Post by Tristan Zindler

Last week at data:unplugged in Muenster Sascha Lobo argued that we need positive visions for AI. Because if we don't develop them doom is certain. I tried to write one. Less than half of psychology studies replicate. The US Department of Defense started auditing the robustness of social science research. And the scientific corpus we're training AI systems on is full of contradictions. My argument: the relationship between AI and science is bidirectional. Science needs AI to clean up its own mess. And AI needs clean science to become intelligent. New essay: "Poison in the Corpus: The Antidote in Sight" https://lnkd.in/duEX9Xi2