Andreas Hepp on ChatGPT and pioneer communities

On the occasion of ChatGPT’s third anniversary, ZeMKI member Andreas Hepp shares his expertise on how the visions of pioneering communities have shaped ChatGPT’s development and social positioning.

ChatGPT brings the visions of the 1980s – with all their problems

The vision of having AI as a “talking desktop” and operating AI-based “narrowcasting” dates back to the 1980s: personalized content for everyone, as in conversation. It then took four decades for these visions to be technically implemented and seep into people’s everyday lives. However, the problems with the idea remain: AI technology is becoming deeply woven into people’s lives, but the focus is not on the common good and shared interests, but on the “user” as a data generator and commercial object.

Current ideas go much further

At present, pioneering communities and technology developers are envisioning an even more radical spread of AI: in organizations, for example, language models are to continuously monitor all communication—via email and voice. The aim is not only to automatically generate protocols and to-do lists. The idea is also to network beyond hierarchies, discover innovation clusters, and much more. The old vision of the cybernetic organization is coming, and from there it is only a small step to think such visions on the level of entire communities and countries.

The AI vision itself is crucial—once the technologies are here, it will be too late.

We are used to only dealing with technologies once they are here. People only started to think critically about platforms and social media after they became widespread, not when they were first conceived. But once they are widespread, it is too late. We should take the visions of pioneering communities and early development communities much more seriously and ask ourselves: Which technologies do we want that suit us? And, very currently, which AI would really be productive for our society?