The compound of technologies we use to call 'AI' promises to revolutionize many sectors. However, there is a substantial gap between what firms say they do with AI and what they actually do with it [1]. Several factors could account for such a gap: the adoption of AI tools is costly since it tends to jeopardize an incumbent's operations [2]; there is a shortage of human capital trained in the area of AI [3]; developing AI applcations may require businesses to cope with ethical/societal implications [4, 5] and regulatory issues [6]. In the context of knowledge intensive industries, there is yet another obstacle to the diffusion of AI, namely, 'people.' While some professionals may be thrilled to integrate AI in their daily work, some others may just feel threatened. This project deals with the distribution of security traders opinions about the impact AI can make on 'trading floors.'
Analyse how the diffusion of AI could be completed on the trading floor.
The resulting dataset (trading_floor.xml) contains 192 responses regarding: the undirected network of knowledge exchange between traders (traders A and B are connected when A says he/she shares technical and industry knowledge with B and vice versa) a trader's opinion about the contribution of AI to his/her productivity and effectiveness in evaluating securities (1 = not at all; 10 = to a great extent). In the datasets, this variable is reported as the node attribute ai. Thanks to the cooperation of the client, you also know the traders' location in the floor. There are six zones, each of which hosts 32 individuals (16 individuals on each side of the zone). The above-displayed picture gives you an idea of the layout of the trading floor. In the dataset, the location of traders is reported as two node attributes, that is, x-pos and y-pos.