Travail de recherche/Working paper
| Résumé : | While the adoption and use of artificial intelligence (AI) are now significant, overall adoption rates remain low (Kazakova et al 2020; McElheran et al 2023; Zolas et al 2020). One recent line of work has argued that AI represents a type of general-purpose technology (GPT) (Bresnahan and Trajtenberg 1995; Trajtenberg 2019; Furman and Seamans 2019), and so requires significant downstream innovation to adapt general solutions to unique user needs (Brynjolfsson, Rock, and Syverson 2019, 2021; Goldfarb, Taska, and Teodoridis 2023). At the firm level, this will require a combination of business process innovation but also the development of software that will fit the unique needs of firms. This need for complementary adaptation and innovation has historically proven to be difficult (Bresnahan and Greenstein 1996). |
