I read an interesting Substack this morning by Stephen Kinsella, a professor of economics at the University of Limerick. He argues that AI wealth is beginning to reshape philanthropy, but it may also reshape economics.

The argument is that the resources now available to organisations such as OpenAI and Anthropic mean that they can fund economic research on a scale comparable to, and maybe way beyond, established public research programmes. And there is evidence that they may not simply finance individual academics. As is the way with the newly wealthy, many of whom seek to establish a legacy for themselves, those behind these organisations, and others like them, may already be creating institutions. These will assemble concentrations of talent, and will seek to influence which questions economists study.

Stephen Kinsell argues that this matters because academic attention is scarce. There need be no censorship and no deliberate suppression of alternative views. If AI research offers money, intellectual excitement, institutional prestige and career opportunities, economists will naturally migrate towards it. And in this case "it" means AI itself. Research into inequality, climate, health, migration and other issues of much greater social significance may consequently receive less attention.

Kinsella offers a contrast with the Institute for New Economic Thinking that is revealing. After 2008, INET sought to challenge economic orthodoxy largely by funding alternative ideas and researchers. Its impact was worthwhile, but it did not transform economics. The emerging AI institutions may have considerably greater influence because they are changing the institutional structures within which economic ideas are produced.

And let's be clear, this matters. We have already seen the enormous impact of funding with less-than-benign intent on the economics profession. The Koch brothers and others funded people like the late Professor James Buchanan, who created the idea of public choice theory, with the intention of deliberately undermining the process of government and democracy itself as part of the neoliberal agenda. And let's not forget that much of that agenda was promoted by quasi-think tanks, or ‘junk tanks', as George Monbiot calls them, often with supposed charitable status and funded by entrepreneurs with deeply right-wing inclinations.

I am not suggesting that everyone in AI has such a political orientation, but there are some deeply worrying signs that there are such relationships, and certainly sufficient of them to massively distort the issues that economists might study and the orientation of the views that they might be expected to deliver. And let me be candid, given the market-oriented thinking of most economists, they are very little better than scribes for hire in most cases, and the outcome of this funding is all too obvious to predict.

I stress that the last two paragraphs are my interpretation of this situation, and do not reflect what Stephen Kinsella said. His opinion seems slightly different. He believes the paradox is that this funding could eventually create the conditions for an intellectual revolt. He suggests that every orthodoxy develops blind spots. If AI economics becomes the new orthodoxy, the neglected questions may eventually generate its challengers.

Stephen Kinsella's questiion in that case is not whether AI companies should fund economic research. Instead, he asks who will fund the heretics when the institutions financing economic research might also have an enormous stake in the technology economists might be asked to understand?

I think that is a good question. I might just broaden it out rather more, when economics already fails to understand, let alone answer, so many of the key questions to which answers are required at present.

PDF of article