Prize Question: Openness vs. Arcane Knowledge in the 21st Century

In 1798, the Königliche Societät der Wissenschaften (Royal Society of Sciences) in Göttingen published the following prize question in the Reichs-Anzeiger: “How can the advantages that are possible through the wandering of craftsmen be promoted and the disadvantages that occur be prevented?” (“Wie können die Vortheile, welche durch das Wandern der Handwerksgesellen möglich sind, gefördert und die dabei vorkommenden Nachtheile verhütet werden?”) Twelve prize papers were examined, amongst others by Georg Christoph Lichtenberg and Johann Christian Blumenbach. The first prize was awarded to a certain Karl Friedrich Mohl, honourable mention was given to the treatise by Johann Andreas Ortloff. The two selected treatises were published in 1798; the Bayerische Staatsbibliothek kindly put a digital copy of them online in February this year.

The background to the prize question is a conflict that we are still familiar with today; in contemporary terms it can be described as the tension between openness of knowledge vs. secrecy as a civic duty. In the 18th century, itinerant journeymen were largely responsible for the international transfer of knowledge and technology. As is clear from the “zwo Preisschriften”, the Central European nation states (and here especially the small German states) expected advantages from the migration of travelling craftsmen. The perfection of their expertise, knowledge about specific products of other countries or about distribution channels, the acceleration of circulation, the ability to innovate: these were the profits seen for the individual states, while the guild-organised journeymen were seen as vehicles of knowledge transfer. In other words: A “brain gain” was expected from the migration of the craftsmen. Mohl also names the disadvantages: poorly trained or ill-prepared craftsmen were unable to acquire specialist knowledge; the manufacturers and factory owners in the countries visited thwarted the transfer through their insistence on arcane knowledge.

With ChatGPT-4, the question of the 18th century arises again: knowledge transfer via labour migration no longer exists today; in the meantime, all developers involved in the creation of AI tools are contractually bound to secrecy for many years. But openness is a principle that must be defended against venture capital-funded companies like OpenAI. ChatGPT was trained, among other things, on special datasets whose creation was financed by taxpayers’ money and which, for example, are put in the public domain online by cultural heritage institutions; open source code — such as from GitHub — was also incorporated free of charge; only in this way is it possible to develop simple applications quickly and effectively … and that seems magic (“Any sufficiently advanced technology is indistinguishable from magic.”). ChatGPT-4, on the other hand, is neither free of cost nor open freeware. Obviously, it is a valuable tool; however, it carries with it the stigma that it requires the intellectual expropriation of millions of developers and is under the complete control of a third party. In contrast to free software, there is not even any source code to examine. ChatGPT-4 is a system that is as completely closed as one can currently imagine: An absolute black box. Not OpenAI, but ClosedAI.

Therefore, the MMK project invites everybody to answer to the following prize question: “How can the advantages that have become possible through open source and open access be promoted and the disadvantages that occur be prevented?” The prize papers can be published here as answers to this blog post.