Scenario: Smoke & Mirrors

Stefan Ullrich, Rainer Rehak

“Are you in it for the cause or in it for yourself?” Hannah doesn’t know how to answer her housemate’s question.

“If you’re in it for the cause,” Jelena continues, “you should accept. Ethics and sustainability are right up your alley. And there’s no such thing as bad publicity!“

Hannah takes another look at the letter from the big organization “AI&I4 Future”—a media conglomerate that is known in social media platforms not least of all for the impressive visuals in its ad campaigns. AI&I4F bills itself as a visionary lab incubator and crash-test institute for artificial intelligence systems. The letter is an invitation to join other experts and activists throughout Europe in formulating a manifesto for sustainable AI with an eye for the common good. The list of invitees reads like a roster of A-listers from all walks of life—established scientists and researchers, popular actors and actresses, veterans of the anti-nuclear movement, young influencers from the eco-tech scene—in short, the list has been curated to include every imaginable stakeholder.”

“Stakeholder,” Hannah suddenly repeats the term to herself out loud. “Right. As if I were on the same page as THEM. What does “multi-stakeholder” even mean in this context?”

Jelena is somewhat taken aback and slightly star-struck, fiddling at the multiple layers of her homemade outfit. “What’s the problem? Is it just that these AI people are so hip and put out such awesome visuals?” As much as she hates to admit it, above and beyond the substantive issues, somehow this is indeed her problem. She feels a bit like a fool, but after some back and forth, she decides to join the panel of experts after all.

A few weeks later the first gathering of all participants takes place at an old estate in the country’s south, where AI&I4F is headquartered. Electric cars pick them up from the train station. She’s seated in the back with the big-name transformation researcher Mats Hillebork. She can scarcely understand him through the plexiglass partition and mask, and she’s not a native speaker of English, either, but he’s obviously cutting jokes about all the hullabaloo surrounding the meeting. In English, he tells her that the estate once belonged to the Fugger Family of merchants and that AI&I4F ultimately renovated the place as a sustainability showcase. Hadn’t Hannah ever wondered about the historical windmill adorning the AI&I4F-logo? But it’s not exactly ecologically sustainable, Hillebork claims, as they drive past the mill. “That’s a mill from the Habsburg dynasty in Spain. Ten years ago, it was dismantled and brought here brick by brick and beam by beam—the old trestle windmill wasn’t snazzy enough!” Hillebork chuckles, leaving Hannah to wonder whether it was just an off-color joke or whether she might have missed something in its telling.

For an organization that specializes in artificial intelligence audits, there is surprisingly little by way of technology here. Sure, there are the E-cars, and the surveillance cameras, but aside from that, scores of employees are bustling around the sprawling grounds. Women washing clothes on washboards, young men drawing water from the well and an old butler posted at the entrance to the house giving directions. There are bowls of fruit and vegetarian snacks everywhere. It is pleasantly cool in the castle, yes, everything about the place is very castle-like. She has two rooms and a modern bathroom all to herself; everything here looks ultra-modern, in contrast to the rest of the building. The latest weather data and community news appear in a corner of the bathroom mirror. That’s how she finds out that the kick-off meeting will last three days. „There I go saying ‘kick-off’ again, I’m already picking up their language: that is not A-OK-AI. “ Those last words are the advertising slogan of a well-known AI voice assistant, and indeed: the characteristic double ring can be heard in the bathroom as the system waits for input from the unsuspecting user.

Hannah gets ready for the first meeting before the scheduled “AI Dinner.” She’s expecting it to be a work meeting, so doesn’t bother getting all done up for it. And she’s especially surprised when she arrives downstairs in the lobby to find a slew of TV cameras running. People are being interviewed and photographed, and she sees journalists wandering around asking the experts to explain the extraordinary potential of AI and how important it is that this technology be dealt with ethically—not the way it is in the USA or China, as they say. What a wonderful world it will be with AI taking so many problems off our hands—and it can even be exported. The remaining events that follow in the next few days at the castle aren’t especially big on content, either. Moderators often trade in generalizations and platitudes, and the agenda is peppered with PR-heavy breaks—plenty of photo-ops and interviews, in some cases, even with an international media presence. Hannah is rather disappointed: she was expecting much more along the lines of substantive work. At the same time, she met a lot of interesting people active in her area of expertise, and she must concede that she’s enjoyed these few days of respite from her daily grind. Maybe things like this are needed now and then, and she should just relax a bit—she certainly has earned it by working so diligently the rest of the time.

These types of meetings take place every couple of months, and within two years the experts have penned a final report, signed by all parties. With its smartly designed visuals, the document features prominently in AI&I4F’s current ad campaign promoting the option of obtaining an AI Quality Seal for European value-based AI products. But, in Hannah’s view, the document is anything but “visionary” in substance. At its core, it merely outlines the overall state of discourse on the subject and otherwise falls far short of presenting the participants’ ideas, visions, and potentialities. Besides, there was never any mention of this campaign at the outset, and she’s not exactly on board with it because it essentially calls for “voluntary self-regulation of the economy.” Ultimately, Hannah is annoyed that she invested so much time and energy into the project that turned out to be exactly what she was critical of and had feared it would be from the beginning: the paper is riddled with shallow generalizations—after all, who would be opposed to “trustworthy AI that supports sustainability and the common good”? Although the signatories include all the relevant figures from critical circles in culture and science, the document contains little well-founded critique and few concrete constructive ideas. On the other hand, she’s come away from it with greater notoriety as an activist-influencer–she’s getting a lot more media inquiries than she used to and is thus better equipped to disseminate the results of her work and her ideas in the media.

Meanwhile, she reads with piqued interest that, while she was busy working on the AI&I4F project, her fellow activists were actively participating in political and legislative processes, both nationally and EU-wide —and she hadn’t even bothered to respond to the emails from her old clique. These activities certainly didn’t get the same kind of media attention, but maybe they are having an under-the-radar effect because the EU Parliament and the German federal government had actually asked for assessments of digital regulation in general and AI in particular within the context of sustainability—precisely her area of specialization. She’s not sure what to make of the whole thing, or how to respond to future initiatives involving AI&I4F because there won’t be any immediate action resulting from the legislative consultations either. That night, she remembers something else she told Jelena two years ago: “Activists and scientists should really think about where they invest their precious time—and where they don’t.”


  1. What do you think about these types of events? Are they essential in fostering meaningful social discourse? What dangers do they entail?
  2. The whole castle looks like it’s been sustainably designed, but it actually reproduces clichés and artifice. Does this succeed in supporting a good work atmosphere?
  3. What might Hannah have done to make the meetings more productive for her?
  4. Isn’t it better to attract media attention than it is to work silently on deep thematic issues encapsulated in your own little cubicle? Is Hannah hypercritical in this regard?
  5. What forms of participation in the democratic process are advantageous? What criteria are applicable here? Do the activities of AI&I4F meet them?
  6. Irrespective of this scenario: What recent social issues have elicited calls to consider “ethics” and “values”? Was this topic in fact a new and uncharted territory so that it merited widespread thematization and discussion, or was the call to action perhaps merely a ploy to forestall regulation of the industry?
  7. Do we need a code of ethics to regulate codes of ethics? That is, a set of guidelines and red lines applied to ethical guidelines to ensure that they actually serve their purpose and aren’t merely misused as a smokescreen?

Published in Informatik Spektrum 44 (2), 2021, S. 131–133.

–Translated from German by Lillian M. Banks

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>