History and Artificial Intelligence
The end of History?
A survey in
late 2025 has suggested the human roles most at risk from redundancy from
technological innovation. The second job on the list is that of the historian[1].
It seems to me that this reveals an important misunderstanding between the
function of research, from the true purpose of the historian and genealogist.
My own
research has relied increasingly on technological intelligence over recent
years, to exponentially enhance the process of assembling factual data. Factual
research can be undertaken far more efficiently than Martin Farndale could
achieve when he started this genealogical journey. The interpretation of increasingly electronic
databases of source material has fundamentally changed the process of research.
Indeed it is often possible to get accurate clarification about a particular
time and place in history by simply asking a question of google, who when asked
will provide sources for its answer which can be verified. Google Gemini, Chat
GPT, or Microsoft Co-pilot will now generate unique new images of historical
events through carefully constructed descriptions within a few minutes, which
can be a powerful tool for historians to use in constructing a narrative. The
picture which introduces this webpage was drawn by Microsoft co-pilot, based on
my short description. Google’s Notebook LM can produce a podcast discussion to
simplify and summarise historical text which is also a helpful tool. Like all
aspects of human pursuit, artificial intelligence seems to be at the cusp of
extraordinary advance in ancient processes undertaken by humans.
The task of
historical research is about to change, fundamentally.
Yet research
is not the real purpose of a historian. Being a historian is a very human
endeavour, to draw out human purpose, emotion, and meaning from the underlying
research. It is also to engage in that most human of activity, to tell stories.
Humans have told stories which have drawn out the richness of their ancestry
since they developed the cognitive capacity to exchange ideas, swap gossip and
assume the characteristics that made us human. Podcasters such as Tom Holland,
William Dalrymple and Alice Loxton continue that tradition today.
The stage of
evolution in our cognitive journey means that, with technical assistance, we
can now ask an artificial intelligence tool to write a play about a subject in
the style of William Shakespeare, or to write a short history in the style of a
historian, on a particular historical theme. Yet what the technology will then
do, is to reproduce material once produced by humans, in a particular human
style. This is not originality, which can only be achieved by a member of the
species homo sapiens, using a particular skill, to draw out his or her
own meanings and significance.
It may be
that, in time, technology will get very close to mimicking that human ability.
However only humans can create original ideas in the first place. Once we
accept that the artificial reproduction of original human endeavour is the end
point, humans will become unthinking, losing the ability to themselves
understand and draw meaning from the historical journey we have all been
following. The array of facts can be assembled with greater efficiency than
ever. It is for humans to draw deep and emotional meaning from those facts. If
we deceive ourselves that artificial intelligence can replace that most human
of skills, and allow historians to be replaced by something mechanical, then we
will have forgotten what it is to be human.
Unless we
accept our future roles as automatons, the Historian should be last on the list
for redundancy by artificial intelligence.