At Rise of AI 2022 in Berlin, a panel moderated by futurist Tanja Schindler with Daria Saharova (World Fund), Janette Wiget (Merantix / AI Campus Berlin) and Mario Brandenburg (Parliamentary State Secretary, German Federal Ministry of Education and Research) discusses whether AI can solve major world problems. Saharova argues climate tech is where AI has the clearest positive impact, citing satellite-based crop and wildfire monitoring and new low-carbon materials such as cement alternatives; Wiget points to breast cancer screening AI and to her portfolio company Cambrium's protein-engineered materials replacing petrochemicals, and stresses that algorithms alone do not heal anything without usable products and cross-stakeholder collaboration; Brandenburg frames AI mainly as a tool for better, data-based political decision-making. The panel debates AI's own energy footprint, noting German data-center emissions reportedly exceed aviation's, and the tension between entrepreneurial curiosity and regulation. On risks, panelists argue technology is neutral and that visible bias in AI data is preferable to hidden bias, while agreeing Europe needs a common regulatory framework rather than each country acting alone. Audience questions cover European collaboration, encouraging corporate entrepreneurship, and public fear of AI, closing with a call to build trust through concrete products, not only ethics boards.
Key takeaways
- Daria Saharova identifies food/agriculture, land use and the built environment as the three highest-CO2 industries where AI has the clearest climate impact.
- Janette Wiget argues algorithms alone do not heal anything; impact requires usable products and collaboration across researchers, corporates, startups and policymakers.
- A panelist cites data showing Germany's data-center and AI infrastructure emissions exceed those of the entire aviation sector.
- Mario Brandenburg argues Europe's self-positioning as the 'ethical AI' region is not enough without competitive products, citing Skype's roughly 20-year head start over European videoconferencing.
- Panelists agree current AI is not general enough to be truly uncontrollable, since problematic systems can still simply be shut down and fixed.
Detailed notes
Where AI's positive impact is clearest
Saharova, founding partner of World Fund (described as Europe's largest climate-tech investor), names food and agriculture, land use, and the built environment as the highest-CO2 industries where AI helps most: satellite companies combining infrared data with AI to detect crop pathogens and, via the Munich company OroraTech, wildfires; personal glucose-sensor data to optimize nutrition; and new low-carbon materials such as cement alternatives. Wiget, associated with Merantix and co-founder of the non-profit AI Campus Berlin, highlights the portfolio company Vara, which screens for breast cancer and is expanding from Germany into Latin America, and Cambrium, which engineers proteins to replace petrochemical and slaughterhouse-derived materials. She stresses that algorithms alone do not heal anything; impact requires products and cross-stakeholder collaboration between researchers, corporates, startups and policymakers. Brandenburg says AI's clearest political value is enabling data-based rather than gut-feeling decisions, which he says are still rare in daily politics.
AI's own energy cost
Wiget cites a figure that data centers and AI infrastructure in Germany emit more CO2 per year than all flights combined. Saharova points to quantum computing as a longer-term route to higher performance with better energy efficiency. Brandenburg argues energy spent on AI is often well spent if it produces new materials or cures, but says no country has yet answered how growing AI energy use should be limited, and calls for a common European regulatory framework, noting it is cheaper to host data centers in Sweden or the Nordics due to their energy mix.
Bias, control and the case against overreacting to AI risk
Asked how AI could ruin the world, Wiget says most risk comes down to human responsibility and education, not the technology itself. Saharova argues technology is neutral, and that the priority is building regulatory frameworks and incentives flexible enough to still allow innovation, alongside a more positive public narrative about where technology is heading. Brandenburg argues that AI making societal bias visible, for instance showing the internet was already "white and halfway racist," is preferable to bias staying hidden, and that current AI is not general enough to be uncontrollable; he cites a chatbot that developed problematic behavior and was simply shut down as proof current systems remain correctable.
Curiosity versus regulation, and Europe's competitive position
Brandenburg says curiosity is not a political KPI and that regulatory sandboxes have emerged as a way to let ideas prove themselves before facing the full weight of regulation, alongside a new social-entrepreneurship unit at the ministry. He warns Europe risks trading dependence on Russian oil for dependence on Chinese solar panels, and more broadly on Chinese or American AI, and argues Europe's niche of "ethical AI" is not enough without competitive products; he cites Skype as a warning that Europe took roughly 20 years to build a comparable alternative (Webex/videoconferencing).
Q&A highlights
Asked how Europe should balance collaboration against regional strengths, Brandenburg argues shared data governance is more manageable and decentralizable than physical goods, and that collaboration should be selective (research and healthcare more readily shared than, say, national eID systems). Saharova adds that colder countries appear more innovation-driven historically and that Europe should better understand and learn from such regional differences. Asked how large corporates can be encouraged toward more AI-driven entrepreneurship, Brandenburg describes his own political role as a translator and mediator between startups and slow-moving institutions. Asked about public fear of AI, Schindler argues that people wrongly imagine the future as a single fixed path rather than a space of multiple possible futures they can actively steer, like a sailboat captain navigating storms rather than drifting down a river.
Quotes
A technology is neutral, right? It's what you do with the technology.
Daria Saharova
AI would be something like a fridge. Everybody has it, nobody knows how it works, but there's lots of trust that this thing is good.
Mario Brandenburg
As futurists, we never talk about the future, there's always futures, in multiples.
Tanja Schindler
Questions answered in this talk
Where can AI have the biggest positive climate impact?
According to Daria Saharova on the 'Can AI Heal the World?' panel at Rise of AI 2022, the biggest opportunities are in the highest-emitting industries: food and agriculture, land use, and the built environment, for example through satellite-based crop and wildfire monitoring and protein-engineered materials replacing petrochemicals.
Does AI's own energy use undermine its climate benefits?
Panelists at Rise of AI 2022 note that data centers and AI infrastructure in Germany reportedly emit more CO2 per year than the entire aviation sector, and call for continued hardware efficiency gains and a common European energy and regulatory framework.
Is Europe competitive in AI compared to the US and China?
Mario Brandenburg argues on the Rise of AI 2022 panel that Europe's positioning around 'ethical AI' is not sufficient on its own, warning that Europe risks dependence on Chinese or American AI unless it also builds competitive products, not just ethics boards.