Commentary on the pending initial public stock offerings for Anthropic and OpenAI highlight the widespread fears and hopes around artificial intelligence. Not exactly overlooked, but not central in the discourse, is the transformative potential they raise for finance and capital, and through that sector for society as a whole.
The AI commentary has a new quality to it, arising from the concerns and aspirations around AI. Unlike prior new technologies, AI raises a “doomsday” fear, of AI replacing or ruling or perhaps destroying humanity, plus some utopian ideas that it will make life better in yet-unimagined ways. The feelings, perhaps foreshadowed by the growth of social media, are particularly palpable.
The companies’ founders were apparently alive to the implications. Presumably OpenAI’s initial formation as a not-for-profit organization, and Anthropic’s control by its own public benefit corporation, reflect this awareness. Those governance forms are part of a long-gestating movement to somehow reconcile the interests of corporate capital with social and ethical values. That movement may now see its purposes advanced, or set back, as the IPO processes play out.
Another angle on the values concern arises around national security and Sino-American strategic competition. Chinese releases of new AI systems at raise the specter of Communist domination of social processes. OpenAI’s warning about China’s progress might be a patriotic alarm, a move to exploit this public concern for its competitive advantage (in revenues and in funding), or both.
The very philosophy of corporate ownership thus becomes a natural next issue. The old socialist answer was raised by Sen. Bernie Sanders, proposing half ownership of both firms by the government, with stakes held in a sovereign wealth fund. Interestingly, his proposal follows closely on the Trump Administration’s taking a stake in chipmaker Intel, premised on an industrial policy goal of ensuring domestic self sufficiency in semiconductor manufacturing. Do public concerns raised by AI override the tenet of private property?
A more esoterically financial question also addresses the question of property, regarding the nature of share ownership as opposed to contractual debt financing. The magnitude of investment that AI firms will need to develop their systems is staggering. Anthropic and OpenAI are developing new technologies, which normally would be financed by venture capital and other equity ownership stakes. But both firms are borrowing in debt markets too. In traditional doctrine – and tax policy – ownership stakes entitle shareholders to a company’s profits and future growth, while also exposing them to losses and failure. Debt is notionally used for particular business activities, so that bonds and loans have contractual maturities, and compensation comes to lenders in the form of negotiated interest. These insulates (or excludes) bondholders or noteholders from downside and upside variability. Interest payments are treated as a business expense for the company, and tax-deductible in consequence. But essential startup projects, i.e. the data centers and developmental costs of the AI firms, look a lot like founding investments. Should debt-holders who finance them be insulated from up- and down-sides, and should the firms be able to deduct the interest from their earnings? The roles of debt and ownership start to blur.
The question becomes more meaningful as private vehicles have gone far in displacing publicly traded stocks and bank loans. The historical separation of equity and debt stakes – shares versus bonds and loans – assigned clear risks and exposures to the various parties, and clear rationales for tax policy and bankruptcy courts. Now, a family of investment funds can hold all sorts of stakes in a mélange of terms, so that risks and taxes and entitlements are subject to interpretations and manipulation, in a practical sense beyond the legal bickering that has always been part of finance. Who should absorb – or be insulated from – any losses, and is there a public interest in the answer? What role should government have and what should be reserved to investors and entrepreneurs?
These and other issues arise, or are crystallizing, today, with these IPOs. They have not yet been fully engaged by our political / cultural / ideological factions. As those do get involved, they may well invoke questions such as whether America can be a socialist nation, whether we were founded as a capitalist nation, what are proper limits to government involvement in markets, and what economic effects may create public interests that call for government measures.
There are even short term economic possibilities that could raise controversy. What effects will a new investment environment have on retirement plans – in a time when the baby boom population bulge is drawing on them? What taxes can or can’t be charged, and how will the US fiscal position be affected? As deficits and shortfalls grow, expedient measures in such matters could overwhelm any principles that the political partisans may have left.
Our national bedrock is our founding creed: our holding of universal unalienable personal rights and that government exists to secure them, legitimated by consent of the governed. These tenets are abstract and admit of infinite pathways to better realize them – even as they exclude an infinity of government pathways. Debate over which is which will now be stretched, and clarity in the abstract terms, including the fact that they are abstract rather than working prescription, is essential. For instance, we are not necessarily “capitalist,” incidentally a term that draws as much on Karl Marx and on anyone for its meaning. But we do believe in an unalienable right to the pursuit of happiness, which strongly suggests freedom of enterprise as a core value. Government protects rights and does not guarantee welfare – but is there any level of deprivation that inherently restricts rights? Such questions, as we enter into a new age of technology – and also of finance – may yet emerge. As a whole society, as individuals, and in society’s diverse sectors, will we continue to meet our tangible needs in fidelity to these existential tenets?
Common Conviction 