The world’s complex challenges need the most effective solutions.
A group of four major companies, including Microsoft, formed a new organization. This group aims to create self-regulation for the AI industry, which is a significant step forward.
There is no certainty about which models will achieve advanced, safe, and reliable AI. Yet, there is optimism among AI founders. They are seeking new governance models.
With some luck, these efforts might lead to significant economic benefits. These gains, driven by advanced AI technologies, could be widely shared and have a positive impact on society as a whole.
Where Does the AI Windfall Go?
Profits from AI companies have the potential to benefit the public in three primary ways: through taxes, charity, and direct company donations.
Taxes are the most substantial method, involving the taxing of capital income—such as AI company profits—and then redistributing the funds via social programs. Governments have various ways to tax this income, ensuring a balanced distribution that supports societal needs.
Charity is another method by which profits can reach the broader public.
Companies like Anthropic actively promote charitable donations by offering matching programs.
For instance, they offer a 3-1 match on donations of shares, up to 50% of an employee’s shares. If an employee donates half of their 10,000 shares, the company will contribute an additional 15,000 shares.
The third way involves AI companies themselves donating a large portion of their profits.
This concept is central to “The Windfall Clause,” a proposal by the Centre for the Governance of AI.
This clause is a voluntary but binding commitment where AI firms agree to donate a set percentage of their profits beyond a certain threshold to charitable entities.
Authors of the “The Windfall Clause” include experts like Cullen O’Keefe from OpenAI, Jade Leung from OpenAI, and Allan Dafoe from Google DeepMind.
Binding Commitments for AI Companies
The Windfall Clause suggests thresholds based on profits as a portion of the gross world product, which refers to the world’s total economic output. If AI proves to be transformative technology, it could generate immense profits.
The tech industry already produces large profits with relatively small workforces compared to older industrial titans like General Motors. AI is likely to continue this trend while potentially replacing some labor-intensive jobs outright.
This shift transforms what would have been wages into revenue for AI companies. Without methods to redistribute this revenue, there could be significant increases in economic inequality.
An example provided by the “The Windfall Clause” suggests the following donation structure:
- 1% of profits between 0.1% and 1% of the world’s economy
- 20% of profits between 1% and 10%
- 50% of profits above 10%
Currently, no company globally, even those valued at trillions like Apple, makes profits large enough to reach these thresholds. However, the proposal aims to establish obligations for extraordinarily successful companies, not to replace taxes for typical companies.
Distribution Questions
Despite these ideas, the proposal does not specify where the donated money should go. Misallocating these funds could lead to negative outcomes, raising crucial questions about distribution. For example, in a global scheme:
- Should all countries receive equal shares?
- Should distribution be based on population?
- Should poorer countries receive more or faster aid?
These questions are crucial to ensure that the redistribution of AI windfalls benefits society fairly and effectively. The approach taken must be thoughtful to prevent exacerbating issues of inequality or inefficiency.
By implementing these three strategies—taxation, charity, and company direct donations—society can potentially share in the prosperity generated by AI advancements.
Each method has its own complexities and potential drawbacks, but together they represent a multifaceted approach to ensuring the broader public can benefit from AI-driven progress.
A Global UBI
If artificial intelligence becomes a transformative force, companies may earn profits equivalent to a large portion of the global economy.
This could potentially result in significant economic disparity. To prevent this, there is a proposal to distribute these profits as a universal basic income (UBI) to individuals worldwide.
This approach aims to ensure that benefits are shared widely, rather than concentrated in the hands of a few.
The idea is to distribute monthly payments to as many people as possible. This would be managed by the company responsible for the AI, in collaboration with host country governments.
To maintain transparency and proper use of funds, regular audits would be conducted. In cases where distributing funds universally is challenging, the focus should be on the poorest countries with robust financial systems.
Why Direct Cash Distributions?
- Reduces Fraud: Direct payments to individuals minimize the risk of funds being mishandled by local authorities.
- Avoids Complex Decisions: It eliminates the need for the company to decide which projects or causes deserve funding.
- Equity Over Taxes: Unlike taxes imposed by rich countries, which may provide minimal benefit to those in need, a global UBI ensures that a significant portion of profits reach the impoverished.
Steps for Implementation
- Collaboration: Work with host country governments to set up and manage the UBI program.
- Audits: Regular audits to ensure funds are used as intended.
- Target Areas: Initially focus on regions with the least resources but sufficient financial infrastructure, like areas using M-Pesa mobile payments.
- Scale: Gradually expand to reach more people as the system proves effective.
Benefits of a Global UBI
- It offers a straightforward way to share the economic gains from AI technology broadly.
- It empowers individuals by providing them with regular income support.
- It creates a more equitable global economic landscape.
A global UBI not only addresses economic inequality but also aligns with a vision of using transformative AI for the greater good of humanity. While establishing such a program will face challenges, it represents a practical and fair approach to wealth distribution in an AI-driven future.
