What are academics for? Can we be replaced by AI?

A friend asked me Friday night how I saw social media and contracting as fitting into being an academic.  I answered at the superficial level of what I would do in particular contexts, but (partly because I've also been thinking about universities from the perspective of my local community) I've decided to answer at greater length.

The role of universities

I think an academic is best understood as a constituent part of a university.  Not that there can't be independent academics, but that even they are best understood if we first think about what role universities play in a society.  I think there are four interdependent roles:
  1. Absorbing risk
  2. Transmitting knowledge
  3. Innovating
  4. Forming identities
 Many people put innovation (research) first.  That is certainly the principle criteria by which academics are at least explicitly judged for promotion, but that may be at least partly because it's the easiest criteria to judge by external evidence.  But universities are not the only ones that innovate.  A great deal of innovation is done by the institutions that will exploit it directly: industry and government.

Universities absorb two kinds of risk for a society.  First, universities perform the riskiest, "blue sky", research.  Rather than looking for solutions to particular problems, universities generically push forward the boundaries of knowledge, discovering the truly unknown.  They employ brilliant, creative, well-educated people who know what society already knows, and recognise problems we nearly understand or ought to understand, and invest their effort to be the first to understand them.  Academic prestige is associated with being unique in knowledge held.

Nevertheless, and sometimes neglected internally, the role of universities that dominates our public image is actually fantastically important.  Universities are composed of the people who proved since grade school the best at absorbing, retaining, retransmitting and exploiting our cultural heritage: our knowledge.  You can think of universities as knowledge batteries, or maybe better as capacitors.  Sure, you can have a lot of information stored in databases or libraries, but what is actively in and on the minds of academics is what will be communicated to students, the press, companies, governments–whoever bothers to consult with us.

But we don't only transmit knowledge.  Although it's the vast majority of academics' least favourite part of our job, we also assess how well students have absorbed that knowledge by giving them marks and ranks–what Americans rightly call "grading".  I think part of why we hate grading is that we know we are assessing our own capacity to transmit as much as the students' capacity to receive, but assessing that latter is really a critical social role, and the second part of risk we absorb.  Universities take potential employees and classify them for their abilities to absorb knowledge and follow instruction.  A students' final GPA or degree classification is just a grotesque aggregate assessment.  What often matters more is a student's more detailed profile: which courses they did best in, which courses they did badly in, and the letters of reference their tutors or research supervisors provide about what they are like to work with.

Finally, again more recognised by the public than most academics, universities form human identity.  Here I am stretching into my own current research, but increasingly I think identity is shared mental models:  the people you consider yourself to be like because you can predict what they are likely to do.  Again, we do identity construction in two different ways.  Universities have individual "brand" identities.  We create cohorts of people with similar educations, and age-related clusters of these who know each other.  University-educated people will exploit these ties for the rest of their careers, and institutions will exploit the brands to find particular types of employees and expertise.  But second, we create what used to be known as "a college man."  Pollsters and employers ask as a fundamental description of an adult whether they have advanced degrees, have completed college, have some college, or none.  As much as universities worry about their individual brand, this gross level of categorisation is also strikingly effective.  The university-educated identity creates a bond not only across institutions but across countries.  It rivals and ameliorates the importance of religion, nationality, and employer.

Present and future roles of academics

You can read the roles of individual academics off of the above roles of universities:
  1. Academics retain the knowledge of previous generations, and exploit it to discover new knowledge.
  2. Academics transmit that knowledge, and 
  3. assess other individuals' abilities to retrain and exploit it.
  4. We are also a part of a university's identity.  A university is defined by its portfolio of academics.
Note about the phrasing of my third point: academics not only judge our students, but each other.  We are constantly asked to assess each other not only for peer review of articles and grants, but in helping companies, governments and through the press ordinary citizens assess each others' claims and discoveries.  We are also sometimes called on to assess non academics as consultants or expert witnesses.

So what are the roles of social media and contracting?  They are two mechanisms of transmitting both knowledge and identity.  Every academic has to trade these off with the time they spend in other forms of transmission, learning, and innovation.  I doubt there's a single best answer for how to allocate time; in fact a fundamental principle of biology is that every trade off affords an enormous diversity of equally good solutions.  But as new opportunities and mechanisms for communication, education, and discovery become available, academics as well as students will begin allocating their time and other resources in different ways.

Can universities be entirely replaced by AI and MOOCs?

Not really.

First, MOOCs can only give part of the university identity.  Spending early adulthood colearning with a specific cohort of people will continue to have advantages.  Being known to academics who can write individual letters of reference for you will still matter for some jobs.  Learning how to do research by apprenticeship will also still play a role.  Spending time in what is effectively an incubator coming up with both ideas and relationships, and proving you can meet protracted learning and other goals, will probably both retain value.

But clearly teaching and research done by AI and other IT will have a huge impact on universities and our societies.  Again, where there are trade offs there are a large diversity of possible solutions. Many students will prefer a weaker brand over an enormous financial investment, though those for whom money is less of a barrier may still prefer to signal their commitment to their identity with a massive academic investment.  Also, the monastic value of having a dedicated location and time to learn probably has some value, though in contrast, so does developing the capacity to continue learning throughout a career and lifetime.  But universities also strive to teach that capacity, and will probably increasingly do some of that teaching with MOOCs.

net:art - near in the distance 2 – a multimedia, multinational live performance

But most importantly, societies still need universities to absorb risk.  Students known to specific academics or institutions, who are in turn known to companies and other organisations, will probably continue to be seen as the best vetted, though to be fair the full assessment capacities of big data may surpass these.  Some combination of MOOC ranking and psychological assessment may prove a better predictor of employee behaviour, and even prove effective at recommending the forging of new collaborations and start ups.  Nevertheless, the shared experience of an academic history helps make people into teams who collaborate easily.  But at least as important is the university-based alternate structure for knowledge discovery.  Academic research is driven by individual's curiosity, and their desire for status and uniqueness.  University research will continue to lead to fundamentally different and complementary innovations compared to the pursuit of known practical problems.  Even though our richest companies can and will do a certain amount of blue-sky research in house, the value of universities comes also from their universality.  Multiple disciplines and interests are bound together and cross-fertilised by the bright, open minds we educate.  Countries that give up on supporting that source of innovation will have to exploit its outcomes second hand, with less connection to and understanding of the innovators, their students, and their professors.

Addendum:  here is a great related article from April 2017: In Defense of the Lecture.  I'm doing flipped classroom + remote Q&A sessions Spring of 2017, I will need to blog about that this summer.


Joanna Bryson said…
Argh, there's some really interesting comments about this post on Google+. I guess Google is trying to move us all there since they haven't made the comments come up here.


Here's my own response: (Weirdly +Google+ didn't notify me about your post). You are certainly right that not all knowledge is transmitted through academics, but what is is carefully curated. Many brilliant minds spend a lot of time coming up with the curriculum they will allocate to various degrees. Even determining which departments get space within a campus footprint is hugely controversial, and again reflects an enormous amount of intellectual work and curation.

With respect to the risk element, that's something that struck me while I was pondering how to communicate why the UK's obsession on "impact" was so wrong. In fact, it's not entirely wrong to focus on impact: you do need to counterweight against academics who pursue uniqueness by being intentionally useless. Having academics solve real problems for a society is great. But it's not a unique role of academia, and something terrible is lost if we throw out all the folks pursuing baroque interests for reasons they can't explain. Not every academic needs a theory of academia, but it would be useful if governments had a good one.