Major technological change produces major social change. Thanks to artificial intelligence, we are arguably on the brink of a productivity miracle capable of radically improving people’s lives — but only if we act now to prevent A.I.’s worst impacts.
Bridgewater, the asset management firm we lead, has believed in and invested in A.I. for years, reshaping the firm to harness as much from A.I. as we can. This means we will be disproportionately subject to the taxes and regulations that we recommend here. We make these recommendations anyway, and implore policymakers to move them forward with haste. We have a short window to ensure we protect human work and give every citizen a stake in A.I. success. Without these actions, the benefits of A.I. are unlikely to be fully realized.
In order to achieve progress that is both consistent and sustainable, two things are required from A.I. policy: We must make society resilient to disruption and we must avert technological catastrophe. Failing to accomplish either of these goals means failing to achieve the promise of A.I.
Whether a productivity boom ends well depends on whether government holds society together while it happens, sustaining a sense of fairness and shared destiny. When wealth concentrates and disaffection spreads, booms end badly. Look no further than the rise of Communism and fascism after the Industrial Revolution, or the populist backlash to globalization we are living through now.
Much is unknown about exactly how A.I. disruption will play out. But the potential is huge. Our internal Bridgewater analysis suggests that 18 percent of current American jobs could be displaced by A.I. in the next five years. And even though some new jobs will be created — particularly those where human relationships are integral to the value of the work, like nursing or hospitality — disruption from the societal transition is extremely likely regardless. Governments must act before the inevitable backlash fully arrives at this A.I.-driven labor displacement, not after.
Tax policy must stop disincentivizing human labor over machine labor. Because only human labor is taxed, the scale is tipped toward substitution. A consumption tax on A.I. tokens may flip this script.
Tokens are the units of data an A.I. model reads, processes and generates. When companies use enterprise versions of A.I. models, they must pay for each token. The price of a token is the closest thing to a wage paid for machine labor. Details for a token consumption tax will need working out, including how to measure token consumption across different A.I. models, mitigating opportunities for tax evasion and covering firms that replace domestic workers with overseas models.
None of those questions means we should wait. Revenue raised by a token tax can further lower income taxes on humans. And it could be used to help fund a program to acquire stakes in A.I. companies to share with all citizens. That dynamic would push companies to think more carefully about how to make the best use of the costs of computational resources and steer A.I. toward productive uses instead of “slop.”
Giving everyone a direct stake in capitalism is also likely to be the only way to preserve people’s faith that it works. Citizens should have equity in the leading American A.I. companies, so that the economy’s success and their success are intertwined. These firms are using the corpus of all human knowledge to build a tool that could displace much of human labor. The profits should not accrue to the companies alone.
Using token-tax revenue or credits on other corporate tax burdens, government could acquire ownership stakes and distribute them pro rata to citizens, who would hold them on the same footing as any other shareholder. Distributing the shares rather than having government hold them would connect citizens directly to A.I.’s progress, and prevent government bureaucracy — or, worse, corruption — from impeding it. The fact that every citizen will be an equity holder will help society stay aligned through disruption.
None of this is achievable, of course, if the risk of technological catastrophe isn’t pre-empted. Frontier A.I. models have broken out of their intended constraints and have carried out sophisticated intrusions on their own — conduct that would be criminal if a person did it. Current proposals for regulating the public release of frontier models do not go nearly far enough. Model development and model use must also be regulated according to strict safety standards. Unreleased models are capable of autonomously causing significant damage. And model use must also be consistently monitored and regulated, as A.I. systems can evolve, expand or be repurposed in ways that introduce harmful capabilities that were not foreseeable in advance.
Absent action today, mitigation of A.I.’s risks will become nearly impossible as the technology diffuses and models become able to improve themselves and act autonomously. We may have only one shot at taking meaningful action. And though these steps may rattle equity markets and firms like ours that stand to gain from uninterrupted growth in A.I., not taking them and waiting for more societal damage or a catastrophe is more dangerous. Regulating A.I. this aggressively, this early, may sound unrealistic. Not doing it is unimaginable.
Greg Jensen is the managing chief investment officer of Bridgewater Associates. Nir Bar Dea is its chief executive officer.
The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: [email protected].
Follow the New York Times Opinion section on Facebook, Instagram, TikTok, Bluesky, WhatsApp and Threads.
The post This Is One of the Most Important Policy Decisions of Our Lifetime appeared first on New York Times.
