MIT

Position Title/Short Description

Full Text of JOE Listing:

Postdoc Job Posting:
MIT Sloan / MIT Computer Science & Artificial Intelligence Lab

Position: Multiple Renewable Full-time Post-Doctoral Associate positions, available starting in 2024 for 2-3 years

Employer
MIT FutureTech(futuretech.mit.edu)at:

MIT Sloan School of Management’s Initiative on the Digital Economy (IDE) and
MIT Computer Science and Artificial Intelligence Lab (CSAIL)

Location
Cambridge, Massachusetts, USA

Job Description
Dr. Neil Thompson, Director of MIT FutureTech, seeks to hire Postdoctoral Associates for the following research projects:

(i) AI Scaling and its Impacts on what can be Automated
The ever-growing AI models being released by OpenAI and its other competitors are a
natural consequence of AI scaling laws, which characterize a potential roadmap for the
path of growth in AI capabilities as a function of the dedicated data and computational
resources (see https://ide.mit.edu/insights/whats-next-ai-scaling-and-its-implications/)
In this work, we will map the economic consequences of these scaling laws, in particular
the capabilities and skills that these models will acquire and the automation
consequences (e.g. in robotics).

(ii) AI Bottlenecks
Computer scientists and economists disagree profoundly about the extent to which AI may
lead to the acceleration of growth in aggregate productivity and in consumer welfare
across advanced economies. On the one hand, computer scientists point to the rapid
improvements in the performance of AI systems captured, for instance, by the scaling laws
that have governed these improvements over the past decades. On the other hand,
economists emphasize the presence of bottlenecks, such as production complementarities
with other scarce factors, that may diminish productivity gains. In this work, we will
empirically explore which of these bottlenecks do, or do not, have the potential to slow
the pace of automation or stymie its benefits.

(iii) Analyzing AI Job Automation at the Process Level
Current analyses of job automation focus on task level feasibility. But such analyses
ignore the interdependencies between tasks. If a task is a bottleneck in a production
process, then automating it can increase productivity across many workers. Conversely,
automating some others tasks may provide almost no benefit. In this joint work with
Martin Fleming (former IBM chief economist) and Christophe Combemale (CMU professor) we
explore how automation is occurring in the context of business processes and how this
changes conclusions about the Future of Work.

(iv) Estimating the Productivity Effects of IT Improvement
Measuring the effects of IT on firm performance can substantially understate the real
effect of IT (“You can see the computer age everywhere but in the productivity
statistics” -Solow). We believe that this is in large part due to poor measures of
quality of IT inputs: when researchers measure IT inputs using spending without properly
accounting for changes in IT performance(see https://arxiv.org/abs/2206.14007) when
important categories of improvement are missing (algorithm improvement).
(see https://ide.mit.edu/wp-content/uploads/2021/09/How_Fast_Do_Algorithms_Improve.pdf).
We are seeking an IO economist with an interest in using Census data (or similar) and
FutureTech’s detailed data sets on IT progress (see https://futuretech.mit.edu/community-
resources) to re-estimate these effects. The postdoc may also apply some of our new
research on the potential of quantum computing (see https://arxiv.org/abs/2310.15505)for
firms.

In addition to researching these issues, the post-doc will be expected to become a team leader, supervising and mentoring younger students.

Qualifications
• PhD or equivalent in economics or related field
• Demonstrated expertise in computing / AI
• Interest in working in an interdisciplinary lab with computer scientists and engineers
• Strong data analysis skills, including statistical / econometric training
• Proficiency with a modern programming language , e.g. Python, R
• Solid English communication skills (verbal and written)
• Desire and skill (e.g. organizational abilities) to work in teams

The post-doc’s primary appointment would be at the MIT Sloan School of Management, and would have a secondary appointment at the MIT Computer Science and Artificial Intelligence Lab.

Salary
$70,000 per year, plus benefits.

To apply
By January 3rd 2024, through JLE with the following materials:
(i) CV
(ii) Job market paper (or other sample of previous research)
(iii) Sample of code
(iv) Cover letter discussing the applicant’s experience and which research projects would be
of interest

Selected candidates will be first interviewed in January via zoom.

About MIT FutureTech
Led by Dr. Neil Thompson (see http://www.neil-t.com/), MIT FutureTech (see https://futuretech.mit.edu/)is an interdisciplinary group of computer scientists, engineers, and economists who study the foundations of progress in computing and Artificial Intelligence: the trends, implications, opportunities and risks.

Our work is supported by grants from Open Philanthropy, the National Science Foundation, Microsoft, Accenture, IBM, the MIT-Air Force AI accelerator, and the MIT Lincoln Laboratory.

The following are examples of our recent work:
• How industry is dominating AI research
(see https://www.science.org/stoken/author-tokens/ST-1055/full

• The Quantum Tortoise and the Classical Hare: A simple framework for understanding which
problems quantum computing will accelerate (and which it will not)
(see https://arxiv.org/abs/2310.15505)

• A workshop on AI scaling and its implications for AI development, automation, and more
(see https://futuretech.mit.edu/workshop-on-ai-scaling-and-its-implications)

• The Great Inflection? A Debate About AI and Explosive Growth
(see https://asteriskmag.com/issues/03/the-great-inflection-a-debate-about-ai-and-
explosive-growth)

• There’s plenty of room at the Top: What will drive computer performance after Moore’s
law?
(see https://www.science.org/doi/10.1126/science.aam9744
?ijkey=v.ccqFuTmgrY6&keytype=ref&siteid=sci)

• Deep Learning's Diminishing Returns: The Cost of Improvement is Becoming Unsustainable
(see https://ieeexplore.ieee.org/abstract/document/9563954)

• America’s lead in advanced computing is almost gone
(see https://gppreview.com/2023/02/28/americas-lead-in-advanced-computing-is-almost-
gonepart-1-systems-and-capabilities/)

• The Decline of Computers as a General Purpose Technology: Why Deep Learning and the End
of Moore’s Law are Fragmenting Computing
(see https://cacm.acm.org/magazines/2021/3/250710-the-decline-of-computers-as-a-general-
purpose-technology/fulltext)

• How Fast Do Algorithms Improve?
(see https://ide.mit.edu/wp-content/uploads/2021/09/How_Fast_Do_Algorithms_Improve.pdf)

About Dr. Neil Thompson, the Director of MIT FutureTech - futuretech.mit.edu/team/neil-thompson(see https://futuretech.mit.edu/team/neil-thompson). Dr. Thompson is the Director of MIT FutureTech. Prior to starting FutureTech, Dr. Thompson was a professor of Innovation and Strategy at the MIT Sloan School of Management. His PhD is in Business & Public Policy from Berkeley. He also holds Master’s degrees in: Computer Science (Berkeley), Economics (London School of Economics), and Statistics (Berkeley). Prior to joining academia, Dr. Thompson was a management consultant with Bain & Company, and worked for the Canadian Government and the United Nations.

About the MIT Initiative on the Digital Economy (IDE) - ide.mit.edu (see https://ide.mit.edu/) The MIT Initiative on the Digital Economy is a team of internationally recognized thought leaders and researchers examining how people and businesses work, interact, and prosper in a time of rapid digital transformation. It is housed at the MIT Sloan School of Management.

About the MIT Computer Science and Artificial Intelligence Lab (CSAIL) - www.csail.mit.edu (see https://www.csail.mit.edu/). CSAIL is one of the world’s top research centers for computer science and artificial intelligence (currently ranked #1). It has hosted 9 Turing awards winners (the “Nobel Prize of Computing”) and has pioneered many of the technologies that underpin computing.