## Cafe Economics — Embracing Artificial Intelligence

## Source details

**Canonical URL:** [Cafe Economics — Embracing Artificial Intelligence](https://www.imf.org/-/media/files/publications/fandd/article/2023/september/cafe-economics-joshua-gans.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/fandd/article/2023/september/cafe-economics-joshua-gans.pdf.md)
- [Structured JSON version](/-/media/files/publications/fandd/article/2023/september/cafe-economics-joshua-gans.pdf.json)

---

### Key distinction from previous automation
- Previous waves of automation focused on "predominantly physical tasks," whereas AI "tends to provide the potential to automate cognitive ones."
- Past automation also automated cognitive intent and application embedded in physical tasks; information technology automated computation—also a cognitive process.
- The main difference attributed to AI is "the ability of a machine to interact in variable and non-standardized environments."
- Example contrast: Whereas machines could pick up "a specific object in a specific location," AI "holds the promise of picking up and moving a random object that is not in a particular location."
- Concluding characterization: AI’s distinguishing property is "the ability to automate tasks in more varied contexts."

### Near-term effects (next five years)
- Evidence to date: AI tools in workplace tasks "have often involved tools that allow people with lower skills and less experience to perform at the level of those with higher skills and more experience."
- Example: Demand prediction tools in Tokyo allowed "less experienced taxi drivers to waste less time looking for riders," while doing "little to improve the productivity of experienced drivers."
- Projected impacts over the next five years:
  - "Skill premia in certain occupations will be reduced."
  - "Employment opportunities will open up for a wider set of people."
  - Effects expected: "increase productivity, raise employment, and lower income inequality—at least at certain ranges of the income distribution."
- Beyond the next five years: "it is harder to forecast."

### Overhyped concerns and adoption speed
- Generative AI successes cited: LLMs "are able to do writing tasks faster and better than people"; generative art "can produce images that may have taken days or longer"; "the boost to coding tasks has been phenomenal."
- Public reaction: Success has generated "an unusual amount of concern and worry about the adverse consequences that might follow from the use of generative AI."
- Gans’s assessment:
  - The angst is "overhyped" and constitutes "more vague speculation than something involving clear or even less clear evidence."
  - He argues that precaution must be balanced against "the losses from slow adoption for productivity."
  - Learning about adverse consequences requires putting AI "in the hands of people."
  - Policy implication he advances for the current moment: "letting it rip" is the right policy.

### Policy guidance and recommendations
- Monitoring and experimentation:
  - If adoption is rapid, "it behooves us to monitor for those consequences, identify their causes, and consider experimentation with policy interventions that can mitigate them."
  - "Speed has benefits but also means we have to work more intensively to ensure the best outcome."
- Competitive provision and trade:
  - AI "needs to be as competitively provided as possible and available as widely as possible (through trade) so that it can be deployed wherever it can increase productivity in the world."
  - Framing: The mission for organizations should mirror "programs to encourage information technology and internet access."

### Ethics, discrimination, and regulatory responsibility
- Ethical complexity:
  - "The ethical issues are far from clear. Even issues of discrimination that might arise are complex."
  - He suspects "AI will be easier to reprogram to be less discriminatory than people are."
- On suitable regulators or ethical overseers:
  - If ethical issues were clear, an existing institution "whether legislative or legal" might hold ultimate authority, but "the ethical issues are far from clear."
  - Therefore, AI "does require strong policy guidance to induce the necessary changes."
- Role of institutions:
  - Organizations aiming to facilitate growth and prosperity (such as the IMF) should promote widespread competitive access and trade in AI to maximize productivity gains.

*SEPTEMBER 2023 — F&D Cafe Economics: Embracing Artificial Intelligence (Interview with Joshua Gans).*

---


_Source: https://www.imf.org/-/media/files/publications/fandd/article/2023/september/cafe-economics-joshua-gans.pdf_