Abstract:

This presentation investigates the potential influence of Artificial Intelligence (AI) as a technology designed to augment research and boost the rate of new ideas production. As such it may help answer questions on whether and how AI can accelerate economic growth - to even cause a growth explosion - and counter the possible adverse consequences of a declining population on economic growth. The paper focuses on the Ideas Production Function (IDF) as the key growth-driver, considering that AI can automate the discovery of new ideas. Various forms of an IDF incorporating AI are discussed. It is shown that while AI is recognised as a tool that enables higher efficiency in the production of knowledge, so far in the literature this is done with but  no law of motion. Hence the further explores how AI can augment the IPF and raise R&D productivity by addressing this lacuna. Two IDF's with laws of motion are proposed: one that specifies AI as a research-augmenting technology, and one that specifies AI as a facilitator of innovation.  These IPFs are calibrated using specific parameter values obtained from the literature including projections for population growth rates in the U.S. A central finding is that AI alone may not be sufficient to accelerate the growth rate of ideas production indefinitely. The growth rate of ideas closely follows population dynamics in the long-run, despite exhibiting a short-run overshoot. This overshooting shows explosive growth that subsequently declines rapidly. The size and duration of this overshoot vary, being substantial but short-lived (around 10 years) when fishing-out effects are present in traditional ideas production alongside positive spillovers in AI ideas production, and smaller but longer-lasting when positive spillovers exist in both areas. Therefore, while AI offers promising avenues for enhancing innovation in the short term, its long-term effect on sustaining an accelerated pace of ideas generation appears fundamentally constrained by demographic trends.

This presentation will draw on:

Naudé, W., Gries, T. & Dimitri, N. (2024). Artificial Intelligence: Economic Perspectives and Models. Cambridge: Cambridge University Press.

Almeida. D., Naudé, W. And Neves Sequiera, T. (2024). Artificial Intelligence and the Discovery of New Ideas: Is an Economic Growth Explosion Imminent? IZA Discussion Paper No. 16766,  Bonn: IZA Institute of Labor Economics.

Gries, T. And Naudé, W.  (2022).  Artificial Intelligence in Economics, Journal for Labour Market Research, 56 (12).

The public research seminars of the Department of Economics and Finance (DEF) at Tallinn University of Technology  usually take place on the second and fourth Wednesdays of the month both in online  and onsite format, unless announced otherwise. The seminar will last one hour, presentation will last approximately 45 minutes followed by 15 minutes of discussion. The seminars are held in English. Questions about the seminar can be sent to the seminar coordinator Karsten Staehr: karsten.staehr@taltech.ee.