WHAT DOES HUMAN INQUIRY EVEN MEAN ANYMORE? / THE FINANCIAL TIMES OP EDITORIAL

What does human inquiry even mean any more?

Universities and other knowledge-based institutions will need to be radically redesigned in the age of AI

Nicholas Dirks

© Ewan White

A passage from the French philosopher Michel Foucault’s book The Order of Things keeps resurfacing in discussions of artificial intelligence, and with good reason. 

“Man,” Foucault wrote, “is an invention of recent date. 

And one perhaps nearing its end.” 

His claim was not that human beings might vanish, but that “man” — the modern figure of the sovereign knowing subject at the centre of the production of knowledge — is a fragile historical construction. 

Foucault died decades before the advent of generative AI, but his prophecy no longer sounds preposterous.  

When machine systems can occupy some of the functions previously reserved for “man” as knowing subject — generating text, making inferences, producing outputs that look like the product of human understanding — something like the displacement Foucault anticipated becomes newly plausible. 

Alan Turing anticipated this in 1950. 

He predicted that “at the end of the century the use of words and general educated opinion will have altered so much that one will be able to speak of machines thinking without expecting to be contradicted”. 

The following year he put things more starkly: “If a machine can think, it might think more intelligently than we do, and then where should we be?” 

In an unpublished paper from the same period he went further still, warning that “at some stage therefore we should have to expect the machines to take control”.

Ideas of the human have never been fixed, of course. 

They have shifted before in response both to religious and theological upheaval and to scientific and technological discovery. 

Copernicus shifted our understanding of the Earth’s centrality in the universe, while Charles Darwin revolutionised our sense of humans’ place in relation to the animal kingdom. 

And today, whatever we think of predictions about the coming of the “singularity” or “artificial general intelligence”, we broadly accept that the categories that organise knowledge do not last for ever and that current understandings of what is “human” will be no exception.

AI is not just a technological challenge; nor is it a mere philosophical puzzle. 

It is a major wake-up call for our universities and other knowledge-based institutions. Such institutions embed assumptions — about human values, the sources and limits of expertise, the proper relationship between society, government and knowledge. 

When assumptions about who or what creates, controls and uses knowledge begin to shift, the institutions built upon them come under growing pressure to shift as well.

Like Frankenstein’s monster, AI has moved from the laboratory into everyday life. 

AI systems now influence the information we consume, the social relationships we form, the cultural activities we pursue, the jobs we perform, the medical diagnoses we receive, the education we are offered and the scientific discoveries and policy recommendations that shape our future. 

Emerging technologies such as quantum computing, synthetic biology, advanced robotics and blockchain systems are likely to accelerate these transformative effects.   

While many scientists are excited by the possibility of unprecedented scientific breakthroughs, others are concerned not just about the impact of AI on conventional scientific procedures, but about the “black box” nature of new modes of computational reasoning. 

What happens when new discoveries are based on machine intelligence rather than human understanding, which may become increasingly outpaced by AI?

And what about other institutions of knowledge? 

How will the ubiquitous availability of powerful AI models affect classroom teaching, student learning and assessment, the development of critical thinking, writing and reading as well as statistical and mathematical skills, not to mention the role of faculty and the value of scholarship? 

Will universities, research centres and think-tanks survive intact?  

Many of our institutions have already shown weaknesses and vulnerabilities which the new knowledge economy of AI will only exacerbate. 

It is time to rethink, and in some cases redesign, them. 

Machines will only become more intelligent. 

They will find new ways to outwit humans, and to hide from us how they seek to attain their goals and arrive at their conclusions. 

Humans, however, still have the capacity to set those goals, and to establish the criteria by which we measure success, for machines as well as ourselves. 

To do so we’ll need to find the resolve to sit for the exam rather than either hack the answers or put our heads in the sand.  

The writer is the president of The New York Academy of Sciences and former chancellor of the University of California, Berkeley