A Prodigal Son Returns

Almost two decades after dropping out of Harvard to found Facebook, Meta CEO Mark E. Zuckerberg returned with Priscilla Chan ’07, bearing half-a-billion dollars and a new Harvard institute to match. The Kempner Institute will study natural and artificial intelligence, operating at the confluence of two fields that have often been studied in isolation.

This institute is an extraordinary development for Harvard and the world. For Harvard, which has sometimes lagged behind more traditional leaders in AI, this is an opportunity to leap forward in an important area of research. Harvard is uniquely well-suited to study the intersection of natural and artificial intelligence, and this work has the potential to significantly improve the cutting edge of AI.

The world, meanwhile, will benefit from the institute’s fundamental research. Insights into the brain could one day be leveraged for clinical application. Insights into machine learning theory — the type that is often too abstract to inspire enough attention in industry — will one day improve the interpretability and performance of one of the most powerful tools we have for shaping the world around us. Finally, giving Harvard’s talented technical students more opportunities to develop expertise in machine learning has the potential to redound huge social benefits.

We acknowledge concerns that the institute seems broadly unconcerned with exploring pressing and complex issues in the field of AI ethics. Machine learning techniques, for example, have at times exhibited racist and sexist tendencies. More generally, the prospect of Artificial General Intelligence whose values do not align with that of humans may represent an existential risk.

While the Kempner Institute’s work of broadly increasing understanding may obliquely improve our ability to regulate AI, it’s true that these ethical issues deserve to be tackled directly. Harvard should fund an AI ethics institute that can centralize and add to existing efforts across the University ranging from Professor Cynthia Dwork’s research on information privacy to the Safra Center for Ethics (funded by another billionaire!) and our Embedded EthiCS program.

But these concerns don’t warrant icing the Kempner Institute. In no other field would we wish to require that every new initiative include space for researching ethics. Instead, we tend to require them to adhere to ethical principles promulgated elsewhere. This is the approach we should take in AI as well.

In fact, there may be a silver lining in Zuckerberg and Chan’s neglect of AI ethics research. These donors have demonstrated great expertise in and passion for the task of advancing technical understanding — which is, appropriately, a focus of this institute. Are they as adept at using this technological progress ethically? Perhaps not. Are they really the donors we want funding Harvard’s exploration of fundamental questions of right and wrong in the use of AI?

We acknowledge, too, the obvious objections to a broader University funding model that implicitly gives billionaire donors influence over which questions get resources and, to some degree, how they are studied. We wish the University could do just as much good without taking Zuckerberg’s money and the risk of undue influence it entails — without letting the fox into the henhouse.

But if Zuckerberg wants to devote resources to the study of these questions, he will. We are concerned about the potential for unethical approaches to this research. Would we rather it take place in the confines of private industry or in the relative (if still often limited) transparency of academia? Conducted by direct reports to a donor we may distrust or by professors at Harvard, accustomed as they are to decentralized autonomy and academic freedom?

This logic is not absolute. The mantra of harm-mitigation would not justify an institute committed to research we find unethical on its face. But the Kempner institute will do necessary research with extraordinary potential for good. If it has potential for abuse, that is an argument for Harvard to step up to the task. We don’t require absolute faith in Harvard’s upstanding character (far from it): only a hard-headed examination of the available alternatives.

Of course, this puts a great burden on our researchers to hold themselves to the highest ethical standards in an area where doing so requires particularly critical thinking and unusual restraint.

It can be easy to lose, amid the righteous imperative to do things correctly, the righteous imperative to do — to strive, to seek, to find. These researchers stand on one of the great frontiers of human knowledge in the 21st century. We can’t wait to see what they learn.

This staff editorial solely represents the majority view of The Crimson Editorial Board. It is the product of discussions at regular Editorial Board meetings. In order to ensure the impartiality of our journalism, Crimson editors who choose to opine and vote at these meetings are not involved in the reporting of articles on similar topics.

Have a suggestion, question, or concern for The Crimson Editorial Board? Click here.