We are finally starting to learn how to give a post hoc narrative thick description of what should have been visible in the gathering that brings a thing together (similarly, after the shuttle’s explosion a tough inquiry was pursued). And yet we still don’t know how to assemble, in a single, visually coherent space, all the entities necessary for a thing to become an object…when we have learned how to do that, we might finally get our (material) materialism back – and our cosmic things to boot. That’s when the plot will really thicken.
Latour, 2007, Can we get our materialism back, please?, in: Isis, p. 142
Summary
Can we design a machine that creates successful inventions?
This page picks up this impossible-sounding question. Leveraging dormant Latourian and other under-specified science and technology studies agendas, I illustrate how we might synthesize our knowledge of science and innovation with a view to assembling its objects in advance (rather than disassembling them in retrospect).
The proposed output of this synthesis – a mathematical game that can reliably predict the future of technology.
In my view, the AI model of training algorithms to mimic the human discovery process is not the only means towards this goal. Could we not also use our formal knowledge of science and innovation to design such an algorithm?
Notes:
AI and scientific discovery bibliography
Ex-ante technological determinism, or where to place bets
Please contact me if you would like to know more.
Dr. William Burns PhD MSc
Email: william@resorg.news
Science Think Tank 