The four-day workweek and the robot tax: OpenAI's rules of AI capitalism
A ChatGPT developer presented his ideas on how to deal with the devastating effects of artificial intelligence adoption
OpenAI's vision is for AI to make work easier and safer, not just help companies squeeze more out of employees / Photo: Unsplash / Y M
The development of artificial intelligence requires a new industrial policy that puts people's interests first, according to OpenAI CEO Sam Altman. In an April 6 interview with Axios, he said that the transition to "superintelligence" has already begun, and the benefits and risks of this technology without a new set of rules could be very unevenly distributed.
On the same day, OpenAI published proposals on how to change the situation where capital may have too much leverage because of AI. In a 13-page document titled "Industrial Policy in the Age of Superintelligence," OpenAI writes about a public welfare fund, a four-day work week, and an automatic expansion of social support if AI starts pushing humans out of the labor market.
The universal right to AI
OpenAI equates access to AI models with the underlying infrastructure - electricity or the internet.
"You pay your electric bill and plug in whatever you want: computer, oven, kids' toys, whatever. People can come up with ingenious things, I plug them in and I don't think much about it. Although a tremendous amount of work goes into generating electricity, it comes into my house at a surprisingly low cost, when you think about it. Its cost has dropped significantly over time. I think it will be the same with AI
If AI becomes part of labor, education, and everyday services, then lack of access to it becomes not just a technological gap, but a new kind of inequality. OpenAI writes that AI should be accessible to everyone: in terms of cost, connectivity, training and use of such tools. Access points are needed, for example, in schools, libraries, and for low-income communities.
A voice for the workers
OpenAI believes that employees should not just adapt to new tools, but should work with management to influence how AI changes their work. Comagna believes that employees have a better understanding of where AI can improve the quality and safety of work, and where it will only increase workload or worsen working conditions. AI, OpenAI envisions, should make work easier and safer, not just help companies squeeze more productivity out of employees.
"We truly believe in the democratization of AI. We believe in putting this tool in the hands of people and letting them understand it. To determine what rules society collectively wants to set around it and, most importantly, how we're going to integrate it into our lives and get all the incredible benefits that we think are possible"
Dividends of AI efficiency
OpenAI proposes to create a state wealth fund - a mechanism through which a part of the income from AI will be distributed among the population. Thus, AI will enrich not only a narrow circle of capital owners, but also those who have no investments on the stock market.
The company envisions lawmakers working with AI companies to determine how to populate such a fund so that it can invest in long-term diversified assets - not just in AI developers, but in the wide range of companies deploying the technology.
In addition, OpenAI proposes to determine in advance in what form workers should receive "dividends". The plan calls for a four-day work week with no reduction in wages: employers and unions should test whether the same output and service levels can be maintained in less time. If it can, the hours freed up are proposed to be converted into a permanent shorter week, additional paid days off, or both.
OpenAI writes that the "efficiency dividend" can take other forms: higher pension contributions from an employer, more coverage for medical expenses, or support for care for children and elderly relatives.
Social guarantees
OpenAI proposes to tie support measures to the state of the labor market. If AI starts displacing workers, temporary relief measures - extended unemployment benefits, quick cash payments, wage insurance, and training vouchers - should be triggered. Such measures should not be triggered manually, after much political wrangling, but when predetermined economic thresholds are reached.
OpenAI does not propose to extend social support indefinitely: the paper emphasizes that measures should be temporary, targeted and commensurate with the scale of the shock. When the market situation stabilizes, the support will be gradually phased out.
Taxes on robots
AI could change not only the labor market, but also the structure of tax revenues. OpenAI writes that as automation progresses, some revenue will shift from wages to corporate profits and capital gains. The government will receive less money from wages and payroll taxes - which is what the U.S. currently uses to fund Social Security, public health care for the poor, and other government programs.
OpenAI therefore proposes shifting the tax base: relying less on labor income and more on corporate profits, capital gains, and possibly taxes related to automated labor. Such reforms should be accompanied by tax incentives tied to wages: this would encourage companies to retain, retrain, and invest in employees, similar to existing R&D tax credits.
This article was AI-translated and verified by a human editor
