Research Grants
Research summary:
Prior related work:
Robin Hanson's "Economic Growth Given Machine Intelligence".
Existing work on this project:
A rough draft (though with sections missing) already exists. However, the paper is in hiatus for lack of free researcher-hours. It also requires a few decisions as to the nature and complexity of the mathematical models to be described. Since much of the work has been done, we may get good risk reduction returns on hours by finishing this paper.
Target dates for:
Total Budget: $3,600
How research costs are estimated:
- Person-months for research and writing: 1.5 (obtained by taking our standard estimate[1] of 4 person-months per journal paper, multiplying by 1.5 because doing mathematical modeling is more time-intensive than just writing text, but reducing the number to a quarter of its full value because most of the work has already been done).
- Dollars per person-month[2]: $2,400.
[2] This billing rate reflects an estimate of financial outlays for SIAI to create the equivalent of one full-time skilled researcher-month, including stipend or hosting expenses, workspace, and administrative or management time, and other supporting expenses. Actual person-months may be greater or lower depending on the labor mix for a particular project, with shortfalls made up from general funds. This rate is not reflective of the money researchers could earn in the competitive labor market. Think of this as a matched donation. You donate the living expenses; our researchers donate the surplus value of their labor.
How this paper will help reduce existential risk:
- The paper will explore economic models of takeoff speeds and mechanisms. It will thereby attempt to translate the dynamics of AI growth into the language of economics, specifically that of endogenous growth models. Translating existing questions across disciplines in this manner can often spark new insights by bringing a question into contact with an established but as yet untapped library of intuitions and predictive models.
- Accurately modeling AI takeoff speeds and mechanisms is critical to understanding what kind of technological singularity may occur and what types of safety mechanisms may be effective.
- Publication of this paper would help build an academic consensus that hard takeoff scenarios are at least plausible. This would aid good decision-making, specifically by making it clear that hard takeoff scenarios need to be considered, and that safety under a hard takeoff scenario requires planning well in advance of the actual takeoff.
Human capital benefits, or network benefits (Will writing this paper help new Visiting Fellows become familiar with key research domains? Will it help create relationships with outside co-authors? Will it give folks interested in existential risk entry into new communities where valuable contacts may be found?):
-
Writing the paper will help Visiting Fellows become acquainted with growth economics.
- It may also create the opportunity for SIAI people to come into contact with economists and with other academics interested in forecasting the pace of technological development.
Donate Online
Credit card transactions are securely processed through PayPal.