academy (which offers a bachelor’s degree from Link Campus University for game developers), and an avid player, explored diversity in game culture.
“We see that when the industry is well-developed, there are less issues about inclusion and diversity, because they have more students applying for programs, and also more professionals,” she said. “In Italy, the industry’s not so big, so we still have many issues. We see more women working in localization, because it’s a study that’s ‘for women,’ and also communication and marketing, while there are less women in programming, design and leadership roles.”
“There is a need to change the mindset,” Romanini said. “We see parents that give the girls only dolls, and the boys only Lego.”
Difficult Choices Ahead
As IoT, personal robots, and machine learning expand, ethics will increasingly come into play to help autonomous systems make difficult choices, such as prioritizing the safety of passengers vs. pedestrians. Stanford’s One Hundred Year Study on Artificial Intelligence
is an example of critical research at the institutional level, but it’s reassuring that individual engineers are also considering ethics. That makes sense, because the power to build and replicate at internet scale is now available to all developers.
“This is a global issue, we are in global companies,” Lewis told the audience during the panel discussion on AI and bias in Berlin. “The conversation in Germany will be different than in the UK or the US.” This time, the walls in Berlin weren’t listening—but these developers hope the world is.